[{"data":1,"prerenderedAt":1267},["ShallowReactive",2],{"/en-us/blog/tags/startups/":3,"navigation-en-us":19,"banner-en-us":437,"footer-en-us":449,"startups-tag-page-en-us":660},{"_path":4,"_dir":5,"_draft":6,"_partial":6,"_locale":7,"content":8,"config":10,"_id":12,"_type":13,"title":14,"_source":15,"_file":16,"_stem":17,"_extension":18},"/en-us/blog/tags/startups","tags",false,"",{"tag":9,"tagSlug":9},"startups",{"template":11},"BlogTag","content:en-us:blog:tags:startups.yml","yaml","Startups","content","en-us/blog/tags/startups.yml","en-us/blog/tags/startups","yml",{"_path":20,"_dir":21,"_draft":6,"_partial":6,"_locale":7,"data":22,"_id":433,"_type":13,"title":434,"_source":15,"_file":435,"_stem":436,"_extension":18},"/shared/en-us/main-navigation","en-us",{"logo":23,"freeTrial":28,"sales":33,"login":38,"items":43,"search":374,"minimal":405,"duo":424},{"config":24},{"href":25,"dataGaName":26,"dataGaLocation":27},"/","gitlab logo","header",{"text":29,"config":30},"Get free trial",{"href":31,"dataGaName":32,"dataGaLocation":27},"https://gitlab.com/-/trial_registrations/new?glm_source=about.gitlab.com&glm_content=default-saas-trial/","free trial",{"text":34,"config":35},"Talk to sales",{"href":36,"dataGaName":37,"dataGaLocation":27},"/sales/","sales",{"text":39,"config":40},"Sign in",{"href":41,"dataGaName":42,"dataGaLocation":27},"https://gitlab.com/users/sign_in/","sign in",[44,88,184,189,295,355],{"text":45,"config":46,"cards":48,"footer":71},"Platform",{"dataNavLevelOne":47},"platform",[49,55,63],{"title":45,"description":50,"link":51},"The most comprehensive AI-powered DevSecOps Platform",{"text":52,"config":53},"Explore our Platform",{"href":54,"dataGaName":47,"dataGaLocation":27},"/platform/",{"title":56,"description":57,"link":58},"GitLab Duo (AI)","Build software faster with AI at every stage of development",{"text":59,"config":60},"Meet GitLab Duo",{"href":61,"dataGaName":62,"dataGaLocation":27},"/gitlab-duo/","gitlab duo ai",{"title":64,"description":65,"link":66},"Why GitLab","10 reasons why Enterprises choose GitLab",{"text":67,"config":68},"Learn more",{"href":69,"dataGaName":70,"dataGaLocation":27},"/why-gitlab/","why gitlab",{"title":72,"items":73},"Get started with",[74,79,84],{"text":75,"config":76},"Platform Engineering",{"href":77,"dataGaName":78,"dataGaLocation":27},"/solutions/platform-engineering/","platform engineering",{"text":80,"config":81},"Developer Experience",{"href":82,"dataGaName":83,"dataGaLocation":27},"/developer-experience/","Developer experience",{"text":85,"config":86},"MLOps",{"href":87,"dataGaName":85,"dataGaLocation":27},"/topics/devops/the-role-of-ai-in-devops/",{"text":89,"left":90,"config":91,"link":93,"lists":97,"footer":166},"Product",true,{"dataNavLevelOne":92},"solutions",{"text":94,"config":95},"View all Solutions",{"href":96,"dataGaName":92,"dataGaLocation":27},"/solutions/",[98,123,145],{"title":99,"description":100,"link":101,"items":106},"Automation","CI/CD and automation to accelerate deployment",{"config":102},{"icon":103,"href":104,"dataGaName":105,"dataGaLocation":27},"AutomatedCodeAlt","/solutions/delivery-automation/","automated software delivery",[107,111,115,119],{"text":108,"config":109},"CI/CD",{"href":110,"dataGaLocation":27,"dataGaName":108},"/solutions/continuous-integration/",{"text":112,"config":113},"AI-Assisted Development",{"href":61,"dataGaLocation":27,"dataGaName":114},"AI assisted development",{"text":116,"config":117},"Source Code Management",{"href":118,"dataGaLocation":27,"dataGaName":116},"/solutions/source-code-management/",{"text":120,"config":121},"Automated Software Delivery",{"href":104,"dataGaLocation":27,"dataGaName":122},"Automated software delivery",{"title":124,"description":125,"link":126,"items":131},"Security","Deliver code faster without compromising security",{"config":127},{"href":128,"dataGaName":129,"dataGaLocation":27,"icon":130},"/solutions/security-compliance/","security and compliance","ShieldCheckLight",[132,135,140],{"text":133,"config":134},"Security & Compliance",{"href":128,"dataGaLocation":27,"dataGaName":133},{"text":136,"config":137},"Software Supply Chain Security",{"href":138,"dataGaLocation":27,"dataGaName":139},"/solutions/supply-chain/","Software supply chain security",{"text":141,"config":142},"Compliance & Governance",{"href":143,"dataGaLocation":27,"dataGaName":144},"/solutions/continuous-software-compliance/","Compliance and governance",{"title":146,"link":147,"items":152},"Measurement",{"config":148},{"icon":149,"href":150,"dataGaName":151,"dataGaLocation":27},"DigitalTransformation","/solutions/visibility-measurement/","visibility and measurement",[153,157,161],{"text":154,"config":155},"Visibility & Measurement",{"href":150,"dataGaLocation":27,"dataGaName":156},"Visibility and Measurement",{"text":158,"config":159},"Value Stream Management",{"href":160,"dataGaLocation":27,"dataGaName":158},"/solutions/value-stream-management/",{"text":162,"config":163},"Analytics & Insights",{"href":164,"dataGaLocation":27,"dataGaName":165},"/solutions/analytics-and-insights/","Analytics and insights",{"title":167,"items":168},"GitLab for",[169,174,179],{"text":170,"config":171},"Enterprise",{"href":172,"dataGaLocation":27,"dataGaName":173},"/enterprise/","enterprise",{"text":175,"config":176},"Small Business",{"href":177,"dataGaLocation":27,"dataGaName":178},"/small-business/","small business",{"text":180,"config":181},"Public Sector",{"href":182,"dataGaLocation":27,"dataGaName":183},"/solutions/public-sector/","public sector",{"text":185,"config":186},"Pricing",{"href":187,"dataGaName":188,"dataGaLocation":27,"dataNavLevelOne":188},"/pricing/","pricing",{"text":190,"config":191,"link":193,"lists":197,"feature":282},"Resources",{"dataNavLevelOne":192},"resources",{"text":194,"config":195},"View all resources",{"href":196,"dataGaName":192,"dataGaLocation":27},"/resources/",[198,231,254],{"title":199,"items":200},"Getting started",[201,206,211,216,221,226],{"text":202,"config":203},"Install",{"href":204,"dataGaName":205,"dataGaLocation":27},"/install/","install",{"text":207,"config":208},"Quick start guides",{"href":209,"dataGaName":210,"dataGaLocation":27},"/get-started/","quick setup checklists",{"text":212,"config":213},"Learn",{"href":214,"dataGaLocation":27,"dataGaName":215},"https://university.gitlab.com/","learn",{"text":217,"config":218},"Product documentation",{"href":219,"dataGaName":220,"dataGaLocation":27},"https://docs.gitlab.com/","product documentation",{"text":222,"config":223},"Best practice videos",{"href":224,"dataGaName":225,"dataGaLocation":27},"/getting-started-videos/","best practice videos",{"text":227,"config":228},"Integrations",{"href":229,"dataGaName":230,"dataGaLocation":27},"/integrations/","integrations",{"title":232,"items":233},"Discover",[234,239,244,249],{"text":235,"config":236},"Customer success stories",{"href":237,"dataGaName":238,"dataGaLocation":27},"/customers/","customer success stories",{"text":240,"config":241},"Blog",{"href":242,"dataGaName":243,"dataGaLocation":27},"/blog/","blog",{"text":245,"config":246},"Remote",{"href":247,"dataGaName":248,"dataGaLocation":27},"https://handbook.gitlab.com/handbook/company/culture/all-remote/","remote",{"text":250,"config":251},"TeamOps",{"href":252,"dataGaName":253,"dataGaLocation":27},"/teamops/","teamops",{"title":255,"items":256},"Connect",[257,262,267,272,277],{"text":258,"config":259},"GitLab Services",{"href":260,"dataGaName":261,"dataGaLocation":27},"/services/","services",{"text":263,"config":264},"Community",{"href":265,"dataGaName":266,"dataGaLocation":27},"/community/","community",{"text":268,"config":269},"Forum",{"href":270,"dataGaName":271,"dataGaLocation":27},"https://forum.gitlab.com/","forum",{"text":273,"config":274},"Events",{"href":275,"dataGaName":276,"dataGaLocation":27},"/events/","events",{"text":278,"config":279},"Partners",{"href":280,"dataGaName":281,"dataGaLocation":27},"/partners/","partners",{"backgroundColor":283,"textColor":284,"text":285,"image":286,"link":290},"#2f2a6b","#fff","Insights for the future of software development",{"altText":287,"config":288},"the source promo card",{"src":289},"/images/navigation/the-source-promo-card.svg",{"text":291,"config":292},"Read the latest",{"href":293,"dataGaName":294,"dataGaLocation":27},"/the-source/","the source",{"text":296,"config":297,"lists":299},"Company",{"dataNavLevelOne":298},"company",[300],{"items":301},[302,307,313,315,320,325,330,335,340,345,350],{"text":303,"config":304},"About",{"href":305,"dataGaName":306,"dataGaLocation":27},"/company/","about",{"text":308,"config":309,"footerGa":312},"Jobs",{"href":310,"dataGaName":311,"dataGaLocation":27},"/jobs/","jobs",{"dataGaName":311},{"text":273,"config":314},{"href":275,"dataGaName":276,"dataGaLocation":27},{"text":316,"config":317},"Leadership",{"href":318,"dataGaName":319,"dataGaLocation":27},"/company/team/e-group/","leadership",{"text":321,"config":322},"Team",{"href":323,"dataGaName":324,"dataGaLocation":27},"/company/team/","team",{"text":326,"config":327},"Handbook",{"href":328,"dataGaName":329,"dataGaLocation":27},"https://handbook.gitlab.com/","handbook",{"text":331,"config":332},"Investor relations",{"href":333,"dataGaName":334,"dataGaLocation":27},"https://ir.gitlab.com/","investor relations",{"text":336,"config":337},"Trust Center",{"href":338,"dataGaName":339,"dataGaLocation":27},"/security/","trust center",{"text":341,"config":342},"AI Transparency Center",{"href":343,"dataGaName":344,"dataGaLocation":27},"/ai-transparency-center/","ai transparency center",{"text":346,"config":347},"Newsletter",{"href":348,"dataGaName":349,"dataGaLocation":27},"/company/contact/","newsletter",{"text":351,"config":352},"Press",{"href":353,"dataGaName":354,"dataGaLocation":27},"/press/","press",{"text":356,"config":357,"lists":358},"Contact us",{"dataNavLevelOne":298},[359],{"items":360},[361,364,369],{"text":34,"config":362},{"href":36,"dataGaName":363,"dataGaLocation":27},"talk to sales",{"text":365,"config":366},"Get help",{"href":367,"dataGaName":368,"dataGaLocation":27},"/support/","get help",{"text":370,"config":371},"Customer portal",{"href":372,"dataGaName":373,"dataGaLocation":27},"https://customers.gitlab.com/customers/sign_in/","customer portal",{"close":375,"login":376,"suggestions":383},"Close",{"text":377,"link":378},"To search repositories and projects, login to",{"text":379,"config":380},"gitlab.com",{"href":41,"dataGaName":381,"dataGaLocation":382},"search login","search",{"text":384,"default":385},"Suggestions",[386,388,392,394,398,402],{"text":56,"config":387},{"href":61,"dataGaName":56,"dataGaLocation":382},{"text":389,"config":390},"Code Suggestions (AI)",{"href":391,"dataGaName":389,"dataGaLocation":382},"/solutions/code-suggestions/",{"text":108,"config":393},{"href":110,"dataGaName":108,"dataGaLocation":382},{"text":395,"config":396},"GitLab on AWS",{"href":397,"dataGaName":395,"dataGaLocation":382},"/partners/technology-partners/aws/",{"text":399,"config":400},"GitLab on Google Cloud",{"href":401,"dataGaName":399,"dataGaLocation":382},"/partners/technology-partners/google-cloud-platform/",{"text":403,"config":404},"Why GitLab?",{"href":69,"dataGaName":403,"dataGaLocation":382},{"freeTrial":406,"mobileIcon":411,"desktopIcon":416,"secondaryButton":419},{"text":407,"config":408},"Start free trial",{"href":409,"dataGaName":32,"dataGaLocation":410},"https://gitlab.com/-/trials/new/","nav",{"altText":412,"config":413},"Gitlab Icon",{"src":414,"dataGaName":415,"dataGaLocation":410},"/images/brand/gitlab-logo-tanuki.svg","gitlab icon",{"altText":412,"config":417},{"src":418,"dataGaName":415,"dataGaLocation":410},"/images/brand/gitlab-logo-type.svg",{"text":420,"config":421},"Get Started",{"href":422,"dataGaName":423,"dataGaLocation":410},"https://gitlab.com/-/trial_registrations/new?glm_source=about.gitlab.com/compare/gitlab-vs-github/","get started",{"freeTrial":425,"mobileIcon":429,"desktopIcon":431},{"text":426,"config":427},"Learn more about GitLab Duo",{"href":61,"dataGaName":428,"dataGaLocation":410},"gitlab duo",{"altText":412,"config":430},{"src":414,"dataGaName":415,"dataGaLocation":410},{"altText":412,"config":432},{"src":418,"dataGaName":415,"dataGaLocation":410},"content:shared:en-us:main-navigation.yml","Main Navigation","shared/en-us/main-navigation.yml","shared/en-us/main-navigation",{"_path":438,"_dir":21,"_draft":6,"_partial":6,"_locale":7,"title":439,"button":440,"config":444,"_id":446,"_type":13,"_source":15,"_file":447,"_stem":448,"_extension":18},"/shared/en-us/banner","GitLab Duo Agent Platform is now in public beta!",{"text":67,"config":441},{"href":442,"dataGaName":443,"dataGaLocation":27},"/gitlab-duo/agent-platform/","duo banner",{"layout":445},"release","content:shared:en-us:banner.yml","shared/en-us/banner.yml","shared/en-us/banner",{"_path":450,"_dir":21,"_draft":6,"_partial":6,"_locale":7,"data":451,"_id":656,"_type":13,"title":657,"_source":15,"_file":658,"_stem":659,"_extension":18},"/shared/en-us/main-footer",{"text":452,"source":453,"edit":459,"contribute":464,"config":469,"items":474,"minimal":648},"Git is a trademark of Software Freedom Conservancy and our use of 'GitLab' is under license",{"text":454,"config":455},"View page source",{"href":456,"dataGaName":457,"dataGaLocation":458},"https://gitlab.com/gitlab-com/marketing/digital-experience/about-gitlab-com/","page source","footer",{"text":460,"config":461},"Edit this page",{"href":462,"dataGaName":463,"dataGaLocation":458},"https://gitlab.com/gitlab-com/marketing/digital-experience/about-gitlab-com/-/blob/main/content/","web ide",{"text":465,"config":466},"Please contribute",{"href":467,"dataGaName":468,"dataGaLocation":458},"https://gitlab.com/gitlab-com/marketing/digital-experience/about-gitlab-com/-/blob/main/CONTRIBUTING.md/","please contribute",{"twitter":470,"facebook":471,"youtube":472,"linkedin":473},"https://twitter.com/gitlab","https://www.facebook.com/gitlab","https://www.youtube.com/channel/UCnMGQ8QHMAnVIsI3xJrihhg","https://www.linkedin.com/company/gitlab-com",[475,498,555,584,618],{"title":45,"links":476,"subMenu":481},[477],{"text":478,"config":479},"DevSecOps platform",{"href":54,"dataGaName":480,"dataGaLocation":458},"devsecops platform",[482],{"title":185,"links":483},[484,488,493],{"text":485,"config":486},"View plans",{"href":187,"dataGaName":487,"dataGaLocation":458},"view plans",{"text":489,"config":490},"Why Premium?",{"href":491,"dataGaName":492,"dataGaLocation":458},"/pricing/premium/","why premium",{"text":494,"config":495},"Why Ultimate?",{"href":496,"dataGaName":497,"dataGaLocation":458},"/pricing/ultimate/","why ultimate",{"title":499,"links":500},"Solutions",[501,506,509,511,516,521,525,528,532,537,539,542,545,550],{"text":502,"config":503},"Digital transformation",{"href":504,"dataGaName":505,"dataGaLocation":458},"/topics/digital-transformation/","digital transformation",{"text":133,"config":507},{"href":128,"dataGaName":508,"dataGaLocation":458},"security & compliance",{"text":122,"config":510},{"href":104,"dataGaName":105,"dataGaLocation":458},{"text":512,"config":513},"Agile development",{"href":514,"dataGaName":515,"dataGaLocation":458},"/solutions/agile-delivery/","agile delivery",{"text":517,"config":518},"Cloud transformation",{"href":519,"dataGaName":520,"dataGaLocation":458},"/topics/cloud-native/","cloud transformation",{"text":522,"config":523},"SCM",{"href":118,"dataGaName":524,"dataGaLocation":458},"source code management",{"text":108,"config":526},{"href":110,"dataGaName":527,"dataGaLocation":458},"continuous integration & delivery",{"text":529,"config":530},"Value stream management",{"href":160,"dataGaName":531,"dataGaLocation":458},"value stream management",{"text":533,"config":534},"GitOps",{"href":535,"dataGaName":536,"dataGaLocation":458},"/solutions/gitops/","gitops",{"text":170,"config":538},{"href":172,"dataGaName":173,"dataGaLocation":458},{"text":540,"config":541},"Small business",{"href":177,"dataGaName":178,"dataGaLocation":458},{"text":543,"config":544},"Public sector",{"href":182,"dataGaName":183,"dataGaLocation":458},{"text":546,"config":547},"Education",{"href":548,"dataGaName":549,"dataGaLocation":458},"/solutions/education/","education",{"text":551,"config":552},"Financial services",{"href":553,"dataGaName":554,"dataGaLocation":458},"/solutions/finance/","financial services",{"title":190,"links":556},[557,559,561,563,566,568,570,572,574,576,578,580,582],{"text":202,"config":558},{"href":204,"dataGaName":205,"dataGaLocation":458},{"text":207,"config":560},{"href":209,"dataGaName":210,"dataGaLocation":458},{"text":212,"config":562},{"href":214,"dataGaName":215,"dataGaLocation":458},{"text":217,"config":564},{"href":219,"dataGaName":565,"dataGaLocation":458},"docs",{"text":240,"config":567},{"href":242,"dataGaName":243,"dataGaLocation":458},{"text":235,"config":569},{"href":237,"dataGaName":238,"dataGaLocation":458},{"text":245,"config":571},{"href":247,"dataGaName":248,"dataGaLocation":458},{"text":258,"config":573},{"href":260,"dataGaName":261,"dataGaLocation":458},{"text":250,"config":575},{"href":252,"dataGaName":253,"dataGaLocation":458},{"text":263,"config":577},{"href":265,"dataGaName":266,"dataGaLocation":458},{"text":268,"config":579},{"href":270,"dataGaName":271,"dataGaLocation":458},{"text":273,"config":581},{"href":275,"dataGaName":276,"dataGaLocation":458},{"text":278,"config":583},{"href":280,"dataGaName":281,"dataGaLocation":458},{"title":296,"links":585},[586,588,590,592,594,596,598,602,607,609,611,613],{"text":303,"config":587},{"href":305,"dataGaName":298,"dataGaLocation":458},{"text":308,"config":589},{"href":310,"dataGaName":311,"dataGaLocation":458},{"text":316,"config":591},{"href":318,"dataGaName":319,"dataGaLocation":458},{"text":321,"config":593},{"href":323,"dataGaName":324,"dataGaLocation":458},{"text":326,"config":595},{"href":328,"dataGaName":329,"dataGaLocation":458},{"text":331,"config":597},{"href":333,"dataGaName":334,"dataGaLocation":458},{"text":599,"config":600},"Sustainability",{"href":601,"dataGaName":599,"dataGaLocation":458},"/sustainability/",{"text":603,"config":604},"Diversity, inclusion and belonging (DIB)",{"href":605,"dataGaName":606,"dataGaLocation":458},"/diversity-inclusion-belonging/","Diversity, inclusion and belonging",{"text":336,"config":608},{"href":338,"dataGaName":339,"dataGaLocation":458},{"text":346,"config":610},{"href":348,"dataGaName":349,"dataGaLocation":458},{"text":351,"config":612},{"href":353,"dataGaName":354,"dataGaLocation":458},{"text":614,"config":615},"Modern Slavery Transparency Statement",{"href":616,"dataGaName":617,"dataGaLocation":458},"https://handbook.gitlab.com/handbook/legal/modern-slavery-act-transparency-statement/","modern slavery transparency statement",{"title":619,"links":620},"Contact Us",[621,624,626,628,633,638,643],{"text":622,"config":623},"Contact an expert",{"href":36,"dataGaName":37,"dataGaLocation":458},{"text":365,"config":625},{"href":367,"dataGaName":368,"dataGaLocation":458},{"text":370,"config":627},{"href":372,"dataGaName":373,"dataGaLocation":458},{"text":629,"config":630},"Status",{"href":631,"dataGaName":632,"dataGaLocation":458},"https://status.gitlab.com/","status",{"text":634,"config":635},"Terms of use",{"href":636,"dataGaName":637,"dataGaLocation":458},"/terms/","terms of use",{"text":639,"config":640},"Privacy statement",{"href":641,"dataGaName":642,"dataGaLocation":458},"/privacy/","privacy statement",{"text":644,"config":645},"Cookie preferences",{"dataGaName":646,"dataGaLocation":458,"id":647,"isOneTrustButton":90},"cookie preferences","ot-sdk-btn",{"items":649},[650,652,654],{"text":634,"config":651},{"href":636,"dataGaName":637,"dataGaLocation":458},{"text":639,"config":653},{"href":641,"dataGaName":642,"dataGaLocation":458},{"text":644,"config":655},{"dataGaName":646,"dataGaLocation":458,"id":647,"isOneTrustButton":90},"content:shared:en-us:main-footer.yml","Main Footer","shared/en-us/main-footer.yml","shared/en-us/main-footer",{"allPosts":661,"featuredPost":1246,"totalPagesCount":1265,"initialPosts":1266},[662,690,713,734,755,776,798,818,838,857,877,899,921,942,963,983,1003,1025,1043,1062,1082,1103,1124,1144,1164,1185,1206,1226],{"_path":663,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":664,"content":672,"config":683,"_id":686,"_type":13,"title":687,"_source":15,"_file":688,"_stem":689,"_extension":18},"/en-us/blog/adsoul-devops-transition-to-gitlab-ci",{"title":665,"description":666,"ogTitle":665,"ogDescription":666,"noIndex":6,"ogImage":667,"ogUrl":668,"ogSiteName":669,"ogType":670,"canonicalUrls":668,"schema":671},"How adSoul transitioned to GitLab CI from Jenkins","adSoul, a marketing automation company, outlines a successful three-phase migration plan for moving to GitLab CI from Jenkins.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749678442/Blog/Hero%20Images/londoncommit.png","https://about.gitlab.com/blog/adsoul-devops-transition-to-gitlab-ci","https://about.gitlab.com","article","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"How adSoul transitioned to GitLab CI from Jenkins\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"Brein Matturro\"}],\n        \"datePublished\": \"2019-11-05\",\n      }",{"title":665,"description":666,"authors":673,"heroImage":667,"date":675,"body":676,"category":677,"tags":678},[674],"Brein Matturro","2019-11-05","\nadSoul is a Germany-based data-driven online marketing company that aims to improve search engine advertising and scalability for businesses. The core of adSoul relies heavily on API interfaces and entity recognition to post keywords on Google and Bing with marketing automation. \n\nAt GitLab Commit London, [Philipp Westphalen](https://www.linkedin.com/in/philipp-westphalen-a83318188/), fullstack developer at adSoul and GitLab Hero, shares how the company transitioned from Jenkins to GiLab CI. adSoul is a startup company with five developers, and as Philipp says “We literally have no time for everything we need to do.” They were looking for a tool that requires less time-consuming maintanence, and with Jenkins the team found it hard to read their existing files. “Our Jenkins was not so stable at all and it was tough to change because it was managed by our provider,” Philipp says. Cost and visibility were also huge motivators in moving away from [Jenkins to Gitlab CI](/blog/docker-my-precious/).\n\n## GitLab migration in three phases\n\nPhase 1: Move the repository.\nThe [adSoul team](https://www.adsoul.com) used the GitHub Import by GitLab, but had setbacks with migrating their issues, so they created a GitHub open source issue migrator as a resolution. Following that, they modified scripts with the new origin by exchanging the GitHub API call with a GitLab API. “This was really easy and we had a stable build with our new repository, so we could move our product management to GitLab and not need GitHub anymore,” Philipp says.\n\nPhase 2: Migrate the CI/CD pipeline.\nThe team started to create a GitLab CI YAML and tried to do a simple ‘lift and shift,’ however their processes were more complicated than anticipated. Though this phase was time consuming, it became clear the team could move to phase three without hiccups. “Quick pro tip,” says Philipp. “If you’re running your own GitLab runners, increase the log limit if you have to debug your building step.” \n\nPhase 3: Improve the CI/CD pipeline.\nThe team thought about ways of building their software, so they split projects into steps. “Our idea was that one job does one thing perfectly. Each job is simple and everyone can modify it easily” Philipp says. They improved their build time by moving to Gradle, created parallel job processing, and by using standard Docker images for ease of management. \n\n## Takeaways from a successful migration\n\n1. Plan your migration. Get every member of the team involved and aware of the upcoming changes, including how tools are working together and what the expectations are moving forward. “Take your time for the migration,” Philipp says. “It’s not two days and then we are finished.” \n\n2. Go step by step. adSoul used a three phase plan which allowed the team to deploy a new version and still continue to work on existing projects. “We could improve our application without having to wait for a better infrastructure,” Philipp says.\n\n3. Rethink your [DevOps strategy](/blog/better-devops-with-gitlab-ci-cd/). In the time leading up to the migration, examine things like security automation and other important pieces in a DevOps overall strategy.\n\n4. Start with a small project. Work closely with colleagues to create small GitLab CI projects to familiarize everyone before creating larger, overwhelming projects.\n\nPro tip: Keep your pipeline user friendly. Create a good user experience for the team with clear job names, style your config for a better overview, and write comments for variables and hard to understand code. \n\n## Why GitLab works for a small team\n\n“The most important thing is that GitLab is a powerful CI/CD solution with high customization,” Philipp says. There is one home for all projects, without dependencies on one another. With Jenkins, even small exploratory changes can impact the larger job. “With GitLab, you don’t have dependency between branches. So, if you’re trying something new for your CI, you can do it simply in your branch and the master branch will not be affected by the changes,” Philipp says.\n\nThe CI is low maintenance, which is a useful timesaver for a smaller team. “The CI provides us with really low maintenance time. So, usually we don’t have to care about our CI for a month or more,” Philipp says.\n\nTo learn more about adSoul’s migration to GitLab, watch Philipp’s talk from GitLab Commit London.\n\n\u003C!-- blank line -->\n\u003Cfigure class=\"video_container\">\n  \u003Ciframe src=\"https://www.youtube.com/embed/C5xfw0ydh2k\" frameborder=\"0\" allowfullscreen=\"true\"> \u003C/iframe>\n\u003C/figure>\n\u003C!-- blank line -->\n","engineering",[108,679,680,681,9,682],"DevOps","open source","performance","cloud native",{"slug":684,"featured":6,"template":685},"adsoul-devops-transition-to-gitlab-ci","BlogPost","content:en-us:blog:adsoul-devops-transition-to-gitlab-ci.yml","Adsoul Devops Transition To Gitlab Ci","en-us/blog/adsoul-devops-transition-to-gitlab-ci.yml","en-us/blog/adsoul-devops-transition-to-gitlab-ci",{"_path":691,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":692,"content":698,"config":707,"_id":709,"_type":13,"title":710,"_source":15,"_file":711,"_stem":712,"_extension":18},"/en-us/blog/all-remote-fundraising",{"title":693,"description":694,"ogTitle":693,"ogDescription":694,"noIndex":6,"ogImage":695,"ogUrl":696,"ogSiteName":669,"ogType":670,"canonicalUrls":696,"schema":697},"How to raise funds as an all-remote startup","GitLab CEO Sid Sijbrandij and podcast host Maren Kate unpack why venture firms struggle to fund all-remote startups.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749673152/Blog/Hero%20Images/remotefundraisinghurdle.jpg","https://about.gitlab.com/blog/all-remote-fundraising","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"How to raise funds as an all-remote startup\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"Valerie Silverthorne\"}],\n        \"datePublished\": \"2019-08-16\",\n      }",{"title":693,"description":694,"authors":699,"heroImage":695,"date":701,"body":702,"category":703,"tags":704},[700],"Valerie Silverthorne","2019-08-16","\nIt’s possible to be an all-remote startup and get venture capital – GitLab is proof of that – but that doesn’t mean it’s easy. GitLab CEO [Sid Sijbrandij](/company/team/#sytses) spoke with [Maren Kate](https://www.linkedin.com/in/marenkate), host of the From 5 to 50 podcast, about why venture capitalists don’t love all-remote companies and how to work around that challenge. The [Remote AF](https://podcasts.apple.com/us/podcast/gitlab-raised-$100m-got-valued-at-over-billion-by-starting/id1467214647?i=1000444691471) podcast is aimed at startups looking to scale their companies from 5 to 50 employees and beyond.\n\nMaren starts by asking Sid how the concept of all-remote was received by the venture capital community. Sid’s response: “They don’t like remote. We missed out on investors because we are remote. We have skepticism from investors because we are remote.”\n\n## Stages of fundraising\n\nFundraising changes as a company grows, and it gets easier with time, Sid explains. “In the beginning they assess your team, then they assess your product, and then they assess your financials.” That’s why it can be hard for a newly-minted, all-remote startup to get fundraising traction in the early stages, he says. “When it comes to the team, they’re super skeptical they will be able to create something with all-remote. Then when it’s about the product they say, ‘Yes, maybe, but what about scaling?’ And then when it’s about the financials you can let the numbers speak for themselves so it’s less of a concern.”\n\nAnd in the early days of GitLab, even Sid was skeptical enough about all-remote to open an office. That office made our [series A financing](/blog/gitlab-announces-4m-series-a-funding-from-khosla-ventures/) easier, he says. But Sid soon realized that people weren’t coming into the office (San Francisco Bay Area traffic being what it is) so committed to an [all-remote philosophy](/company/culture/all-remote/). That decision made [series B fundraising efforts](/blog/gitlab-master-plan/) difficult. Some investors just said no to GitLab, but a few at least asked for an explanation. Even after an explanation, many remained dubious and in the end it took an enthusiastic VC who’d actually stayed up late reading through the handbook and vouched for GitLab to seal the deal.\n\n>Some of the best ideas are the least plausible.\n\nAll-remote companies are getting a toehold today, Sid offers, pointing to [InVision](/blog/pyb-all-remote-mark-frein/), WordPress, and Zapier. But there are still some factors that can inhibit fundraising options. “If we were to be acquired there’s probably a 50% discount, because for the acquiring company it’s so hard to bring people over to their headquarters,” Sid says. “Since an acquisition is the most likely outcome (for most startups), if you raise venture capital that depresses the evaluation you will get.”\n\n## Has co-location hit the wall?\n\nOn the upside, though, Sid thinks the limits of co-location are being made very clear. “Investors in San Francisco are all battling it out. They’re saying ‘Our portfolio companies are getting outbid by Google, by eBay, by Airbnb for engineering talent.’ Retention is an enormous problem at these companies. So they don’t like remote yet, but they’re starting to sour on the co-located model and all the disadvantages.”\n\nAnd while the all-remote path might be tough, Sid continues to stress the benefits to startups. “Remote offers you much easier hiring and scaling. Remote forces you to do the things you should be doing anyway, but you do them sooner.”\n\nAt the end of the day, Maren wonders whether some unconscious bias is at play. “If you see someone in an office, it makes them more successful,” she says, “but it’s not really that, it’s just human instinct.”\n\nSid agrees, and then adds perhaps the strongest argument in favor of all-remote – co-location can have a very dampening effect on innovation. “There's a lot of detrimental things that happen because some of the best ideas are the least plausible, like run an illegal taxi service, have people rent out their own home to strangers, or start competing with GitHub. And if you co-locate people, they're going to have to tell everyone what they do. And when they see people frowning, they're going to switch to something more plausible. And that's what you want to prevent.”\n\nListen to the whole conversation [here](https://podcasts.apple.com/us/podcast/gitlab-raised-$100m-got-valued-at-over-billion-by-starting/id1467214647?i=1000444691471).\n\nCover image by [Pau Casals](https://unsplash.com/@paucasals) on [Unsplash](https://unsplash.com)\n{: .note}\n","insights",[705,706,9],"inside GitLab","remote work",{"slug":708,"featured":6,"template":685},"all-remote-fundraising","content:en-us:blog:all-remote-fundraising.yml","All Remote Fundraising","en-us/blog/all-remote-fundraising.yml","en-us/blog/all-remote-fundraising",{"_path":714,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":715,"content":721,"config":728,"_id":730,"_type":13,"title":731,"_source":15,"_file":732,"_stem":733,"_extension":18},"/en-us/blog/announcing-100m-series-d-funding",{"title":716,"description":717,"ogTitle":716,"ogDescription":717,"noIndex":6,"ogImage":718,"ogUrl":719,"ogSiteName":669,"ogType":670,"canonicalUrls":719,"schema":720},"Announcing $100 million in Series D round funding led by ICONIQ Capital","Today we announced $100M in new funding to beat nine best-in-class products with a single application.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749663655/Blog/Hero%20Images/gitlab-live-sept-2018.png","https://about.gitlab.com/blog/announcing-100m-series-d-funding","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"Announcing $100 million in Series D round funding led by ICONIQ Capital\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"GitLab\"}],\n        \"datePublished\": \"2018-09-19\",\n      }",{"title":716,"description":717,"authors":722,"heroImage":718,"date":724,"body":725,"category":298,"tags":726},[723],"GitLab","2018-09-19","Today we are thrilled to announce our $100 million Series D funding led by ICONIQ Capital, bringing our valuation to over $1 billion and validating our position as the world’s first single application for the entire DevOps lifecycle. We’re elated that ICONIQ shares our vision of beating out nine other products with a single, best-in-class application that integrates each stage of the DevOps lifecycle. We plan to use the funding to become best-in-class in every DevOps software category from planning to monitoring. More than 100,000 organizations use GitLab to deploy to multiple clouds, implement [cloud native](/topics/cloud-native/) architectures, and practice Concurrent DevOps. This results in a 200 percent faster DevOps lifecycle with unmatched visibility, higher levels of efficiency, and comprehensive governance.\n\n> “GitLab is emerging as a leader across the entire software development ecosystem by releasing software at a pace that is unmatched by any competitor. They’re taking the broad software development market head-on by developing an application that allows organizations to churn out software at an accelerated rate with cost and time savings.” - Matthew Jacobson, General Partner at ICONIQ Capital\n\n\u003C!-- blank line -->\n\u003Cfigure class=\"video_container\">\n  \u003Ciframe src=\"https://www.youtube.com/embed/ZgFqyXCsqPY\" frameborder=\"0\" allowfullscreen=\"true\"> \u003C/iframe>\n\u003C/figure>\n\u003C!-- blank line -->\n\n\u003Cfigure class=\"video_container\">\n\u003Ciframe src=\"https://docs.google.com/presentation/d/e/2PACX-1vTO_mVE0psqDSIOwmrv30ebL0IMdAIhYFHqBcoqI6b8_Cl1yl8f6FaAIm-d7qwsOWhhiUIqPxo6fjhH/embed?start=false&loop=false&delayms=3000\" frameborder=\"0\" width=\"1280\" height=\"749\" allowfullscreen=\"true\" mozallowfullscreen=\"true\" webkitallowfullscreen=\"true\">\u003C/iframe>\n\u003C/figure>\n\nEnterprises are facing a tool chain crisis, investing time and resources into piecing together disparate tools from different stages of the software development and operations lifecycle. A typical enterprise needs to integrate tools like VersionOne, Jira, GitHub, Jenkins, Artifactory, Electric Cloud, Puppet, New Relic, and BlackDuck. This causes poor visibility because data lives in so many different tools, slow cycle time because teams need to wait on each other, and bolted on security as an afterthought.\n\nOur open core application, which has more than 2,000 contributors, is breaking down this barrier by building features for each stage of the DevOps lifecycle into a single application. This enables Concurrent DevOps, the ability for teams to manage, plan, create, verify, package, release, configure, monitor, and secure software together so they can have visibility into what matters, start without waiting, and ship with confidence.\n\n> “Our goal is to strive for less people managing processes and more automation within our workflow. GitLab does just that by eliminating the complicated web that tied all of our development tools together, so we now have a single, automated application that makes our team more efficient.” — Adam Dehnel, product architect, BI Worldwide\n\nWe started as source code management based on Git. But when announcing our series C last year, we [voiced the ambition](/blog/gitlab-raises-20-million-to-complete-devops/) to grow beyond that to cover every product category of the DevOps lifecycle. Some features have already become best in class, like the continuous integration system that scored highest in the current offering category in The Forrester Wave™: Continuous Integration Tools, Q3 2017 report, and was recognized as a \"Strong Performer\" in The Forrester New Wave™ for Value Stream Management Tools, Q3 2018. With this latest funding round and the growing number of people contributing code to GitLab, we are well positioned for all of our features to become best-in-class.\n\n> “Two of the defining characteristics of modern application development are an open, collaborative process and toolchains that transcend individual product categories from planning out to operations. By marrying an open source development process to a comprehensive set of functional capabilities, GitLab aims to leverage those precise qualities in a single application that meets enterprises’ DevOps needs.” - Stephen O’Grady, Principal Analyst with RedMonk\n\n> “Since raising a Series C round last year, we’ve delivered on our commitment to bring a single application for the entire DevOps lifecycle to market, and as a result have been able to reach over $1 billion in valuation. With this latest funding round, we will continue to improve upon our suite by building out our management, planning, packaging, monitoring and security features for a more robust DevOps application.” – Sid Sijbrandij, CEO of GitLab\n\nGitLab is purpose-built for organizations that are undergoing a digital transformation. Our focus is on supporting organizations that are aiming for faster DevOps lifecycles, cloud native architectures, and multi-cloud deployments. Some of our recent product milestones include the release of [Auto DevOps](/press/releases/2018-06-22-auto-devops-gitlab-11.html) to accelerate the DevOps lifecycle by 200 percent, a [Kubernetes integration](/releases/2018/03/22/gitlab-10-6-released/) so clusters can be spun up from within GitLab, and [enhancements to the Web IDE](/releases/2018/08/22/gitlab-11-2-released/) to make code changes easier for everyone.\n\nThe latest round brings our valuation to over $1 billion and validates our position as the world’s first single application for the entire DevOps lifecycle. As other DevOps tools become locked in to a single cloud, our customers are embracing a multi-cloud future, so their tooling must be independent of any single cloud provider and work across on-premises, private and public clouds. Customers want their tooling to be consistent, whether they deploy it themselves or use it as a service.\n\n> “Deployments are no longer limited to a traditional software release cycle. Using GitLab-CI for our Continuous Deployments over-the-air, we have empowered the customer to be their own technician while increasing the rate at which software can be delivered.” - Chris Hill, head of systems engineering, next generation infotainment at Jaguar Land Rover.\n\nWe began in 2014 with a mission to change all creative work from read-only to read-write, so that everyone can contribute. Since then, our all-remote company has grown from fewer than 10 to more than 350 team members in over 40 countries across the globe. And we’re not slowing down – we’re still [hiring for 77 positions](/jobs/)!\n\nWe were recently [recognized by Inc. Magazine as No. 44 out of the 5,000 fastest-growing private companies in the United States](/blog/gitlab-ranked-44-on-inc-5000-list/). We attribute our success to our open core model and our value of transparency. We have an emphasis on co-creation and a commitment to open source – [over 2,000 users and customers have contributed to GitLab’s code base](http://contributors.gitlab.com/). This philosophy helps build stronger customer relationships, which in turn result in a direct influence on feature updates to the product.\n\n## Get involved\n\nWe owe GitLab’s existence to your enthusiasm, drive, and hard work. Without our contributors’ belief in open source software, we would not be where we are today. We need your help to make our collective vision a reality. Everyone can contribute!\n\nWe are committed to standing by our [promise to be good stewards of open source](/blog/being-a-good-open-source-steward/),\nand keeping communication and collaboration amongst the community a high priority.\n\nTo get started you can learn more about [Concurrent DevOps](/topics/devops/ ), ready stories of [how others are using GitLab](/customers/), or visit our [contribution guide](/community/contribute/).",[727,705,9],"news",{"slug":729,"featured":6,"template":685},"announcing-100m-series-d-funding","content:en-us:blog:announcing-100m-series-d-funding.yml","Announcing 100m Series D Funding","en-us/blog/announcing-100m-series-d-funding.yml","en-us/blog/announcing-100m-series-d-funding",{"_path":735,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":736,"content":742,"config":749,"_id":751,"_type":13,"title":752,"_source":15,"_file":753,"_stem":754,"_extension":18},"/en-us/blog/expanding-gitlab-for-startups-program",{"title":737,"description":738,"ogTitle":737,"ogDescription":738,"noIndex":6,"ogImage":739,"ogUrl":740,"ogSiteName":669,"ogType":670,"canonicalUrls":740,"schema":741},"Expanding GitLab for Startups program","Learn how startups can access the complete DevSecOps platform.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749664087/Blog/Hero%20Images/tanukicover.jpg","https://about.gitlab.com/blog/expanding-gitlab-for-startups-program","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"Expanding GitLab for Startups program\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"Emilio Salvador\"}],\n        \"datePublished\": \"2023-03-01\",\n      }",{"title":737,"description":738,"authors":743,"heroImage":739,"date":745,"body":746,"category":727,"tags":747},[744],"Emilio Salvador","2023-03-01","\nGitLab is committed to supporting the startup community, and we are excited to announce the expansion of our [GitLab for Startups program](/solutions/startups/). Our goal has always been to help businesses of all sizes streamline their development processes and improve collaboration, and we recognize the unique challenges that startups face. \n\nStarting on March 15, 2023, our GitLab for Startups program will include qualifying startups backed by external funding. Based on funding level and years in the program, qualifying startups are able to utilize our complete DevSecOps platform at a highly discounted price or possibly at no cost for the first year. The program will continue supporting eligible startups with deep discounts in the second year.  \n\nWith GitLab, startups are able to reduce software development cycles, reduce time to market, and improve product quality. By adopting a single application for the entire DevSecOps lifecycle, startups can develop software faster and focus more on growing their customer base, increasing revenue, and differentiating their product in the market.\n\n![Chart listing eligibility requirements](https://about.gitlab.com/images/blogimages/gitlabforstartupscriteria.png)\n\nQualifying startups can submit the application form on our [GitLab for Startups page](/solutions/startups/join/) starting on March 15, 2023. For additional questions regarding this offer, please see our FAQ section on the GitLab for Startups page or feel free to reach out to us at startups@gitlab.com.\n",[748,9,727],"DevSecOps",{"slug":750,"featured":6,"template":685},"expanding-gitlab-for-startups-program","content:en-us:blog:expanding-gitlab-for-startups-program.yml","Expanding Gitlab For Startups Program","en-us/blog/expanding-gitlab-for-startups-program.yml","en-us/blog/expanding-gitlab-for-startups-program",{"_path":756,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":757,"content":763,"config":770,"_id":772,"_type":13,"title":773,"_source":15,"_file":774,"_stem":775,"_extension":18},"/en-us/blog/from-idea-to-production-on-thousands-of-clouds",{"title":758,"description":759,"ogTitle":758,"ogDescription":759,"noIndex":6,"ogImage":760,"ogUrl":761,"ogSiteName":669,"ogType":670,"canonicalUrls":761,"schema":762},"From idea to production on thousands of clouds","Deliver cloud native applications in more places consistently at scale with GitLab and Gravity.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749679266/Blog/Hero%20Images/blue-lights.jpg","https://about.gitlab.com/blog/from-idea-to-production-on-thousands-of-clouds","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"From idea to production on thousands of clouds\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"Ev Kontsevoy\"}],\n        \"datePublished\": \"2019-11-20\",\n      }",{"title":758,"description":759,"authors":764,"heroImage":760,"date":766,"body":767,"category":298,"tags":768},[765],"Ev Kontsevoy","2019-11-20","\nToday, deploying an application with GitLab is easier than ever: just create a Kubernetes cluster on your cloud of choice, connect it to GitLab with the Kubernetes integration, and Auto DevOps creates a full deployment pipeline for you.\n\nBut what if you need your app to run in two clusters in two separate regions? Ten clusters across multiple cloud providers? A hundred clusters and also on a fleet of self-driving trucks?\n\nAt [Gravitational](https://gravitational.com), we believe the future should not belong to a single cloud provider and developers should be able to run their applications anywhere with the same simplicity as having a single Kubernetes cluster.\n\nI am a huge fan of GitLab. I’ve had the great pleasure of getting to know much of the founding team [over the years](https://about.gitlab.com/blog/gitlab-joins-forces-with-gravitational/) and was happy to provide my [own contribution](https://gitlab.com/gitlab-org/gitlab-foss/issues/22864) to the community a while back. Today, I’m happy to share some thoughts on how to build with GitLab and deploy applications into dozens or even hundreds of cloud environments. \n\n## The rise of multicloud\n\nHow do you run applications in different data centers? Do you need to rewrite them from scratch? AWS may still be the dominant cloud provider, but cloud competitors are eating into their lead. It’s not just the big public cloud companies either. [Private cloud data centers](https://www.forbes.com/sites/jasonbloomberg/2019/02/02/have-private-clouds-finally-found-their-place-in-the-enterprise/#2f859685604f) are growing just as rapidly.\n\nMany companies that need to meet tough security and compliance requirements will require applications to run in their bare metal data centers. Running an application on an on-premises or even air-gapped data center adds additional complexity due to the hundreds or even thousands of dependencies in modern applications.\n\nGravitational has built Gravity, an open source [Kubernetes packaging solution ](https://gravitational.com/gravity/)that allows developers to build “cluster images” (similar to VM images) that can contain an entire Kubernetes cluster pre-loaded with multiple applications. You would use GitLab to go from idea to production, and Gravity to expand your production to anywhere in the world. \n\nStatements like, “I have snapshotted our entire production environment and emailed it to you, so you can run it in your private data center,” will not seem completely crazy.\n\nGravity uses standard, upstream CNCF-supported tooling for creating \"images\" of Kubernetes clusters containing the applications and their dependencies. The resulting files are called cluster images which are just .tar files.\n\nA cluster image can be used to recreate full replicas of the original environments for any deployment environment where compliance and consistency matter, i.e. in locked-down AWS/GCP/Azure environments or even in air-gapped server rooms. Each image includes all dependencies to spin up a full cluster, as well as the Gravity daemon that handles the most common operational tasks associated with Kubernetes applications, and it monitors and alerts human operators of problems.\n\n## Deploy with GitLab, scale with Gravity\n\n![Gravity dashboard](https://about.gitlab.com/images/blogimages/gravity-dashboard.png)\n\nDevelopers can leverage a GitLab repository as a single source of truth for rolling out a Kubernetes app and leverage [GitLab CI/CD](https://docs.gitlab.com/ee/ci/) for continuous delivery.\n\nAny project of meaningful scale begins by defining an [epic](https://docs.gitlab.com/ee/user/group/epics/) with goals, milestones, and tasks. An [issue](https://docs.gitlab.com/ee/user/project/issues/#issues) is the main object for collaborating on ideas and planning work. GitLab’s [package and container registry](https://about.gitlab.com/stages-devops-lifecycle/package/) helps you manage and package dependencies. \n\n[The GitLab Kubernetes integration](https://docs.gitlab.com/ee/user/project/clusters/) allows customers to create Kubernetes clusters, utilize review apps, run pipelines, use web terminals, deploy apps, view pod logs, detect and monitor Kubernetes, and much more. For deploying a Kubernetes cluster in a single destination, GitLab provides everything you need from start to finish. \n\nHowever, if your customers need to run your application in their private data centers, they can use Gravity, which essentially copy/pastes the entire Kubernetes cluster environment you’ve built in GitLab. \n\n[Download](https://gravitational.com/gravity/download/) and set up the Gravity open source edition following our [quickstart guide](https://gravitational.com/gravity/docs/quickstart/). From Gravity, you can build a cluster image of your Kubernetes application. Gravity’s [documentation](https://gravitational.com/gravity/docs/overview/) will walk you through the steps required to build an image manifest that describes the image build, the installation process, and the system requirements for the cluster. \n\nYou can build empty Kubernetes cluster images to quickly create a large number of identical, production-ready Kubernetes clusters within an organization, or you can build a cluster image that also includes Kubernetes applications to distribute your application to third parties. \n\n## Next steps\n\nIf you want to learn more about working with Kubernetes, start with [Kubernetes 101](https://www.youtube.com/watch?v=rq4GZ_GybN8). You’ll learn how GitLab and Kubernetes interact at various touchpoints. And, if you’re looking for a way to port your applications to new environments, check out [Gravity](https://gravitational.com/gravity). \n\n## About the guest author\n\nEv is a co-founder and the CEO of Gravitational. Before Gravitational, he launched the on-demand OpenCompute servers at Rackspace. Prior to Rackspace, he co-founded Mailgun, the first email service built for developers. Ev has been a fighter against unnecessary complexity in software for 20 years. He abhors cars but loves trains and open source software that doesn't require an army of consultants to operate.\n\n## About Gravitational\n\n[Gravitational](https://gravitational.com) helps companies deliver cloud applications across cloud providers, on-premises environments, and even air-gapped server rooms. Products include Teleport for multi-cloud privileged access management that doesn't get in the way of developer productivity, and Gravity, a Kubernetes packaging solution that takes the drama out of on-prem deployments. Gravitational was founded in 2015 and recently [announced their Series A](https://gravitational.com/blog/gravitational-series-a-funding/). \n\nCover image by [Sharon McCutcheon](https://unsplash.com/@sharonmccutcheon) on [Unsplash](https://unsplash.com/photos/TMwHpCrU8D4)\n",[682,679,230,769,108,9],"kubernetes",{"slug":771,"featured":6,"template":685},"from-idea-to-production-on-thousands-of-clouds","content:en-us:blog:from-idea-to-production-on-thousands-of-clouds.yml","From Idea To Production On Thousands Of Clouds","en-us/blog/from-idea-to-production-on-thousands-of-clouds.yml","en-us/blog/from-idea-to-production-on-thousands-of-clouds",{"_path":777,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":778,"content":784,"config":792,"_id":794,"_type":13,"title":795,"_source":15,"_file":796,"_stem":797,"_extension":18},"/en-us/blog/fuzzit-acquisition-journey",{"title":779,"description":780,"ogTitle":779,"ogDescription":780,"noIndex":6,"ogImage":781,"ogUrl":782,"ogSiteName":669,"ogType":670,"canonicalUrls":782,"schema":783},"Fuzzit - GitLab journey","From a bootstrap startup to integral part of GitLab.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749664102/Blog/Hero%20Images/gitlab-values-cover.png","https://about.gitlab.com/blog/fuzzit-acquisition-journey","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"Fuzzit - GitLab journey\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"Yevgeny Pats\"}],\n        \"datePublished\": \"2020-10-22\",\n      }",{"title":779,"description":780,"authors":785,"heroImage":781,"date":787,"body":788,"category":789,"tags":790},[786],"Yevgeny Pats","2020-10-22","\n\n{::options parse_block_html=\"true\" /}\n\n\n\n## Fuzzit Story\n\nFuzzit started in early 2019 by myself as a spin-off project from my consulting company.\nThe consulting revenue gave me the opportunity to dedicate time and explore the fuzzing-as-a-service idea a bit more without taking VC money too early and becoming “locked-in”.\n\nAfter about 6 months, Fuzzit started gaining traction and becoming a leader in the open-source community.\nBeing the first commercial product to offer languages such as: Go, Rust and more, while at the time OSS-Fuzz only supported C/C++ and wasn’t available for all oss projects. \n\nAfter about 8 months once the product matured thanks to input from the open-source users,\nwe went exploring the enterprise market more deeply. We developed that in 3 main directions: \n\n- Enterprise clients interviews and PoCs,\n- Partnerships with various CI providers to expand the reach.\n- Enterprise focused VCs\n\nIn that process we were lucky to meet with GitLab,\nwhere after a few calls it became apparent this could be a great fit for both sides to pursue  an acquisition (I’ll expand on that later on here).\n\nAt that point in time, we had to decide if we were either moving forward with an acquisition or going to raise funding to try and build a large business.\nIn our process of exploring the fuzzing enterprise market,\nwe understood that if we want to build a big DevSecOps company we would need to expand the offering far beyond continuous coverage-fuzzing.\nThis is of-course possible but will create even more fragmentation in the already fragmented market,\nand will require a substantial amount of financial investment.\nThe opportunity to join a unique place like GitLab for me personally and the amazing technological fit \nfor Fuzzit to be supported natively in a complete [DevSecOps platform](/solutions/security-compliance/), made the decision easy for me. \n\n## Acquisition process\n\nBeing part of a few acquisitions (some successful and some not)\nI can say first hand that the acquisition process is always a complex one,\nwhere only few acquisitions close in the end and many fall in various stages of the process.\nThe acquisition process was very transparent and efficient, as documented in the [handbook](https://about.gitlab.com/handbook/acquisitions/).\n\nCompletely by chance the head of corp dev, Eliran Mesika, is an Israeli which made things very easy for me personally as I could speak and negotiate in my mother tongue.\nGitLab grew in the last two years to over 1200 people, doubling the team, so understanding the structure and driving the process are not easy feats.\nThe process was very transparent even with some unexpected delays/bumps on the way.\n\nDuring the acquisition process I had the chance to meet quite a few people from the Secure team\nwhere we discussed the technology, how the integration will look like and make sure it’s a good fit for everyone both in terms of technology and culture.\nAfter term-sheet was signed, it was mainly legal-work and once that was complete I joined GitLab!\n\n## Joining GitLab\n\nMy vision at Fuzzit was to advance continuous coverage-guided fuzzing adoption to make software more secure.\nI’m only 5 months in but I feel that this vision fits perfectly at GitLab with its shift-left strategy and single DevSecOps application.\nI believe native support for continuous coverage-guided fuzzing in GitLab will lower the barrier to entry for developers, increase adoption and will make software more secure.\nI still have a lot of work and learning to do at Gitlab to achieve the above but so far we have made great progress.\n. It has been an awesome experience for me and hopefully for everyone else here who was involved!\n\nYou can checkout the current state and documentation of coverage-guided fuzzing in GitLab [here](https://docs.gitlab.com/ee/user/application_security/coverage_fuzzing)\n\n## Future\n\nStay tuned for future fuzzing features and blogs!\n","unfiltered",[705,791,9],"security",{"slug":793,"featured":6,"template":685},"fuzzit-acquisition-journey","content:en-us:blog:fuzzit-acquisition-journey.yml","Fuzzit Acquisition Journey","en-us/blog/fuzzit-acquisition-journey.yml","en-us/blog/fuzzit-acquisition-journey",{"_path":799,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":800,"content":806,"config":812,"_id":814,"_type":13,"title":815,"_source":15,"_file":816,"_stem":817,"_extension":18},"/en-us/blog/gemnasium-our-gitlab-journey",{"title":801,"description":802,"ogTitle":801,"ogDescription":802,"noIndex":6,"ogImage":803,"ogUrl":804,"ogSiteName":669,"ogType":670,"canonicalUrls":804,"schema":805},"Gemnasium: Our GitLab journey","We joined GitLab as a small startup and quickly became an integral part of the company. We want to share our success story with the startup community.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749679751/Blog/Hero%20Images/gemnasium-gitlab-cover.png","https://about.gitlab.com/blog/gemnasium-our-gitlab-journey","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"Gemnasium: Our GitLab journey\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"Philippe Lafoucrière\"}],\n        \"datePublished\": \"2019-04-30\",\n      }",{"title":801,"description":802,"authors":807,"heroImage":803,"date":809,"body":810,"category":298,"tags":811},[808],"Philippe Lafoucrière","2019-04-30","\n\nGoing through acquisition is never easy, and often fails eventually. Two common scenarios can occur:\nEither the two companies are of similar size, and in this “merger” configuration employees are\nscared of duplicate jobs; or the buyer is slightly bigger than the seller, and there's a risk of losing\nthe culture and cohesion of the team. Ours was the latter: We were afraid of being absorbed and\ndigested completely, and eventually not working on the same subjects anymore, or not together.\nI’ve spent years building a trusting relationship with my team, and I was worried about\ntheir future in this new adventure.\n\nThis story is different. We’ve been at GitLab for over a year now, and it's all been for the best.\nWe still have the pleasure of working together, even though the team has doubled in size since.\nI wanted to share some reflections on Gemnasium's experience of being acquired by and integrated into GitLab:\n\n## Negotiating the acquisition\n\nWe had our share of ups and downs during the development of Gemnasium. I learned to be very\ncautious about my business relations. This due diligence is always a critical step, especially for\nthe buyer, to avoid any surprises and ensure the quality of the purchased product.\nDuring that step, we couldn’t answer all the requests from GitLab, since sharing algorithms and\nsource code was putting us at risk. But we explained why and managed to provide something\nclose enough to fulfil the requirement. We had open and healthy discussions at that point with\nGitLab, and it helped to create the trust we were looking for.\n\n## Joining GitLab\n\nWhen we joined the company, I was amazed to see everyone contributing to the [handbook](/handbook/).\nLiterally everyone, including PeopleOps, Sales, and Marketing. Committing changes with Git is the\nDNA of the company, and really makes a difference. There’s no one left behind, struggling with\nthe inherent technical difficulties of contributing to a shared repository.\nThen I discovered what fuels GitLab to make it so special: Slack. A lot of companies already use\nSlack, often for the best. But with GitLab being an all-remote company, Slack is a main communication\nchannel for everyone, including with other teams. At the time, GitLab was already present in\n40 countries (vs more than 50 as of today), which means a lot of time zones covered. There’s\nalways someone available to help and answer questions.\nEven administrative problems are taken care of by the People Ops team in a few minutes. Not\ndays, not weeks – minutes. It allows all employees to focus on what really matters: Delivering and\nmaking the product better. No need to follow up anymore, nothing to complain about; the burden\nis just gone, and everybody moves on. When you make the life of your employees easier, they are\nhappier and more productive. As simple as that.\n\n## Concluding the Gemnasium story\n\nThe acquisition, like everything else at GitLab, went extremely quickly – so quickly that we didn’t have the\nopportunity to bond one last time together as a team. That was a concern to me, also because\nthe onboarding was overwhelming. So many questions, so many processes and new concepts to\ndigest. Our Product Manager Fabio Busatto was really helpful and did everything he could\nto get up to speed as soon as possible. It felt obvious that we would benefit from having a “retreat”\nin a common place, to close the Gemnasium story, and put the new GitLab one on track. “We don’t\ndo that at GitLab, we’re a remote company,” was the first answer I got. I didn’t have to insist\ntoo much to convince our CEO [Sid](/company/team/#sytses), and I promised to keep everything cheap and neat. We already had two\nteam members in Quebec City, so it made sense to organize something here, to save on travel.\n“[Everything starts with an issue](/handbook/communication/#everything-starts-with-an-issue)” at GitLab, so I created\n[one to make this case](https://gitlab.com/gitlab-com/people-ops/General/issues/99).\nA few days later, the idea and budget were approved without any trouble. We could spend a\nwhole week altogether, and it was a great experience for all of us. The feedback from the team\nwas very positive, and it boosted morale as well.\n\n## Becoming the Secure Team\n\nAs an official part of GitLab, the Gemnasium team became the Security Products\nTeam, now called the [Secure Team](/handbook/engineering/development/sec/secure/). Our scope is much broader than just dependency scanning, and\nwe were expected to deploy SAST, DAST, and Container Scanning solutions. It took us less than a month\nto deliver an [MVC](https://handbook.gitlab.com/handbook/values/#minimal-viable-change-mvc) of dependency scanning, based on Gemnasium. We were already working\nremotely, using GitLab, so the pipeline and other parts of the equation were familiar to us. Before the\nnext milestone following our arrival, Gemnasium was running on GitLab infrastructure.\n[Dmitriy](/company/team/#dzaporozhets) (CTO) and Sid (CEO) were really present, taking the pulse of the team, and helping us\nto remove any roadblocks. They didn’t try to force us to do this integration their way. It was really\na collaboration and every meeting began with them asking, “How can we help you?”\n\n## Transitioning from manager to individual contributor\n\nAfter a few months, it became obvious that the team was performing well and heading in the right direction.\nWe had results, customers, and a huge roadmap ahead of us. It was time to start hiring new engineers.\nBack then, I was meeting with a lot of customers, gathering feedback and ideas to help our product\nmanager, and helping with pre-sales. Hiring new engineers can be very time consuming, and\nwith our expectations for the Secure Team, that means a full-time job for a while. Instead of\nforcing me to stop what I was doing and start right away with the recruiting, my manager\ndecided to leverage my skills. I was recently promoted to a Distinguished Engineer position,\nwhich also means switching from the [management branch to the Individual Contribution path](/handbook/engineering/career-development/#individual-contribution-vs-management).\n\nThis is a big shift for me and the team, but in the end, it results in more space and latitude to\nwork on various subjects: Developing our Security Products is much more than just a roadmap\nand implementation. We need to understand the competition, discuss strategic partnerships,\nidentify risks and opportunities, and many other things left aside during all our regular processes.\nBeing my own boss for the last 10 years taught me to be efficient and put the team in the best position for success.\n\n> The bureaucracy that's often associated with large organizations\nis very limited, even after growing to more than 500 people\n\nMy manager, [Dalia Havens](/company/team/#dhavens), has been nothing but supportive\nin this area since the beginning, and a great [servant leader](https://en.wikipedia.org/wiki/Servant_leadership).\nGitLab has been successful so far because the bureaucracy that's often associated with large organizations\nis very limited, even after growing to more than 500 people. As soon as a roadblock is identified,\nwe can discuss collaborate to fix the problem, sometimes right away.\nReducing the number of steps necessary to actually achieve or deliver something is one of the keys\nto happiness for a team used to iterating daily.\n\nI think this is the main reason for the success of this acquisition. At no time did GitLab try to put us in a box.\nAs soon as the results are there, we’re free to experiment, to innovate, and more importantly, to build our own future.\n\n## Experimenting and innovating\n\nOne good example of this freedom to explore is the [auto-remediation feature](/direction/secure/#auto-remediation).\nIn 2014, we shipped our second iteration of the Auto-Update in Gemnasium. While the\nalgorithm behind the update sets had been improved, we were aware that the setup was far from\nsimple, which was against the philosophy behind Gemnasium: In order to work, our algorithm\nhad to run the pipeline, maybe multiple, consecutive times (with different update sets).\nThis was clearly hard to achieve for our users, and for our developers (we didn’t know anything about the test suite).\nBeing part of GitLab would solve that issue, as we would eventually be able to pilot the pipeline\nfor that. Even better, we would be able to hide the runs from the users.\n\nAfter a few customer meetings, it was obvious that this feature was a competitive advantage, and we decided to push\nit more. The whole team was excited to contribute to what would be the first MVC, as our product\nmanager helped to refine the feature, gluing all the pieces together. This step was essential:\nIt allowed everyone to contribute and influence the roadmap. Even as the company gets bigger\nevery day, we still feel empowered and a part of the decision-making process.\n\nThese past 12 months have been extremely exciting and rewarding. While we’re now fully integrated\ninto GitLab, we still feel the fresh air of freedom we had during the Gemnasium years. Even\nbetter, we can focus on what we love, and stop worrying about the short-term future.\n\nIf you're interested in being acquired by GitLab, we're actively looking for startups to join us.\nPlease visit our [acquisitions handbook](/handbook/acquisitions/) to find out more and to see if you\nare the right fit.\n",[705,791,9],{"slug":813,"featured":6,"template":685},"gemnasium-our-gitlab-journey","content:en-us:blog:gemnasium-our-gitlab-journey.yml","Gemnasium Our Gitlab Journey","en-us/blog/gemnasium-our-gitlab-journey.yml","en-us/blog/gemnasium-our-gitlab-journey",{"_path":819,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":820,"content":826,"config":832,"_id":834,"_type":13,"title":835,"_source":15,"_file":836,"_stem":837,"_extension":18},"/en-us/blog/gitlab-acquisitions",{"title":821,"description":822,"ogTitle":821,"ogDescription":822,"noIndex":6,"ogImage":823,"ogUrl":824,"ogSiteName":669,"ogType":670,"canonicalUrls":824,"schema":825},"A guide to GitLab’s soft landing acquisitions","Find the team a new home, release your technology to a wider user base, and continue to build products you love through a soft-landing acquisition.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749680671/Blog/Hero%20Images/soft-landing-acquisitions.jpg","https://about.gitlab.com/blog/gitlab-acquisitions","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"A guide to GitLab’s soft landing acquisitions\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"Eliran Mesika\"}],\n        \"datePublished\": \"2019-07-30\",\n      }",{"title":821,"description":822,"authors":827,"heroImage":823,"date":829,"body":830,"category":298,"tags":831},[828],"Eliran Mesika","2019-07-30","\n\nA few months ago we published our [acquisitions handbook](/handbook/acquisitions/). A first of its kind, it provides a clear view on how we approach and carry out acquisitions at GitLab. We believe this handbook is the basis for effective communication and expectation setting.\n\nOur unique approach to acquisition is suited for companies which have built great technologies but were unable to reach the desired distribution and are nearing the end of their runway. For companies in this state we are offering an opportunity for soft landing in GitLab through acquisition, finding the team a new home, releasing the technology you developed to the greater GitLab user base, and continuing to build awesome products you love.\n\n## Is this relevant for your company?\n\nIf you’re a technology company:\n1. Operating in the expanded [DevOps space](/direction/)\n2. With a team of 10 employees or fewer\n3. At the end of your runway and/or thinking about winding down\n4. Open to a soft-landing acquisition and ready to move through the process quickly\n\n... then your company is potentially a great fit for our soft-landing acquisition process.\n\n## What GitLab has to offer\n\n1. Assets will be purchased for up to $1M total, all cash. GitLab stock will not be offered as part of the deal for the assets sold.\n2. We believe talent follows leadership they trust. In addition to the purchase price, GitLab will offer cash bonuses for founders and engineers to help in the transition, conditional on employee interviews and offer acceptance:\n   - Each founder with more than 10% ownership of the company will receive $250,000 paid as follows: $50,000 on closing and $200,000 as a retention bonus\n   - Each engineer will receive $60,000 paid as follows: $12,000 on closing and $48,000 as a retention bonus\n1. Triple our normal stock option grants for founders, normal stock option grants for non-founders\n\nWe invite you to take a closer look at our acquisitions [handbook page](/handbook/acquisitions/) and reach out to myself, the acquisitions lead,  eliran@gitlab.com, to start a discussion.\n\nIt's important to add that we're open to other types of acquisitions, aside from the soft-landing type. We've felt it's beneficiary to all sides of a soft-landing acquisition to have a streamlined, fast process, which is why we've created ours at GitLab. If you'd like to engage us in an acquisition conversation, again, feel free to reach out to me at eliran@gitlab.com.\n\nYou can also read about [one startup's experience of being acquired by GitLab](/blog/gemnasium-our-gitlab-journey/).\n\nCover image by [Pascal Meier](https://unsplash.com/photos/UYiesSO4FiM) on [Unsplash](https://unsplash.com)\n{: .note}\n",[679,9,706],{"slug":833,"featured":6,"template":685},"gitlab-acquisitions","content:en-us:blog:gitlab-acquisitions.yml","Gitlab Acquisitions","en-us/blog/gitlab-acquisitions.yml","en-us/blog/gitlab-acquisitions",{"_path":839,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":840,"content":845,"config":851,"_id":853,"_type":13,"title":854,"_source":15,"_file":855,"_stem":856,"_extension":18},"/en-us/blog/gitlab-live-event-recap",{"title":841,"description":842,"ogTitle":841,"ogDescription":842,"noIndex":6,"ogImage":718,"ogUrl":843,"ogSiteName":669,"ogType":670,"canonicalUrls":843,"schema":844},"Here's what went down at #GitLabLive","We went live today to discuss our $100m Series D funding and what's next for GitLab – catch up on the recording here.","https://about.gitlab.com/blog/gitlab-live-event-recap","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"Here's what went down at #GitLabLive\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"Rebecca Dodd\"}],\n        \"datePublished\": \"2018-09-20\",\n      }",{"title":841,"description":842,"authors":846,"heroImage":718,"date":848,"body":849,"category":298,"tags":850},[847],"Rebecca Dodd","2018-09-20","\n\nAfter [yesterday's big news](/blog/announcing-100m-series-d-funding/), we held a\n[#GitLabLive](https://twitter.com/search?q=%23GitLabLive&src=tyah) event today to dive into what this means for us, for you, and for GitLab the product.\n\nAs you can tell, we were all pretty excited about it:\n\n{::options parse_block_html=\"false\" /}\n\n\u003Cdiv class=\"center\">\n\n  \u003Cblockquote class=\"twitter-tweet\" data-partner=\"tweetdeck\">\u003Cp lang=\"en\" dir=\"ltr\">Waiting for \u003Ca href=\"https://twitter.com/hashtag/GitLabLive?src=hash&ref_src=twsrc%5Etfw\">#GitLabLive\u003C/a> like \u003Ca href=\"https://t.co/eqw4ljZXaa\">pic.twitter.com/eqw4ljZXaa\u003C/a>\u003C/p>— Brendan O'Leary 👨🏻‍💻 (@olearycrew) \u003Ca href=\"https://twitter.com/olearycrew/status/1042809056275193856?ref_src=twsrc%5Etfw\">September 20, 2018\u003C/a>\u003C/blockquote>\n  \u003Cscript async src=\"https://platform.twitter.com/widgets.js\" charset=\"utf-8\">\u003C/script>\n\n\u003C/div>\n\nHosted by Director of Cloud Native Alliances [Priyanka Sharma](https://twitter.com/pritianka), the event covered GitLab past,\npresent, and future with GitLab team-members, investors, and customers. You can watch the whole thing below:\n\n\u003C!-- blank line -->\n\u003Cfigure class=\"video_container\">\n  \u003Ciframe src=\"https://www.youtube.com/embed/ZgFqyXCsqPY\" frameborder=\"0\" allowfullscreen=\"true\"> \u003C/iframe>\n\u003C/figure>\n\u003C!-- blank line -->\n\n### Get the slides\n\n\u003Cfigure class=\"video_container\">\n\u003Ciframe src=\"https://docs.google.com/presentation/d/e/2PACX-1vTO_mVE0psqDSIOwmrv30ebL0IMdAIhYFHqBcoqI6b8_Cl1yl8f6FaAIm-d7qwsOWhhiUIqPxo6fjhH/embed?start=false&loop=false&delayms=3000\" frameborder=\"0\" width=\"1280\" height=\"749\" allowfullscreen=\"true\" mozallowfullscreen=\"true\" webkitallowfullscreen=\"true\">\u003C/iframe>\n\u003C/figure>\n\nJust as an aside:\n\n\u003Cdiv class=\"center\">\n\n  \u003Cblockquote class=\"twitter-tweet\" data-lang=\"en\">\u003Cp lang=\"en\" dir=\"ltr\">There’s a rumor going around that we borrowed the set of “Between Two Ferns” for \u003Ca href=\"https://twitter.com/hashtag/GitLabLive?src=hash&ref_src=twsrc%5Etfw\">#GitLabLive\u003C/a>, can neither confirm nor deny 😆\u003C/p>— GitLab (@gitlab) \u003Ca href=\"https://twitter.com/gitlab/status/1042830634366853121?ref_src=twsrc%5Etfw\">September 20, 2018\u003C/a>\u003C/blockquote>\n  \u003Cscript async src=\"https://platform.twitter.com/widgets.js\" charset=\"utf-8\">\u003C/script>\n\n\u003C/div>\n\n## Highlights\n\n### [Company update](https://youtu.be/ZgFqyXCsqPY?t=57s)\n\nOur CEO and co-founder [Sid Sijbrandij](/company/team/#sytses) chats with Priyanka about some major company\nmilestones, from our beginnings at YCombinator, to recently being ranked [#44 on Inc. 5000's\nlist of the fastest-growing companies in the US for 2018](/blog/gitlab-ranked-44-on-inc-5000-list/), and our [CI solution being recognized by Forrester as a Leader in that space](/blog/gitlab-leader-continuous-integration-forrester-wave/). He also talks about our acquisition of Gemnasium, which spurred development of GitLab's integrated security features, as well as our focus on cloud native and [Kubernetes](/solutions/kubernetes/).\n\nReaching 2,000 contributors recently was also a landmark achievement for us, and we're proud to have switched to a [DCO for source code contributions](/blog/gitlab-switches-to-dco-license/) to make it even easier for everyone to contribute. We're thrilled to have a number of [foundational open source projects call GitLab home](/blog/welcome-gnome-to-gitlab/) now, giving weight to [#movingtogitlab](/blog/why-move-to-gitlab/).\n\nAll of this has been building towards yesterday's news: our [series D funding of $100 million](/blog/announcing-100m-series-d-funding/) to help us realize our vision of beating out nine other, best-in-class products with a single application.\n\nHow does it feel being part of the unicorn club?\n\n> \"It's exciting. When we came to Silicon Valley, YCombinator explained to us that if you're going to raise money from external investors, your aim should be to become a billion dollar company. Otherwise you should not raise any money and we seriously considered it. But we opted to raise the money and now our early shareholders can feel confident that we've got here. But we raised more money now so the bar is higher and we're going to try to keep growing the company.\" - Sid Sijbrandij, CEO, GitLab\n\n### [Why invest in GitLab?](https://youtu.be/ZgFqyXCsqPY?t=14m20s)\n\nMatthew Jacobson, General Partner at ICONiQ Capital, joins to share some insight into why they've invested in GitLab. With a focus on growth-stage investments, they look for product velocity and the strength and quality of the team.\n\nThe conversation between ICONiQ and GitLab started over two years ago, where the \"maniacal focus on product\" at GitLab became clear and the breadth of our ambition made a real impression. Nine categories is an ambitious product vision!\n\n### [Scaling a remote work culture](https://youtu.be/ZgFqyXCsqPY?t=22m4s)\n\nOur culture is extremely important to us. Chief Culture Officer [Barbie Brewer](/company/team/#BarbieJBrewer) joins to shed some light on how we preserve it as we scale, keeping people front and center at all times:\n\n> \"We focus on working with the best people, getting the best contributors, and building the best product... We have our [values](https://handbook.gitlab.com/handbook/values/) at the core of everything we do. We give each other feedback and push each other to be better.\" - Barbie Brewer, Chief Culture Officer, GitLab\n\nThe company growing doesn't necessarily mean the workforce needs to grow 1:1: \"We're not just growing fast, we're growing smart,\" said Barbie.\n\nBarbie also reiterated our commitment to [diversity, inclusion and belonging](https://handbook.gitlab.com/handbook/values/#diversity-inclusion), sharing some of the ways we encourage and empower GitLab team-members to uphold these values and help each other to learn and grow along the way. As always, our [handbook](/handbook/) is our single source of truth.\n\n### [Product update](https://youtu.be/ZgFqyXCsqPY?t=33m58s)\n\nTo fill us in on what's new with GitLab the product, we're joined by [William Chia](/company/team/#thewilliamchia), Manager, Product Marketing. We delivered the full software development lifecycle at the end of 2016, then set our sights on [Concurrent DevOps](/blog/from-dev-to-devops/). William shares how conversations with users and customers alerted us to the \"toolchain crisis\" and how this has inspired us to deliver Concurrent DevOps with a single application covering the entire DevOps lifecycle.\n\n#### [User perspective: Why GitLab?](https://youtu.be/ZgFqyXCsqPY?t=51m26s)\n\nWe hear from Michael Sobota, Director of Product Integration at Charter Communications, about their company goals of quick, iterative development, shifting operations concerns left, and how they're using GitLab as their DevOps platform to get there. They've gone from feedback cycles of **two weeks** to a matter of **minutes** – ultimately helping them to deliver a better customer experience.\n\n> \"Gone are the days of managing multiple build machines. It's all in the power of developers.\" - Michael Sobota, Director of Product Integration, Charter Communications\n\n### [Product vision](https://youtu.be/ZgFqyXCsqPY?t=1h2m50s)\n\nHead of Product [Mark Pundsack](/company/team/#MarkPundsack) joins to share our ambitious product vision, and we're so excited about it, we're dedicating a post to it on its own! In case you just can't wait, here's the rundown:\n\n1. GitLab is a complete [DevOps platform](/solutions/devops-platform/), delivered as a single application, enabling [Concurrent DevOps](/topics/concurrent-devops/).\n1. We're going to double down on what's working and focus on depth, breadth, and adding new roles to the product.\n1. In 2019, we plan to become leaders in four new areas: project management, continuous delivery and release automation, application security testing, and value stream management.\n1. We have 26 new product capabilities planned for 2019.\n1. DevOps isn't just about developers and operations. We plan to cover roles like designers and product managers so everyone can work concurrently in a single product.\n\n### [Q&A](https://youtu.be/ZgFqyXCsqPY?t=1h13m51s)\n\nSid's back in the house! He answers some audience questions, and encourages everyone to make suggestions for how to improve GitLab:\n\"Many times the hardest thing is figuring out what to make, not how to make it.\" More proof that [everyone can contribute](/company/strategy/#why)!\n",[276,727,705,727,9],{"slug":852,"featured":6,"template":685},"gitlab-live-event-recap","content:en-us:blog:gitlab-live-event-recap.yml","Gitlab Live Event Recap","en-us/blog/gitlab-live-event-recap.yml","en-us/blog/gitlab-live-event-recap",{"_path":858,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":859,"content":865,"config":871,"_id":873,"_type":13,"title":874,"_source":15,"_file":875,"_stem":876,"_extension":18},"/en-us/blog/gitlab-raises-20-million-to-complete-devops",{"title":860,"description":861,"ogTitle":860,"ogDescription":861,"noIndex":6,"ogImage":862,"ogUrl":863,"ogSiteName":669,"ogType":670,"canonicalUrls":863,"schema":864},"GV leads $20M series C funding for GitLab to complete DevOps","We went live today with The Changelog’s Adam Stacoviak and Jerod Santo to announce $20M in new funding, a new board member, and our vision for Complete DevOps.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749671288/Blog/Hero%20Images/gitlab-live-event.png","https://about.gitlab.com/blog/gitlab-raises-20-million-to-complete-devops","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"Announcing $20 million in Series C round funding led by GV to complete DevOps\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"GitLab\"}],\n        \"datePublished\": \"2017-10-09\",\n      }",{"title":866,"description":861,"authors":867,"heroImage":862,"date":868,"body":869,"category":298,"tags":870},"Announcing $20 million in Series C round funding led by GV to complete DevOps",[723],"2017-10-09","Update: for the most recent status of complete [DevOps](/topics/devops/) please see our [Product Vision](/direction/) page.\n\nToday we are thrilled to announce our $20 million Series C funding led by GV. This follows [our Series B round last September](/blog/gitlab-master-plan/). With the help of our investors (and community!) we’re gearing up to bring you Complete DevOps, a reimagined scope of DevOps that unifies development and operations work into a single user experience.\n\n\u003C!-- more -->\n\nNot a GitLab user? [Install GitLab](/install/) or [sign in](https://gitlab.com/users/sign_in) to get started!\n{: .alert .alert-gitlab-orange}\n\nIn addition to our Series C funding round led by Dave Munichiello, GV General Partner, we’re excited to announce new board member,\nMatt Mullenweg, founder of WordPress.\n\n> \"The Fortune 500 is racing to build world-class software development organizations that mirror the speed,\nproductivity, and quality of the largest tech companies. As these organizations strive to produce high-quality\ncode at scale, they will need best-in-class tools and platforms. GitLab’s platform accelerates\nthe development process with an emphasis on collaboration and automation.\nGitLab’s hybrid, multi-cloud solution is loved by developers, and is seeing tremendous traction in the field.\" – Dave Munichiello, GV General Partner\n\n> \"GitLab’s powerful momentum and scaling have a lot of parallels to Automattic and WordPress in their early days.\nWordPress had to battle a lot of competitors, and ultimately came out on top as a successful company on an open source business model.\nI hope to help GitLab achieve the same triumph. Fundamentally, I want to help create the kind of internet that I want to live in\nand I want my children to live in, one that reaches a global audience and one that is able to make a difference.\"\n– Matt Mullenweg, founder of WordPress\n\nSince our start in 2014, we’ve had one mission: change all creative work from read-only to read-write so that everyone can contribute.\nLast year we unveiled GitLab’s Master Plan on September 13th, committing to shipping every stage of idea to production (which we completed in 8.15!).\nThis was a major step forward in simplifying the software development process.\nNow, we're taking it a step further to unite development and operations in one user experience.\nWatch the recording of our earlier live stream announcing our #CompleteDevOps vision below,\nand keep scrolling for a recap and the slides from the presentation.\n\n\u003Ciframe width=\"560\" height=\"315\" src=\"https://www.youtube.com/embed/5dhjw-TT964\" frameborder=\"0\" allowfullscreen>\u003C/iframe>\n\n## Complete DevOps\n\nBefore DevOps, the world of software iteration was slow, insecure, and error prone.\nDevOps came to the intersection of development and operations to create faster iteration cycles with greater quality and security.\n\nBut it didn't go far enough...\n\nIn the current landscape, developers and operations use different tools,\nthey don't have the ability to fully collaborate, and the need to integrate\nmany disparate tools continues to be a point of friction that slows progress\nand leads to insecure, poor quality code.\n\nComplete DevOps reimagines the scope of tooling to include both developers\nand operations teams in one unified solution. This dramatically reduces friction,\nincreases collaboration, and drives a competitive advantage.\n\nIn [10.0](/releases/2017/09/22/gitlab-10-0-released/), we shipped the first iteration of Auto DevOps,\nwhich just scratches the surface of the Complete DevOps features we have in the works.\nYou can read our Head of Product [Mark Pundsack](/company/team/#MarkPundsack)’s\ndetailed vision in [his blog post](/blog/devops-strategy/), but to summarize:\n\nWe want to build GitLab into the complete DevOps tool chain.\nWe already cover every stage of the software development lifecycle. Why stop at production?\nWhy not go beyond that, into operations? We want to close the loop between Dev and Ops,\nautomating processes and reducing complexity so that you can focus on a great customer experience.\n\n\u003Ciframe src=\"https://docs.google.com/presentation/d/e/2PACX-1vRVKUjMMa7M7lPV04_TMgfmd2Fj_kEQYW9-RvKAtKf799_Dwbfvos8diqinI-Uhm1uTwPYCdAPPzun1/embed?start=false&loop=false&delayms=3000\" frameborder=\"0\" width=\"1280\" height=\"749\" allowfullscreen=\"true\" mozallowfullscreen=\"true\" webkitallowfullscreen=\"true\">\u003C/iframe>\n\n### Why Complete DevOps?\n\n1. A single UI for development and operations means less time is wasted switching tools.\n2. All phases of DevOps are deeply integrated, so development and operations can work together collaboratively with less friction.\n3. The best practices of more than 100K organizations are built in by default.\n4. You benefit from a single install, with upgrades that don't break, no integration work, and one permission model.\n\nShare your thoughts, comments, and questions about #CompleteDevOps with us [on Twitter](https://twitter.com/gitlab)!\n\n### The cloud-native development solution\n\nThe software world is moving from virtual machines to cloud-native development.\nWe want to help ease this transition for companies, by offering a complete development and operations solution for cloud-native development.\n\n## Get involved\n\nWe 💜 our community! At GitLab, everyone can contribute and we owe GitLab’s existence to your enthusiasm,\ndrive, and hard work. Without our contributors’ belief in open source software, we would not be where we are today.\nWe need your help to make our collective vision a reality.\n\nWe are committed to standing by our [promise to be good stewards of open source](/blog/being-a-good-open-source-steward/),\nand keeping communication and collaboration amongst the community a high priority.\nOur open core business model ships both open and closed software.\nIn an effort to maintain an unprecedented level of transparency, we follow three key principles:\n\n1. [Development in the open](/blog/improving-open-development-for-everyone/). You can submit issues in a public issue tracker. This is not a read-only interface.\n1. [Business in the open](/blog/almost-everything-we-do-is-now-open/). Our company handbook and policies are in the open.\n1. [Clear direction](/direction/). Our Direction page clarifies the current project priorities and what is possible in the upcoming releases.\n\nRead more about our company values in our [open source](/blog/our-handbook-is-open-source-heres-why/) [handbook](https://handbook.gitlab.com/handbook/values/), licensed by [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0/).",[727,705,9],{"slug":872,"featured":6,"template":685},"gitlab-raises-20-million-to-complete-devops","content:en-us:blog:gitlab-raises-20-million-to-complete-devops.yml","Gitlab Raises 20 Million To Complete Devops","en-us/blog/gitlab-raises-20-million-to-complete-devops.yml","en-us/blog/gitlab-raises-20-million-to-complete-devops",{"_path":878,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":879,"content":885,"config":893,"_id":895,"_type":13,"title":896,"_source":15,"_file":897,"_stem":898,"_extension":18},"/en-us/blog/gitlab-taught-in-korean-uni",{"title":880,"description":881,"ogTitle":880,"ogDescription":881,"noIndex":6,"ogImage":882,"ogUrl":883,"ogSiteName":669,"ogType":670,"canonicalUrls":883,"schema":884},"Schooled in GitLab: Teaching our handbook at a South Korean university","Students at Hankuk University of Foreign Studies tackled our handbook. The students' favorite topics were compensation and remote work.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749673044/Blog/Hero%20Images/books-internship-post.jpg","https://about.gitlab.com/blog/gitlab-taught-in-korean-uni","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"Schooled in GitLab: Teaching our handbook at a South Korean university\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"Guenjun Yoo\"}],\n        \"datePublished\": \"2020-01-29\",\n      }",{"title":880,"description":881,"authors":886,"heroImage":882,"date":888,"body":889,"category":890,"tags":891},[887],"Guenjun Yoo","2020-01-29","\nBusiness students at [Hankuk University of Foreign Studies](http://mis.hufs.ac.kr/) in Seoul, South Korea are studying the GitLab handbook and business model. The students are enthusiastic about GitLab and its story, says lecturer SanJoon Song in an email interview, but there was one problem: Our 3,000+ page handbook is a lot to swallow in one semester.\n\nSo Song had the class divide the handbook into 15 different categories, which different groups of students researched over the course of the semester. At the end of the term, the groups presented a summary of their category to the class.\n\n“Many engineers in Korea said that the GitLab handbook is good to read before starting up a business,” says Song. “However, there is a lot of reading in the handbook; too many pages for me.”\n\nThe level of transparency in the handbook was a revelation to Song and his students.\n\n“We didn't study [the handbook] only to focus on the content itself, but we tried to understand and share about the context of handbook; what conventions GitLab has and what protocols GitLab is trying to develop with its employees by this handbook,” says Song. “In Korea, this is very unusual to share such details of company goals and protocols with entire employees by handbook and for me, this approach is very new and fresh.”\n\n## Inside information\n\nSong was very surprised by how much “insider” information is available in our handbook and says he’s particularly amazed by the detailed explanations of what to do if things go wrong.\n\nOn the other hand, his students were most impressed by the details on [compensation](/handbook/total-rewards/compensation/compensation-calculator/calculator/) and incentives in the handbook, followed closely by the idea of remote work.\n\n“Personally I liked the concept of [‘accept mistakes’](https://handbook.gitlab.com/handbook/values/#accept-mistakes) in the efficiency section,” says Song. “We also talked a lot about GitLab’s [six values](https://handbook.gitlab.com/handbook/values/).”\n\n![Breaking down the handbook](https://about.gitlab.com/images/blogimages/studyingthehandbook.jpg){: .shadow.medium.center}\nStudents in Song's class breaking down the handbook.\n{: .note.text-center}\n\nRemote work was also a big topic of discussion in Song's classroom.\n\n\"Many Koreans are interested in remote work,\" says Song. \"It is really great that people can work anywhere, anytime without having the stress of commuting. Remote work is not common in Korea yet. Only a few software developers are allowed to work from home but that is also partial and in a limited environment only. Many students also want to do the remote work but this is still kind of a dream.”\n\nSong is currently teaching a second GitLab-focused class, this time diving into project management and DevOps by looking at our product and Pivotal Labs. If there is one benefit Song thinks his students have taken away from studying GitLab it’s the importance of communication.\n\n“Communication between employees is one of the most important matters,\" says Song. \"By studying the GitLab handbook, my students and I learned an efficient way of communication between the employer and employees. The handbook explicitly shows how GitLab is trying to do the best way of communication between stakeholders; what is the company goal, why we established the goal and how we are achieving the goal.”\n\nSong hopes to inspire a future generation of entrepreneurs by studying the GitLab handbook in the classroom.\n\n“My students have studied the GitLab handbook for one semester. I hope this study can be their reference when they start their startup, so they can create their company goals and prototype in the direction of success, like GitLab.\"\n\n_If you’re interested in seeing more of Song’s curriculum, he shared it\n[here](https://docs.google.com/document/d/1u5J6Ypj6zwQJVjmrl1wd0eIv7Q_TYLJysDquhGMJimA/edit). You'll need to scroll down a bit._\n\nCover image by [Patrick Tomasso](https://unsplash.com/@impatrickt) on [Unsplash](https://unsplash.com/)\n{: .note}\n","open-source",[892,266,679,680,706,9],"careers",{"slug":894,"featured":6,"template":685},"gitlab-taught-in-korean-uni","content:en-us:blog:gitlab-taught-in-korean-uni.yml","Gitlab Taught In Korean Uni","en-us/blog/gitlab-taught-in-korean-uni.yml","en-us/blog/gitlab-taught-in-korean-uni",{"_path":900,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":901,"content":907,"config":915,"_id":917,"_type":13,"title":918,"_source":15,"_file":919,"_stem":920,"_extension":18},"/en-us/blog/how-smbs-can-save-with-gitlabs-devops-platform",{"title":902,"description":903,"ogTitle":902,"ogDescription":903,"noIndex":6,"ogImage":904,"ogUrl":905,"ogSiteName":669,"ogType":670,"canonicalUrls":905,"schema":906},"How SMBs can save with the GitLab DevOps Platform","Use our ROI Calculator to understand how a DevOps platform saves money.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749667966/Blog/Hero%20Images/global-compensation-calculator-iteration.jpg","https://about.gitlab.com/blog/how-smbs-can-save-with-gitlabs-devops-platform","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"How SMBs can save with the GitLab DevOps Platform\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"Sharon Gaudin\"}],\n        \"datePublished\": \"2022-09-08\",\n      }",{"title":902,"description":903,"authors":908,"heroImage":904,"date":910,"body":911,"category":912,"tags":913},[909],"Sharon Gaudin","2022-09-08","\n\nMigrating from a complex and costly DevOps toolchain to The One DevOps Platform from GitLab can not only save a small and medium-sized business (SMB) from an inefficient workload, it can result in a big financial savings, too.\n\nAnd that savings could mean the difference between an [SMB failing and thriving](/blog/6-ways-smbs-can-leverage-the-power-of-a-devops-platform/) in a cutthroat and unpredictable market. That’s right… GitLab’s end-to-end platform can turn IT into a business driver that speeds software creation, boosting competitiveness and pulling in more revenue. \n\nThat’s key for SMBs, which have small IT teams or maybe even a team of one. That means there are fewer hands to do the work, and likely less financial resources. SMBs might also have a harder time hiring general IT people who also can develop and deploy code. The benefits of a single DevOps platform help solve several SMB issues.\n\n“Where [migration is an investment](https://learn.gitlab.com/smbmigrationguide/migratedevopssmb) in time and change, it’s an investment that will pay a lot of dividends in time and money savings,” says [Brendan O’Leary](https://gitlab.com/brendan), staff developer evangelist at GitLab. “If it’s done right, the ROI will be very quick. You can get rid of all those other tools, while speeding up your ability to iterate and serve your customers.”\n\n## Use an ROI calculator\n\nWith so many factors to consider, how can IT managers measure potential savings? \n\nGitLab can help with that. Check out our [ROI Calculator](https://about.gitlab.com/calculator/roi/), which can help estimate the financial benefits an SMB could realize by moving to GitLab from their DIY DevOps toolchain.\n\nGetting rid of a tangle of disparate tools means freeing up money spent on licensing fees, as well as on updating and maintenance. It also means freeing up all the time IT people spend context switching between these tools. And don’t forget The One DevOps Platform is going to help teams develop and deploy faster and more efficiently, [making SMBs more nimble and competitive](/blog/can-an-smb-or-start-up-be-too-small-for-a-devops-platform/). \n\n## How to save with GitLab\n\nLet’s look at [how an SMB can save money](https://cdn.pathfactory.com/assets/10519/contents/427544/b901b768-7b0e-4590-b00e-047a80536cdb.pdf) by shifting from a complex toolchain to GitLab’s DevOps Platform:\n\n- License fees are obvious costs that need to be considered when trying to calculate ROI. \n\n- The cost of maintaining the software over time also needs to be factored in. \n\n- Consider how much time and energy is spent on tool upgrades, security patching, and monitoring the performance and overall availability of a multitude of tools.\n\n- Because GitLab’s platform speeds development and deployment, productivity increases and that propels revenue and opportunities to grow the company.\n\n- A complicated toolchain that has users continually jumping between tools and switching interfaces creates a chaotic environment that requires constant management, tweaking, updating, and stitching. That means IT is managing the toolchain instead of actually developing and delivering the code that drives the bottom line. Software isn’t efficiently created in a chaotic environment.\n\n- IT is wasting time, energy, and money getting up to speed on each tool. This goes for every new person who has to learn each tool, instead of a single application.\n\n- SMBs find new customers – and the revenue they bring – by creating software that satisfies customers’ needs. You can do that more quickly with a DevOps platform.\n\n- Because GitLab’s DevOps Platform enables companies to develop and deploy more securely, more quickly, and with less hands-on work, SMBs are more able to change on a dime to meet or get ahead of new demands and even new competitors.\n\n- If an IT team is spending time on the care and feeding of a toolchain instead of doing interesting software development, it can cause stress and job dissatisfaction, which could lead to problematic turnover.\n\n- The One DevOps Platform naturally pushes security left so it’s automatically integrated into every step of the development lifecycle. Detecting errors early in the process is much cheaper and less time consuming than detecting them in production. \n\nThat’s a lot of ways to save money and earn extra revenue. And all of that can be done with one single license, one permission model, and one interface, giving teams the time and resources to focus on creating business value instead of managing a toolchain.\n\n## Drilling down on ways to save\n\nGitLab’s DevOps Platform allows teams to move from, or avoid, that often complex and confusing multitude of tools by using a single, complete software development ecosystem. An SMB may be small enough that it hasn’t amassed a complex toolchain – yet. But it will only grow more unwieldy as the company grows. Now is the time to adopt a single platform and avoid that problem all together. \n\n\"It enables us to write better software more efficiently,” said Dorian de Koning, DevOps lead at [Weave](https://weave.nl), a software technology developer based in The Netherlands. “We went from single manual deployment approximately every two weeks to tens of deployments a day.\"\n","devsecops",[679,914,9],"customers",{"slug":916,"featured":6,"template":685},"how-smbs-can-save-with-gitlabs-devops-platform","content:en-us:blog:how-smbs-can-save-with-gitlabs-devops-platform.yml","How Smbs Can Save With Gitlabs Devops Platform","en-us/blog/how-smbs-can-save-with-gitlabs-devops-platform.yml","en-us/blog/how-smbs-can-save-with-gitlabs-devops-platform",{"_path":922,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":923,"content":929,"config":936,"_id":938,"_type":13,"title":939,"_source":15,"_file":940,"_stem":941,"_extension":18},"/en-us/blog/how-startups-build-it-infrastructure",{"title":924,"description":925,"ogTitle":924,"ogDescription":925,"noIndex":6,"ogImage":926,"ogUrl":927,"ogSiteName":669,"ogType":670,"canonicalUrls":927,"schema":928},"A way for startups to build a solid IT infrastructure","Seven free software solutions to cover your most important use cases.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749679216/Blog/Hero%20Images/startups-it-infrastructure.jpg","https://about.gitlab.com/blog/how-startups-build-it-infrastructure","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"A way for startups to build a solid IT infrastructure\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"plapadoo\"}],\n        \"datePublished\": \"2017-08-07\",\n      }",{"title":924,"description":925,"authors":930,"heroImage":926,"date":932,"body":933,"category":890,"tags":934},[931],"plapadoo","2017-08-07","\n\n *plapadoo is a software startup from Hannover, Germany, providing tailored, high-quality software engineering to their clients. They fill us in on how they chose solutions for their IT infrastructure, including communication, backups, [CI/CD](/topics/ci-cd/) and more.*\n\n\u003C!-- more -->\n\nWe recently founded [our company](https://plapadoo.de/) and so one of the first things to do was to get our infrastructure up and running. As a software startup, our technical infrastructure is the heart of our company. It influences our productivity, has impact on our costs and offers a great chance to set us apart from the competition. Having a good infrastructure is also key to saving us money and increasing development speed.\n\nWhen planning the setup of our infrastructure, we kept two things in mind: First, we wanted to have open source software running wherever possible, and second, we wanted to use strong encryption for both communication and data storage. Also, we prefer lightweight software with few dependencies. Below, you find a small list of important use cases and which software we use to cover them:\n\n- [Chat](#chat) ([Matrix](https://matrix.org/)/[Riot](https://about.riot.im/) web app + Android app)\n- [Email](#email) (self-hosted [Dovecot](https://www.dovecot.org/) + [Postfix](http://www.postfix.org/) + [Sieve](http://sieve.info/) + [SpamAssasin](http://spamassassin.apache.org/))\n- [Calendar and Contacts](#calendar-and-contacts) ([Radicale](http://radicale.org/))\n- [Voice Conferencing](#voice-conferencing) ([uMurmur](http://umurmur.net/)/[Mumble](https://wiki.mumble.info/))\n- [Synchronization of files across multiple devices](#data-storage) ([Syncthing](https://syncthing.net/))\n- [Git and Continuous Integration](#build-and-continuous-integration) ([GitLab](/stages-devops-lifecycle/) & [GitLab CI](/solutions/continuous-integration/))\n- [Backup and Traceability](#backup-and-traceability) ([borgmatic](https://github.com/witten/borgmatic) & [etckeeper](http://etckeeper.branchable.com/))\nBesides this, we have other services (like VPN or HTTP servers) running which are not that special and as such, are not covered on this article.\n\n## Base setup\n\nIt all starts with choosing the platform to run your software on. We decided to use [Arch Linux](https://www.archlinux.org/) as the operating system for our server. Our main reasons for choosing Arch Linux were its active community, good documentation, highly up-to-date repositories with current versions of important software, good support for disk encryption, and finally, the fact that Arch Linux has a rolling update scheme instead of a release-based one. This last point is especially important to us, since we do not want to go through the pain of upgrading our operating system from one version to the next every other year -  which usually causes lots of trouble. Furthermore, release-based distributions tend to have outdated software in their repositories. Instead, we prefer to keep our system always up to date and enjoy the latest version of any software any time.\n\nMost of our software is installed using Arch Linux’ package manager. However, in some cases [Docker](https://docker.com/) is also a good idea to use for running software. This is especially the case when software introduces dependencies you don’t want on your host system or if you are in doubt about the security of a software. Since Docker provides a certain level of isolation, security breaches don’t have as bad consequences as they have when you are running the vulnerable software directly on your host system. However, it should be kept in mind that there is the risk of a so-called container breakout. This basically means that your host system can be subject to an attack even if the vulnerable software is running inside a Docker container. Other reasons for using Docker can be wanting to try something out without messing up your host system or maybe software is simply not available for your Linux distribution. Of course, there are many other advantages to containerization, but we won’t be covering those today.\n\n## Communication\n\nCommunication, and using appropriate communication channels has been central to us since the very beginning. We wanted a means of communicating that was secure, fast, reliable, and easily accessible from any device. This applies to chat, email, contacts and calendar entries.\n\n### Chat\n\nFor chatting, we needed a solution which supported the concept of a “room” or “channel,” so as to keep discussions clear and separated from each other. We found Matrix/Synapse and Riot to be a perfect solution. While we also tried alternatives, such as Rocket.Chat and Mattermost, we liked Riot/Matrix the most because of its native Android app, its active development, and an open API.\n\nWe are using the Matrix API to run custom chat bots. These bots have become quite an important factor in our company, since they massively increase transparency and information distribution among the team. For example, we have bots to inform us about new commits being pushed to our GitLab server, new calendar entries being created in our shared calendar, successful or failed builds and so on. We will cover these bots in detail in an upcoming article.\n\n### Email\n\nSince we want to have complete control over the data belonging to our core business, we use a private mail server. It is indeed challenging to set up securely, but we still decided to go with it because of how important secure and private communication is to us. We had to read a lot of documentation before we could set it up, most importantly to prevent a security hole in the system. Not doing that would possibly mean ending up on a spammer blacklist, since someone could be abusing our mail server, or an attacker gaining access to our mail. It is a lot of work, but we definitely recommend taking the time to understand every step of the process and avoid any mistakes. On the client side, we seek to encrypt our emails using PGP whenever possible.\n\n### Calendar and contacts\n\nIn order to have a shared calendar as well as a shared address book, we are running Radicale, which is a lightweight CalDAV and CardDAV server. Although it is not easy to configure, it comes with support for Git and just quietly does its job in the background. We have never experienced any problems with this software so far and like it for its reliability. For Android and iOS, there are CalDAV and CardDAV adapters available to synchronize everything with your phone.\n\n### Voice Conferencing\n\nFor voice conferencing, it was very important to us to have a trustworthy open source solution in place. Proprietary solutions always come at the risk of backdoors being shipped along with them. We decided to give Mumble a try. Mumble is an open source voice client that requires a central server to handle all the traffic. The official server implementation is called Murmur. When installing Murmur, we learned that it pulls in a giant bunch of dependencies.\n\nAmong those dependencies are things such as X11 which most people don’t want on their servers. The problem with such dependencies is that they introduce potential attack surfaces as well as costing time, money, and other resources to maintain and update them. So you normally want as few dependencies as possible. This alone would make it a bad fit for us, but we still decided to give it a try. One option would have been to run Murmur inside a Docker container where the mentioned dependencies wouldn’t bother us too much. While we were configuring Murmur, we had to choose a server password. As always, we generated a long, strong password with about 60 random characters (including special characters). As we started the server and tried to connect a client, we were completely shocked. Murmur let clients in without requiring a password.\n\nWe found out that Murmur seems to have a problem with long passwords and then just ignores them. So if you configure Murmur with the goal of strong security, you get no security at all. Needless to say that we immediately uninstalled Murmur and all of the crazy dependencies it introduced.\nWhile looking for alternatives, we soon discovered uMurmur which is an alternative Mumble server implementation aiming at embedded systems. It comes with few dependencies and generally seems to be well implemented. We installed it, did not experience any issues with long passwords and have been using it ever since without any problem. The communication is encrypted using a TLS certificate.\n\n## Data storage\n\n![box files](https://about.gitlab.com/images/blogimages/startups-it-infrastructure-body.jpg){: .shadow}\n\nAnother important aspect within a company besides communication is the need to store and distribute documents among its different stakeholders.\nWhen sharing data, most programmers will normally use Git. However, Git is not to best choice for sharing binary data such as documents, photos, videos, etc., because one usually doesn’t need to keep different versions of these files. A common approach is to use ownCloud/NextCloud for data sharing, but since we really don’t like PHP, we precluded these two applications.\n\nInstead, we discovered Syncthing. Once you understand the concept of Syncthing, it is easy to set up, extremely easy to use and it just works out of the box. Syncthing can be described as a software which synchronizes data across several nodes. We have one Syncthing instance running on our server that acts as a kind of master node, although a master is not explicitly needed -  Syncthing is completely decentralized. We also run Syncthing on our desktops and phones. Each Syncthing node has a unique ID, which has to be added using the web interface of the master node in order to share data with them. For the local node, the unique ID of the master node has to be added accordingly. Using this concept of a master node, we don’t have to wire all our devices to each other -  it is enough to just wire each device to the master node.\n\nAfter that, you can select which folders should be shared using Syncthing. Syncthing will then automatically upload any new data you put into these folders to the remote node. Data added by other users is downloaded to the clients on the fly, and deletions of files, changes, etc. are also applied locally. For Android, there is a native Syncthing app available which does exactly the same. By using Syncthing, all our devices always have the latest version of the data stored inside the Syncthing shares on the master node.\n\n## Build and continuous integration\n\nFor Git and continuous integration, we use GitLab, which already comes with integrated CI features. Although GitLab is quite resource-hungry, it provides lots of very nice features such as an integrated issue tracker and the “snippets” area -  where you can paste code snippets and share them. GitLab is well documented and has an open API. It features webhooks that you can use to trigger HTTP requests whenever commits are pushed, CI pipelines start, and so on. We use that to generate notifications in matrix rooms corresponding to the Git repositories. So, for example, if someone pushes a commit to project “foo,” we get a notification in a Matrix room “room about foo,” which is linked to this project.\n\n>GitLab provides lots of very nice features such as an integrated issue tracker and the “snippets” area -  where you can paste code snippets and share them\n\nWe are using the official GitLab Docker image, which already includes [Prometheus](https://prometheus.io/) for monitoring. We are accessing this Prometheus instance from our host system and plot its data in a dedicated [Grafana](https://grafana.com/) dashboard. This way, we can monitor our GitLab server internals with very little effort.\n\nFor building a project using GitLab CI, you need a so-called “gitlab-runner” that acts as a build agent. There are also official Docker images available for those runners, but we have created our own Docker base image, which has some basic tools we constantly need. We use our custom base image to build individual runners for each project on top of it. This way, we have runners tailored exactly to the needs of our projects. Since the Docker socket is mapped into our runners, we can even build and deploy Docker images from within them.\n\nWe like the fact that the build jobs are defined through a “.gitlab-ci.yml” file that is versioned with each project. This way, you can track changes to the build process and always have a running build - even if you checkout an old version of a project.\n\n## Backup and traceability\n\nBacking up your data is very important. Especially nowadays with the widespread use of SSDs, when fatal disk failure is likely to happen. Other reasons for data loss may be accidental deletion or attacks. We are using [BorgBackup](https://borgbackup.readthedocs.io/) together with borgmatic, which is a nice, simple, incremental, and highly automatable backup solution. You can easily specify files to exclude from the backup, and also select how many daily, weekly, monthly and yearly backups you want Borg to keep. By setting up a Cron job or systemd timer, you can fully automate the backup process. We create backups every night and store them on an NFS storage, which is only mounted when the backup process is running. This way, we avoid the backup to be deleted by an accidental `rm -rf /` or some other mishap. Borg encrypts the backups and supports compression to keep your backups safe and small. We like to keep track of any changes we make to the system, especially those to configuration files.\n\nFor Linux, there is a useful little tool called etckeeper, which turns your `/etc` directory into a Git repository. It also adds hooks to your package manager to automatically commit any configuration changes being performed during system updates. Using etckeeper, every configuration change corresponds to a Git commit, with an author, a timestamp and a message. This provides for much more transparency, especially when more than one person administrates a server. Also, the way Git works, accidental changes are detected and bad configurations can be easily reverted.\n\n## Summary\n\nWe explained that we, at plapadoo, prefer lightweight (in terms of dependencies), focused software over bloated solutions and favor open source software. Our custom chat bot gives us a high level of transparency and awareness, and also improves our productivity, since we always know what’s going on, even if working remotely. Lastly, we explained which software solutions we have chosen for which use cases and why.\n\nIf you liked this article, please help us reach more readers by sharing it. If you have any questions, thoughts or recommendations on the topic, feel free to comment. Which software solutions did you choose for your startup?\n\n_This post was originally published on [Medium](https://medium.com/plapadoo/a-way-for-startups-to-build-a-solid-it-infrastructure-a48b222fbff6/)._\n\n[CERN reception, Meyrin, Switzerland](https://unsplash.com/@samuelzeller?photo=JuFcQxgCXwA) by [Samuel Zeller](https://unsplash.com/@samuelzeller) on Unsplash.\n{: .note}\n",[680,9,935],"user stories",{"slug":937,"featured":6,"template":685},"how-startups-build-it-infrastructure","content:en-us:blog:how-startups-build-it-infrastructure.yml","How Startups Build It Infrastructure","en-us/blog/how-startups-build-it-infrastructure.yml","en-us/blog/how-startups-build-it-infrastructure",{"_path":943,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":944,"content":950,"config":957,"_id":959,"_type":13,"title":960,"_source":15,"_file":961,"_stem":962,"_extension":18},"/en-us/blog/how-we-keep-investors-in-the-loop",{"title":945,"description":946,"ogTitle":945,"ogDescription":946,"noIndex":6,"ogImage":947,"ogUrl":948,"ogSiteName":669,"ogType":670,"canonicalUrls":948,"schema":949},"How we keep investors in the loop","Monthly updates to investors and team members ensure transparency and open communication.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749678944/Blog/Hero%20Images/investorupdate.jpg","https://about.gitlab.com/blog/how-we-keep-investors-in-the-loop","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"How we keep investors in the loop\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"Suri Patel\"}],\n        \"datePublished\": \"2018-10-17\",\n      }",{"title":945,"description":946,"authors":951,"heroImage":947,"date":953,"body":954,"category":955,"tags":956},[952],"Suri Patel","2018-10-17","\nI was a bright-eyed and bushy-tailed new GitLab team-member of two months when I emailed\n[Sid](/company/team/#sytses), the CEO of GitLab, and told him that I thought the investor\nupdate format needed a makeover. During my onboarding, I had heard that\n[**everyone can contribute**](/company/strategy/#why), so I decided to take the idea\nfor a test drive.\n\n![My message to Sid.](https://about.gitlab.com/images/blogimages/suriemail.png){: .shadow}\n\nI obsessively refreshed my inbox, waiting to see whether the sentiment was highly\nregarded only in theory, when Sid’s reply arrived.\n\n![Sid’s response.](https://about.gitlab.com/images/blogimages/sidemail.png){: .shadow}\n\n_Challenge accepted, Sijbrandij._ 😎\n\n## Updating the template\n\nOur investor update has gone through several iterations over the years. In the\nearly days, we included sections on hiring, feedback, and upcoming features.\nAfter reading blog posts by\n[Elad Gil](http://blog.eladgil.com/2015/05/investor-update-emails.html) and\n[Aaron K. Harris](http://www.aaronkharris.com/investor-updates), Sid developed\nour current version with their insights in mind and narrowed the scope of our\nupdate to the following sections: thanks, asks, key metrics, lowlights,\nhighlights, and next month expectations.\n\nWhen I joined GitLab, the investor update looked cluttered, and from a reader’s\nperspective, I had difficulty absorbing the information. Below is an example of\nan old update with the former template. Please note that some names have been\nchanged to respect organizations’ privacy.\n\n![Former investor update template.](https://about.gitlab.com/images/blogimages/oldtemplatev3.png){: .shadow}\n\nKnowing that investors can only dedicate a few minutes to each email, I knew that\nI had to organize the sections and copy in a way that would increase comprehension\nand reading speed, so I employed UX copy techniques.\n\n![New investor update template.](https://about.gitlab.com/images/blogimages/originaltemplate.png){: .shadow}\n\nWith this new format, investors can quickly read the update and locate the\ninformation that is most relevant to them.\n\n### Why these categories are important\n\nEach of the seven sections provides investors with a look inside GitLab, offering\na comprehensive assessment of our monthly performance.\n\n1. **CEO foreword**: A brief introduction that will typically coincide with the close of a fiscal quarter. This narrative will provide a high level overview of company operations from the most recently ended quarter as well as key initiatives and expectations for upcoming quarters. \n1. **Thanks**: We express gratitude for investors who have assisted us with\nmaking introductions, providing feedback, or offering assistance. Investing is a\ntype of social engagement, and we like to celebrate people who set aside time to\nhelp us.\n1. **Asks**: We ask our investors to help us connect with people or\norganizations, introduce us to hiring candidates, or provide some other assistance.\nInvestors can be extremely helpful and often say they want to add\nvalue when they invest, so this area of the update gives them the opportunity to\ndrive our business forward.\n1. **Key metrics**: People want to know how their investment is performing.\nOffering figures instills trust and shows a certain discipline and rigor. We\nwant our investors to know how we’re doing - even when we don’t meet our goals -\nbecause we believe in [transparency](https://handbook.gitlab.com/handbook/values/#transparency).\n1. **Lowlights**: Our commitment to open communication extends to this\nsection in which we always list the top three worst things that occurred in the\nmonth. By committing to three items, the question is no longer, “_Should_ I tell\nmy investors?” It’s “_Which_ three things are the most severe?” That's a\nmuch easier question to answer.\n1. **Highlights**: This section gets people excited about the investment and\nillustrates what we’re doing well.\n1. **Expectations**: We discuss what we’re looking forward to, conferences we’re\nattending, and what we’re planning in the next month.\n\n### Have a fixed number of good and bad things\n\nEvery month we send three lowlights and three highlights.\nThis forces us to always tell the three things that are worst.\nWe never have to wonder if something is important enough to include it.\nBy the severity of the items people can tell if it was a good or bad month.\n\n## Every company should send monthly updates\n\nIf you’re not sending investor updates, you’re keeping your biggest proponents\nin the dark. If people invested in your organization, you should keep them up to\ndate on what's happening with their investment. If investors don’t receive regular\ncommunication, they’re forced to go fishing for information and what they might\nhear could be inaccurate.\n\nMonthly updates instill confidence and save you from having to\nfield questions from several directions. When organizations don’t communicate,\ninvestors constantly have to ping their companies to ask how things are going.\nBut, if they receive regular updates, they know they're going to hear from you\nand learn the most challenging things that happened in the previous month. You\ndon’t want to give your investors any reason to worry.\n\nMonthly updates also help you build stronger bonds with investors. Because the\nbasics of an investment are covered each month, conversations with investors\ncan focus on deeper subjects. You can brainstorm about strategy, long-term\ngoals, and emerging trends rather than recap hiring challenges and share\nrelease updates.\n\n## Investors \u003C3 information\n\nInvestors have told us how much they love our updates, specifically expressing\ntheir appreciation of the reliability of our emails. We send the updates around\nthe 10th (give or take 1-3 days) of every month, so investors have come to\nexpect a little GitLab sunshine in their inbox.\n\nInvestors love the format(!) and [Y Combinator](http://www.ycombinator.com/)\nreached out to Sid asking whether the format could be shared with other YC\nfounders in a resource of high quality updates. As Sid says, “The format seems\nto be better than average.”\n\n>“I want to thank you and also commend you for having such\nconsistent, regular, excellent shareholder communications. It’s rare\nto see, and I think it elevates your company and it’s something that\nonly grows in importance as the company scales.” — GitLab investor\n\n## Now it’s your turn\n\nIf you’d like to send your investors a monthly update, we invite you to\n[create a copy](https://docs.google.com/document/d/1TVpESZlemYWLrXQHeDvnASCFvs_tvjpJnLahEFMIxYE/copy)\nof our template. As you work on your update, please remember that it’s important\nto establish a regular cadence and keep the emails concise. Links to spreadsheets\nwith detailed figures and an offer to answer any questions prevents people from\nbecoming overwhelmed.\n\nOur updates are also sent to team members, because we all have stock options and\nSid believes that it’s the company’s duty to inform us of our investment.\nMoreover, team members should know the highlights, lowlights, and next month's\nexpectations, since we’re all working towards a [common goal](/company/strategy/#sequence).\nWe encourage you to send the updates to your team since they invest their talents,\nideas, and efforts into making your organization successful.\n\nUPDATE: To see what we currently do, see our [Investor Relations page on our Monthly Investor Updates](/handbook/finance/investor-relations/#monthly-investor-update-email).\n","culture",[9,705],{"slug":958,"featured":6,"template":685},"how-we-keep-investors-in-the-loop","content:en-us:blog:how-we-keep-investors-in-the-loop.yml","How We Keep Investors In The Loop","en-us/blog/how-we-keep-investors-in-the-loop.yml","en-us/blog/how-we-keep-investors-in-the-loop",{"_path":964,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":965,"content":971,"config":977,"_id":979,"_type":13,"title":980,"_source":15,"_file":981,"_stem":982,"_extension":18},"/en-us/blog/khosla-ventures-gitlab-meeting",{"title":966,"description":967,"ogTitle":966,"ogDescription":967,"noIndex":6,"ogImage":968,"ogUrl":969,"ogSiteName":669,"ogType":670,"canonicalUrls":969,"schema":970},"Acquisitions, growth curves, and IPO strategies: A day at Khosla Ventures","A CEO Shadow’s take on GitLab’s annual investor meeting with Khosla Ventures.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749671620/Blog/Hero%20Images/khosla-ventures-meeting.jpg","https://about.gitlab.com/blog/khosla-ventures-gitlab-meeting","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"Acquisitions, growth curves, and IPO strategies: A day at Khosla Ventures\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"Erica Lindberg\"}],\n        \"datePublished\": \"2019-04-08\",\n      }",{"title":966,"description":967,"authors":972,"heroImage":968,"date":974,"body":975,"category":298,"tags":976},[973],"Erica Lindberg","2019-04-08","\n\nWhen I accepted the opportunity to participate in GitLab’s [CEO Shadow program](/handbook/ceo/shadow/), I knew exactly what to expect. In typical GitLab fashion, there was already a handbook page detailing the goal, the format, and the expectations of the program. Our co-founder and CEO, [Sid Sijbrandij](https://twitter.com/sytses), keeps his calendar [public by default](https://handbook.gitlab.com/handbook/values/#public-by-default) to GitLab team-members, so I was able to get a good understanding of the meetings I’d be attending, who’d be present, and what would be discussed. What I couldn’t have predicted is how an annual meeting between Sid and venture capitalist [Vinod Khosla](https://en.wikipedia.org/wiki/Vinod_Khosla) would turn out. \n\n## Skeptical about in-person meetings \n\nGitLab is an [all-remote](/company/culture/all-remote/) company. We don’t have any offices, and we communicate and collaborate via Zoom, Slack, Google docs, and GitLab. To me, this is normal, and, as an introvert, my biggest concern with participating in the CEO program was the energy drain I knew I would experience living in downtown San Francisco for three weeks and meeting with people in person. Luckily, even our CEO conducts most of his business remotely. \n\nHowever, many of our investors and board members are still, what I call, remote shy, and tend to default to in-person meetings. This is how I found myself traveling an hour and a half south to Menlo Park to meet with one of our investor groups, Khosla Ventures, for their annual meeting with Sid. Khosla Ventures (KV) is a venture capitalist firm founded by Vinod Khosla, co-founder of Sun Microsystems. KV and GitLab have a long history: They invested in our seed round, led our Series A, and has been a part of every fundraising round since. You could say Khosla Ventures is a big fan of GitLab, and the feeling is mutual. \n\n“I’m not sure what we’re going to get out of this. Khosla is the only investor that can get me to travel an hour and a half to meet in person, every year, without an agenda,” Sid told me. \n\nWhile I sincerely appreciate Sid’s dedication to all-remote, to [efficiency](https://handbook.gitlab.com/handbook/values/#efficiency), and to keeping things simple, I found his sentiment surprising. To me, this was a big deal. We were going to spend the day with some of the industry’s brightest minds in Silicon Valley. However, I equally appreciated his emphasis on the lack of agenda. At GitLab, [we work asynchronously](/handbook/communication/) and agendas allow people to prepare ahead of time so meetings can be interactive, discussion-oriented, and productive. A meeting without an agenda is unpredictable and can be a waste of time. \n\n## Khosla Ventures \n\nPulling up to Khosla Ventures in Menlo Park was a refreshing change of scenery from downtown San Francisco where I’d been staying. The mid-century modern style building is tucked behind lush greenery, and as soon as you enter you are unexpectedly greeted by quirky grape purple walls. Inside Khosla Ventures feels more like a creative incubator than an investor firm. \n\nWe first met with Bruce Armstrong, operating partner at Khosla Ventures and a GitLab board member. Top of mind for everyone was the board. With our plans to go public, we needed to hire new board members with public company experience. Earlier that week, I had the opportunity to sit in on a few conversations with potential board members and now I was getting the chance to hear how they were evaluating the candidates. Across the candidates, a pattern was emerging: They all had experience firing CEOs. For some CEOs this would be worrisome. For Sid, it appeared to be a point of pride: He wanted to hire the best people that were going to make the best decisions for the company in the long run. Period. \n\n## The brainstorm \n\nNext we met with Vinod, Bruce, and investment partners Brian Byun and Sven Strohband for a company brief and brainstorm session. Sid began with an update on the financials, and detailed our massive growth and expansion both in terms of people and product. There were no slides presented. Instead, Sid used our website as a visual aid. Nearly every question or discussion point was first addressed with a Google search to pull up the appropriate GitLab web page to reference. When discussing the progress of the product, he defaulted to our [homepage](https://about.gitlab.com/), where our product team has meticulously detailed our current and future feature set. During a discussion about competitors and potential partners, our [DevOps tools page](/competition/) was referenced for a single page view of all our competitors in context of exactly how we compete. \n\nI’ve been working at GitLab for two and a half years as a content writer on our marketing team and at times have been extremely frustrated with our marketing website—the content on it, how it’s organized, what we’re presenting, etc. It doesn’t look or operate in a way I’m familiar with so my instinct was to not trust it. But nothing we do at GitLab is “normal,” and witnessing our CEO use the website as a single source of public truth to inform our investors is just one example of what it means to be a [transparent company](https://handbook.gitlab.com/handbook/values/#transparency). We don’t hide behind “marketing speak” on our public facing website or develop behind closed walls. We tell the same story and share the same information with the company, the customers, the community, and yes, even the investors. \n\nWhile our website doesn’t have the same flashy graphics and pithy marketing copy I’m used to, it speaks the truth even when the truth makes people a little nervous about how we’re going to pull this off. We have an incredibly ambitious product roadmap to be built by an all remote team in a short amount of time.\n\n### Acquisition strategy\n\nSomething I’ve found surprising throughout my entire CEO Shadow experience is how external people underestimate GitLab’s ability to deliver on ambitious plans. The conversation often defaults to, “I see what you’re trying to do, but realistically, which categories are you really able to compete in?” And, unfalteringly Sid answers: all of them. There are some awkward laughs, and the question is reframed to “What part of the product are most of your customers using today?” We move on.\n\nThe conversation with Vinod Khosla was similar but different in tone. Vinod and the rest of the team were skeptical of our ambition but perhaps more attuned to Sid’s commitment to the direction and vision and thus more willing to dig into how we get there instead of why we won’t. \nPotential partnerships to help fill some of our missing functionality were discussed, but it was apparent that our plan was quickly encroaching into competitive territory among the leading contenders. \n\nInstead, there’s an acquisition strategy. To achieve our goal to deliver a single application for the entire DevOps lifecycle that is best-of-breed in every category, we are going to need some help and make some acquisitions. We already acquired [Gitter](/blog/gitter-acquisition/) and Gemnasium in order to enter into the ChatOps and security space more quickly than if we tried to build it all from scratch.\n\nNaturally, our acquisition strategy and offer was already [drafted and public in our handbook](/handbook/acquisitions/). This enabled the conversation to focus on thinking through potential companies and specific areas of our product where we may want to augment the productivity of our soon-to-be 500 internal developers with an acquisition. \n\n### The IPO date  \n\nGitLab [plans to go public on November 18, 2020](/company/strategy/#sequence) and prefers to remain an independent company with no plans of being acquired. While Vinod made it clear it’s strange to pick and make public an IPO date, at GitLab, we are driven by results and deadlines and even an IPO is no exception to the rule. The route we chose to go—traditional or direct listing—was another topic. \n\nDirect listings are historically uncommon. It wasn’t until [Spotify went public via a direct listing](https://techcrunch.com/2018/02/28/spotify-has-filed-to-go-public/) in 2018 that there was even a precedent for tech companies. Now, Slack and potentially Airbnb are rumored to be next, officially [making direct listings *a thing*](https://www.bloomberg.com/opinion/articles/2019-01-11/direct-listings-are-a-thing-now). As for GitLab, like everything else, it will come down to what’s right for us. I can report in good faith all options are being examined carefully and closely. The takeaway here is that while some might think it’s crazy of GitLab to set this ambitious goal, and Vinod might think it’s crazy to set a specific date, one thing is for sure: As a company, we’re ready and already thinking about what’s next. \n\n### Growth curves \nYou know you’re at a successful company when the VCs aren’t focused on how you’re going to meet your short-term goals or current [product vision](/direction/#vision) but are excited about the long-term vision. \n\n“Where is GitLab five years from now?” Vinod asked the room, as he stood up and drew a chart with three staggered S curves on it. He pointed to the first one, “This is where you are now.” Pointing to the other two he asked, “what comes next?” \n\nHe explained how Square started as a payment device, then to a point-of-sale system, found success, and instead of stagnating, entered into a new market via their Square Capital and Cash App offerings. You see similar trajectory with Facebook entering into the devices space with their newest Portal system. What was it going to be for GitLab? It seemed like an outrageous question to ask considering we still have this huge vision to complete but, unsurprisingly, Sid had thought about this and some things are already in the making. \n\n#### Growth curve #1: Meltano\n\n[Meltano](/blog/hey-data-teams-we-are-working-on-a-tool-just-for-you/) (model, extract, load, transform, analyze, notebook, orchestrate) is a start-up within GitLab aimed at becoming a complete solution for data teams. Similar in concept to GitLab, the goal is to create a single application for the entire data science lifecycle. The mission is similar as well: Make analytics accessible to everyone. If successful, Meltano will begin to bridge the gap between systems and data and bring the GitLab vision of everyone can contribute to even more people. \n\n#### Growth curve #2: Product assisted digital transformation  \n\nThe next idea was a real show stopper: product assisted digital transformation. Think code review as a service but expanded to culture, infrastructure, management, pipelines, process, and integrated directly into the product instead of being an outside service. Imagine if you could bootstrap and up-level your engineering teams’ skills with a product that comes with engineering best practices and support out-of-the-box. \n\n## Safe for another year\n\nAs it turns out, agenda-less meetings at Khosla Ventures can provide a ton of value. We walked out of Khosla’s office with a healthy dose of validation and criticism, and our brains buzzing with new horizons of potential to explore. I was already convinced GitLab is a great company to work for, but my experience at Khosla opened my eyes to just how unique our opportunity is. And, the on-site, half day, agenda-less meeting is good for another year. \n\nCover image by [Reza Rostampisheh](https://unsplash.com/@rezarp) on [Unsplash](https://unsplash.com/photos/-hcCm0kIaSg)\n{: .note}\n",[705,9,706],{"slug":978,"featured":6,"template":685},"khosla-ventures-gitlab-meeting","content:en-us:blog:khosla-ventures-gitlab-meeting.yml","Khosla Ventures Gitlab Meeting","en-us/blog/khosla-ventures-gitlab-meeting.yml","en-us/blog/khosla-ventures-gitlab-meeting",{"_path":984,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":985,"content":991,"config":997,"_id":999,"_type":13,"title":1000,"_source":15,"_file":1001,"_stem":1002,"_extension":18},"/en-us/blog/lessons-on-building-a-distributed-company",{"title":986,"description":987,"ogTitle":986,"ogDescription":987,"noIndex":6,"ogImage":988,"ogUrl":989,"ogSiteName":669,"ogType":670,"canonicalUrls":989,"schema":990},"9 Lessons on building a distributed company","GitLab CEO Sid Sijbrandij and Outklip Founder Sunil Kowlgi talk about remote hiring, management, customer support, and more.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749678641/Blog/Hero%20Images/lessons-building-distributed-company.jpg","https://about.gitlab.com/blog/lessons-on-building-a-distributed-company","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"9 Lessons on building a distributed company\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"Sunil Kowlgi\"}],\n        \"datePublished\": \"2019-04-18\",\n      }",{"title":986,"description":987,"authors":992,"heroImage":988,"date":994,"body":995,"category":955,"tags":996},[993],"Sunil Kowlgi","2019-04-18","\n_GitLab CEO Sid Sijbrandij occasionally sits down for a \"[pick your brain](/handbook/eba/ceo-scheduling/#pick-your-brain-meetings)\"\nmeeting with people seeking advice on open source, remote work, or a discussion of other things related to GitLab._\n\nIt is far easier to run an all-remote company than one that’s a hybrid of remote and colocated,\nsays [Sid Sijbrandij](/company/team/#sytses). When a company adopts a colocated\nculture there’s less recording of things and fewer digital artifacts, so it’s going to be hard for\nthe rest of the company to figure out how decisions are made.\n\nI interviewed Sid for lessons on building a distributed company like GitLab. Sid answered\nquestions on topics ranging from hiring to customer support.\n\nMy top takeaways from the interview:\n\n### 1. Remote interviews are more convenient than in-person interviews\n\nDuring an in-person interview, you need to make sure all your interview materials are loaded\nbeforehand on your laptop or iPad. It’s also going to be hard navigating things on your computer\nwhile talking to a person in front of you. You might write down notes that you’ll need to\ndigitize later by scanning, which is redundant work. On the other hand, when interviewing\nsomeone remotely over a video conference, you have all the materials at hand.\nBecause you’re looking at a screen you can look up information online and quickly take notes without interruption.\n\n### 2. Spend more time on the candidate’s questions than on your questions\n\nDuring interviews, you can get a lot of information about the candidate from the questions\nthey come prepared with and their follow-on questions. When Sid interviews, he spends most of the interview on the candidate’s questions.\n\n### 3. It is really important to write things down\n\nPeople are very efficient at reading things. If you write something down you can refer to it,\nso you don’t have to say everything again. In order to have alignment in a distributed company,\nrepetition of goals and strategy is needed. Repetition is easier when you have one writeup and people are able to easily find it.\n\n### 4. Google Docs is superior to a whiteboard\n\nIt is quite common to have meetings where everyone is looking at the same thing.\nBut, because of time zone differences, it’s hard to involve everyone in a meeting.\nWhile whiteboards are commonly used in in-person meetings, they’re not missed that much by remote workers.\nGoogle Docs is superior to a whiteboard because you never run out of space, you can use\nnumbered lists and indentation, and people can view them afterwards.\n\n### 5. Cross-functional teams don’t work well\n\nGitLab doesn’t do cross-functional teams. Teams are composed of people that perform a similar role.\nA team manager is someone who has experience with that role. This way the manager is able\nto assess results, coach, and give career advice, which is very important.\n\n### 6. Focus on the output of employees, not the input\n\nGood remote workers are focused on results. Especially for managers, it’s important that they\ndon’t focus on the input of people – how long they worked or things like that – but rather focus on the output.\nFocus on the input is not healthy in any company, but especially with remote work you have to let it go.\nNo one’s looking over your shoulder to check whether you’re on Facebook or not, and it’s fine if you\nare as long as you deliver the work to a reasonable degree.\n\n### 7. To be a good manager, you have to quickly identify and remedy underperformance\n\nGitLab hires people who are capable of being [managers of one](https://handbook.gitlab.com/handbook/values/#managers-of-one). But in instances where someone\nis underperforming, managers have to identify it, have a conversation, and take remedial action.\nHere’s [GitLab’s process for dealing with underperformance](/handbook/leadership/underperformance/).\n\n### 8. Be quick with recognition\n\nGitLab has various kinds of employee recognition. For quick recognition, there’s a #thanks\nchannel on Slack where people can celebrate their colleagues’ work. There are also $1,000\ndiscretionary bonuses and GitLab tends to be very high velocity with those.\nRecognizing employees and doing it quickly is really important.\n\n### 9. Put customer-reported issues on a level playing field with internally reported issues\n\nThe issue tracking process in GitLab doesn’t distinguish whether the issue reporter is a user,\n a customer, or a team member. If an issue comes from a user or customer, it’s probably\nbecause they care a lot about what you’re building. So, every feature request, everything\nGitLab team-members work on is out there on a level playing field. GitLab tends to have a lot more\ninteraction with customers than other companies.\n\nWatch the full interview below:\n\n\u003C!-- blank line -->\n\u003Cfigure class=\"video_container\">\n  \u003Ciframe src=\"https://www.youtube.com/embed/pDU8lxh1-6U\" frameborder=\"0\" allowfullscreen=\"true\"> \u003C/iframe>\n\u003C/figure>\n\u003C!-- blank line -->\n\n[Visit this page to read the transcript of the interview](https://outklip.com/blog/gitlab-building-a-distributed-company/).\n\n### About the guest author\n\nSunil Kowlgi is the founder of [Outklip](https://outklip.com), a video platform for remote work.\n\nPhoto by [Brett Zeck](https://unsplash.com/photos/eyfMgGvo9PA?utm_source=unsplash&utm_medium=referral&utm_content=creditCopyText) on [Unsplash](https://unsplash.com/search/photos/globe?utm_source=unsplash&utm_medium=referral&utm_content=creditCopyText)\n{: .note}\n",[705,680,706,9],{"slug":998,"featured":6,"template":685},"lessons-on-building-a-distributed-company","content:en-us:blog:lessons-on-building-a-distributed-company.yml","Lessons On Building A Distributed Company","en-us/blog/lessons-on-building-a-distributed-company.yml","en-us/blog/lessons-on-building-a-distributed-company",{"_path":1004,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":1005,"content":1011,"config":1019,"_id":1021,"_type":13,"title":1022,"_source":15,"_file":1023,"_stem":1024,"_extension":18},"/en-us/blog/optimizing-the-value-exchange-a-gentle-introduction",{"title":1006,"description":1007,"ogTitle":1006,"ogDescription":1007,"noIndex":6,"ogImage":1008,"ogUrl":1009,"ogSiteName":669,"ogType":670,"canonicalUrls":1009,"schema":1010},"How to optimize your value exchange system","Part one of a pragmatic, business-driven guide to help teams transition from fixating on output to optimizing the value exchange with their customers.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749672701/Blog/Hero%20Images/post-1-cover.jpg","https://about.gitlab.com/blog/optimizing-the-value-exchange-a-gentle-introduction","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"How to optimize your value exchange system\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"Gabe Weaver\"}],\n        \"datePublished\": \"2019-12-16\",\n      }",{"title":1006,"description":1007,"authors":1012,"heroImage":1008,"date":1014,"body":1015,"category":789,"tags":1016},[1013],"Gabe Weaver","2019-12-16","*Reading time: 8 minutes, 28 seconds*\n\n## The problem \n\nBuilding software products is **really** hard. Building an enduring company is even harder. We are constantly looking to [hire](https://www.christenseninstitute.org/jobs-to-be-done/) solutions to help us do this quickly and with as little risk as possible. Companies hire Agile ([R.I.P](https://pragdave.me/blog/time-to-kill-agile.html)) because they believe it will help them be more adaptable, increase productivity, and accelerate software delivery. They hire DevOps as a natural companion to increase speed, reduce defects, safely deploy code, and improve the resilience of infrastructures. \n\nWhile these solutions are directionally correct, we have a long way to go:\n\n- Software failures cost the U.S. economy over [$1t annually](https://www.cloudcomputing-news.net/news/2017/oct/30/glitch-economy-counting-cost-software-failures/).\n- Only 56% of startups [make it to their fourth year](https://smallbiztrends.com/2019/03/startup-statistics-small-business.html).\n- [~40%](https://newproductsuccess.org/new-product-failure-rates-2013-jpim-30-pp-976-979/) of new products fail.\n- Only [6% of teams](https://www.stateofagile.com/#ufh-i-521251909-13th-annual-state-of-agile-report/473508) report that Agile (long live agility) practices are enabling greater adaptability to changing market conditions.\n\nTo get better, we need to fundamentally shift the way we approach building software products and companies. Through a fictitious story about a company called Acme Co., I'm going to provide a pragmatic, business-driven approach of how we can get there. The first step is to change how we define success. We need to move from **output to outcomes**. \n\n## The basics of the value exchange system\n\nI've worked with a lot of companies over the years. Most of them measure the success of product teams by how much they ship. In [Escaping The Build Trap](https://melissaperri.com/book), Melissa Perri correctly identifies the problem with this and the root cause -- companies misunderstand value. As she succinctly describes in her book:\n\n> Instead of associating value with the outcomes they want to create for their businesses and customers, they measure value by the number of things they produce. Let’s go back to the basics to determine what true value is. Fundamentally, companies operate on a value exchange. \n\nShe uses a simple diagram to illustrate the value exchange system:\n\n\u003Cbr>\n![A Simplified Value Exchange System Diagram](https://about.gitlab.com/images/blogimages/optimizing-the-value-exchange/value-exchange-simple-diagram.png)\n\n\u003Cbr>\nCompanies build products to serve as vehicles for value delivery. When the customer's problems, needs, and wants are fulfilled, they provide value back to the business. Business value is easy to define, as it typically maps to achieving traditional objectives that are universal to all companies:\n\n- **Sustainable Value:** Support the product's core value and create barriers to competition.\n- **Growth:** Grow market share, fulfill more demand, develop new markets, and improve recurring revenue.\n- **Profit:** Support higher prices, improve lifetime value, lower costs, and leverage existing assets.\n\nWhile these are easily measurable, customer value is often more intangible; making it difficult to define and correlate to business objectives. This leads companies to create proxies to represent value that are more straightforward to comprehend and measure -- such as things shipped.\n\nAs we discussed earlier, companies hire solutions like Agile (long live agility) and DevOps because they want to increase the speed and productivity of their value delivery system. The Value Exchange System is a [reinforcing loop](https://thesystemsthinker.com/reinforcing-and-balancing-loops-building-blocks-of-dynamic-systems/), so it naturally follows that if you increase the speed at which you deliver value, you will, therefore, increase the amount of value you capture from your customers. Companies are so fixated on optimizing for speed that a whole market of productivity analytics is emerging to track and report it. For example, the hook on the landing page of a prominent tool on the market promises to help you \"measure your team's success\" by surfacing velocity data to see how fast your team is going. \n\nWhile speed and productivity are good things, optimizing for them in a silo will have a minimal overall impact on increasing the positive reinforcement to the system. This is because output does not necessarily result in outcomes. We need to optimize both how we deliver _and_ capture value. To better understand why, let's jump into an all so common dilemma Acme is currently trying to overcome.\n\n## Let's Meet Acme Co.\n\n> “Bounded rationality means that people make quite reasonable decisions based on the information they have. But they don’t have perfect information, especially about more distant parts of the system.” -- [Thinking In Systems](https://www.amazon.com/Thinking-Systems-Donella-H-Meadows/dp/1603580557)\n\nAcme is a tech company that provides IoT devices and AI-driven insights to help logistics companies improve the efficiency of their operations. Acme has seen remarkable success over the last several years by leveraging a common growth strategy of steadily increasing investment in R&D, marketing, and selling to capture market share as quickly as possible. While not yet profitable, they believe there is an imminent inflection point where their investment in R&D will begin to pay off, resulting in revenue growth disproportionate to operating expenses. This has been playing out as expected, but the executive team is starting to get concerned given some financial trends over the last few quarters. \n\n| | Q1 | Q2 | Q3 | Q4 |\n| --- | --- | --- | --- | --- |\n| Revenue | 36.7 | 44.0 | 48.4 | 50.9 |\n| COGS | 2.2 | 4.0 | 4.8 | 4.6 |\n| **Net Sales** | **34.5** | **40.1** | **43.6** | **46.3** |\n| *Margin* | *94%* | *91%* | *90%* | *91%* |\n|  |  |  |  |  |\n| R&D | 16.5 | 18.3 | 20.3 | 22.7 |\n| Marketing | 4.1 | 5.2 | 6.6 | 7.4 |\n| Sales | 15.0 | 17.8 | 19.9 | 21.2 |\n| G&A | 1.6 | 2.1 | 2.3 | 2.9 |\n| **Total OpEx** | **37.2** | **43.4** | **49.1** | **54.2** |\n| |  |  |  |  |\n| **Net Profit** | **-2.7** | **-3.3** | **-5.5** | **-7.9** |\n| *Profit Margin* | *-7%* | *-8%* | *-11%* | *-16%* |\n\nAcme's board set a -20% acceptable risk threshold for the net profit margin. It believes, that if necessary, it will be able to quickly reduce the deficit through strategic cost-cutting measures. It's clear from the financials that the company is on course to hit the threshold and the executives are scrambling to understand why. They have been easily converting customers from their competitors and are nowhere close to saturating their addressable market. They are investing heavily in delivering new product capabilities and scaling the sales organization to capture the value in return. The executive team decides to create a dedicated cross-functional working group to investigate and solve the problem. \n\n## Practicing kaizen with the improvement kata\n\nThe challenge before the working group is daunting. Acme is an incredibly complex system with hundreds of people and moving parts. They decide to adopt a management processes from the Toyota Product System developed by Taiichi Ohno. The group reviews what it means to practice continuous improvement - [kaizen](https://en.wikipedia.org/wiki/Kaizen) - by employing a technique called the Improvement Kata.\n\n![Diagram of the Improvement Kata](https://about.gitlab.com/images/blogimages/optimizing-the-value-exchange/toyota-kata.png){: .center}\n\nThe first thing the group needs to decide is which direction to focus on first. If the challenge is to increase profits, that could be accomplished in a few different ways - increase the growth rate of revenue or decrease operating expenses. Looking at the financials, revenue growth has declined from an average of 20% per quarter to only 5% in the most recent. Given Acme has a sizeable market left to capture, and decreasing R&D spend at this point could hurt their long term growth targets, the group sets a target condition of getting revenue growth back to 10%. This feels like an obtainable goal within the quarter and will demonstrate forward progress to the executives.\n\nWith a target condition in hand, their next task is to identify the root cause of the decline in revenue growth and conduct experiments to reach their goal. They set out to connect with the sales team to begin their investigation. \n\n##  Why is revenue growth declining?\n\nAs the working group sits down with the sales team to review numbers, they use the [5 why's](https://medium.com/productmanagement101/learn-about-the-five-whys-technique-78283d75800f) to try to understand where things are going wrong:\n\n- **Why is revenue growth declining?** The sales team shares that conversion rates from new revenue are holding steady, but their conversion rates from up-sells have fallen dramatically.\n- **Why are up-sells declining?** According to the sales team, Acme is not delivering capabilities fast enough that it had promised during the sales cycle.\n- **Why did the sales team set expectations for capabilities that weren't built yet?** The sales team explained that enterprise customers have complex needs that aren't supported in Acme's core product capabilities yet. To keep up with sales quotas, they found that walking prospects through Acme's product roadmap usually gets the deal over the line. Based on the roadmap and the planned release dates, they had not seen this as a risk because things were expected to be delivered within the customer's acceptable time ranges. \n- **Why weren't things delivered within the customer's acceptable time range?** In further discussions, the sales team reveals that they feel R&D is delivering new features at an increasingly slower rate quarter over quarter.\n- **Why is R&D slowing down on value delivery?** At first glance, R&D's output metrics are consistent month-over-month relative to headcount. While the sales team helped the working group see that lost opportunities in up-sells were driving lower growth rates; it didn't make sense. The group decides to head over to R&D to better understand the value delivery system.\n\n## The importance of measuring the value delivery stream\n\nTo prepare for collaborating with the R&D team, the working group spends a few minutes reviewing the team's productivity metrics. To the surprise of the group, R&D only has one primary metric everyone on the team tracks consistently - the count of work items delivered. As the working group starts discussions with R&D, they explain that the executives set output objectives as a means of measuring the team's success. The rationale for this was due to the way the product is bundled and priced. The categories all roll up into one pricing model and it was difficult for the finance team to figure out a way to attribute revenue to the various categories. They had ultimately decided that measuring productivity was the next best thing.    \n\nSo if work item output is the goal and R&D's output has been consistent quarter over quarter, why was the sales team convinced R&D was slowing down on value delivery? To understand this, the working group needed to better understand the flow of items through the value delivery process within the Value Exchange System. To do this, they generated a simplified value stream map to visualize the stages a work item goes through as it is converted from requirements to a production feature.  \n\n\u003Cbr>\n\n![Acme's Value Stream Map](https://about.gitlab.com/images/blogimages/optimizing-the-value-exchange/d1.png){: .center}\n\n_Acme's Value Stream Map. Cycle Time = Time in queue + active time + time waiting once started_\n\nThe value stream map reveals that the total lead time for a work item is 2,132 hours (88 days). This is an astonishing revelation to the working group; especially since they didn't include the time a work item spends waiting between when a customer requests a feature and the team starts the planning process. Even though the R&D team delivered ~3,100 work items last month, it took well over three months to satisfy customer requests. The working group needs to collect more data, but they know they are on the right track. Before synthesizing a hypothesis, they finish collecting additional historical metrics to confirm their suspicions. \n\n| | Q1 | Q2 | Q3 | Q4 |\n| --- | --- | --- | --- | --- |\n| Items Delivered | 2,275 | 2,524 | 2,800 | 3,131 |\n| Lead Time | 31 | 52 | 69 | 88 |\n| Cost Per Item Delivered | 7,250 | 7,250 | 7,250 | 7,250 |\n\nBased on the data, the working group notices the strong correlation between the increase in lead time and the decrease in up-sell revenue growth. Given their target condition for the upcoming quarter, they create the following hypothesis: \n\n> Decreasing the lead time by ~35% will enable customer requests to be completed 18 days earlier, resulting in a 5% increase in revenue growth\n\nWith a falsifiable hypothesis in hand, they shift their attention to figuring out the best approach for running an experiment. \n\nContinue reading:\n\n- Part 1: A Gentle(ish) Introduction\n- Part 2 (Next): [Reduce Waste To Increase Flow](/blog/optimizing-the-value-exchange-reduce-waste-to-increase-flow/\n)\n- Part 3: [The Compounding Value Of Shorter Feedback Loops](/blog/optimizing-the-value-exchange-the-compounding-value-of-shorter-feedback-loops/)\n\n ***Giving credit where it is due:** In [Escaping The Build Trap](https://www.oreilly.com/library/view/escaping-the-build/9781491973783/ch01.html), Melissa Perri discusses the Value Exchange System at length and provides unparalleled wisdom for modern-day product managers. The Improvement Kata, Kanban, and Toyota Production System would not exist today if it weren't for Taiichi Ohno. His work has been foundational to many other systems and processes that have evolved over the years.* \n\nCover Photo by Cristina Gottardi on [Unsplash](https://unsplash.com/photos/boxam4k4rQw)\n",[1017,9,1018],"agile","workflow",{"slug":1020,"featured":6,"template":685},"optimizing-the-value-exchange-a-gentle-introduction","content:en-us:blog:optimizing-the-value-exchange-a-gentle-introduction.yml","Optimizing The Value Exchange A Gentle Introduction","en-us/blog/optimizing-the-value-exchange-a-gentle-introduction.yml","en-us/blog/optimizing-the-value-exchange-a-gentle-introduction",{"_path":1026,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":1027,"content":1033,"config":1037,"_id":1039,"_type":13,"title":1040,"_source":15,"_file":1041,"_stem":1042,"_extension":18},"/en-us/blog/optimizing-the-value-exchange-reduce-waste-to-increase-flow",{"title":1028,"description":1029,"ogTitle":1028,"ogDescription":1029,"noIndex":6,"ogImage":1030,"ogUrl":1031,"ogSiteName":669,"ogType":670,"canonicalUrls":1031,"schema":1032},"Optimizing The Value Exchange: Reduce Waste To Increase Flow","Part two of a pragmatic, business-driven guide to help teams transition from fixating on output to optimizing the value exchange with their customers.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749672713/Blog/Hero%20Images/post-2-cover.jpg","https://about.gitlab.com/blog/optimizing-the-value-exchange-reduce-waste-to-increase-flow","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"Optimizing The Value Exchange: Reduce Waste To Increase Flow\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"Gabe Weaver\"}],\n        \"datePublished\": \"2019-12-16\",\n      }",{"title":1028,"description":1029,"authors":1034,"heroImage":1030,"date":1014,"body":1035,"category":789,"tags":1036},[1013],"If you're new to the series, checkout Part 1: [A Gentle(ish) Introduction](/blog/optimizing-the-value-exchange-a-gentle-introduction/\n)\n\n_Reading time: 13 minutes, 32 seconds_\n\n## Learning To Identify Waste\n\n> “Systems behavior is particularly sensitive to the goals of feedback loops. If the goals-the indicators of satisfaction of the rules - are defined inaccurately or incompletely, the system may obediently work to produce a result that is not really intended or wanted.” -- [Thinking In Systems](https://www.amazon.com/Thinking-Systems-Donella-H-Meadows/dp/1603580557)\n\nSince following the Improvement Kata has gotten the group this far, they decide to use another concept from the Toyota Production System (TPS) -- waste reduction. To increase flow through the value delivery system, they start by looking at steps in the workflow that do not add any intrinsic value. TPS categorizes waste into 8 distinct types:\n\n- **Talent:** Underutilizing people's talents, skills, and knowledge.\n- **Inventory:** Excess products and materials not being processed.\n- **Motion:** Unnecessary movements by people.\n- **Waiting:** Wasted time waiting for the next step in a process. \n- **Transportation:** Unnecessary movements of products and materials.\n- **Defects:** Efforts caused by rework, scrap, and incorrect information.\n- **Overproduction:** Production that is more than needed or before it is needed.\n- **Overprocessing:** More work or higher quality than is required by the customer. \n\nWith this in mind, the working group pulls up their value stream map and immediately identifies a common theme across stages in the flow - some stages have a high amount of time spent waiting. The group quickly adds this up to understand the impact.\n\n![Acme's Value Stream Map - Time waiting](https://about.gitlab.com/images/blogimages/optimizing-the-value-exchange/d2.png){: .center}\n\nIn total, work items spend ~85 days waiting between queues or waiting for additional work needed from other parts of the R&D team. The group is speechless as they realize that it only takes around 3 days of actual labor to move a work item through the value delivery stream. They decide this is the right place to start and prioritize their efforts based on which cycles have the highest amount of time in waiting:\n\n1. **Build** - 863 hours in a queue, 613 hours once a work item has started.\n2. **Plan** - 362 hours once a work item has started.\n3. **Review** - 27 hours in a queue, 95 hours once a work item has started.\n4. **Canary** - 73 hours in a queue.\n5. **Prod** - 22 hours in a queue.\n6. **Staging** - 1 hour in a queue.\n\nThe group is excited about their discovery and is eager to share it with the engineering team. When they sit down and show them the numbers, the engineers are surprised. The conversation continues with the engineers talking through what it looks like for a single engineer in a typical week. The working group captures it in a diagram and confirms with the engineers that it is representative of the conversation.\n\n![Sequence Diagram](https://about.gitlab.com/images/blogimages/optimizing-the-value-exchange/sequence-diagram.png){: .center}\n\nThe group asks the engineers how they feel about this workflow. They're generally supportive as they don't see an alternative to this despite it being extremely heavy on the context switching. The working group still needs to figure out how to reduce the waiting time for work items in this stage and eliminated unneccessary waiting is a logical place to start. They present their target condition and ask the engineers for suggestions. The engineers feel strongly that the code review process is imperative to maintain code quality. One engineer suggests the team could try pair programming. He heard that it can increase code quality. It would also remove the Review stage entirely as two people contribute to the same work item until it is fully completed.\n\nThey debated back and forth. Many felt that it would slow everything down even more because they would only be able to do half the work they had been doing up to that point. The engineer who brought it up in the first place pointed to an a [experiment](https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20030012934.pdf) where NASA engineers completed two like-size projects - one as they normally would and one strictly following [Extreme Programming](http://www.extremeprogramming.org/); which includes a mix of practices like pair programming, collective code ownership, TDD, and small iterations. He presented the team with the following data:\n\n| | Non-XP Projet | XP Project |\n| --- | --- | --- |\n| LOC/Hour | 24 | 27 |\n| Util/test LOC | 0 | 1,135 |\n| Prod LOC | 2,144 | 912 |\n| Total LOC | 2,144 | 2,545 |\n\nHe explained that part of the reason why the team was able to produce less production code at a faster overall rate was due to the pair pushing each other not to cut corners by focusing on quality; ruthlessly refactoring code as they went. He also discussed the concept of the driver and navigator roles, and how that led to better architecture decisions the first time around. The engineers were skeptical but were willing to give it a try. Together, everyone agreed that the engineers would run an experiment to try pair programming for the upcoming quarter. They also agreed to only work on one item from start to merged before picking up the next item.\n\n## Increased flow does not equal increased value\n\nAt the end of the quarter, the working group got together to review an updated value stream map. \n\n![Acme's Value Stream Map - Updated](https://about.gitlab.com/images/blogimages/optimizing-the-value-exchange/d3.png){: .center}\n\nThe work in progress limits and pair programming dramatically improved flow through the value stream. Lead time dropped to 893 hours or 37 days. A 57% reduction was much better than the working group had anticipated. The books for the quarter weren't closed, but the working group knew it was going to be a huge win. They eagerly shared the results with the executive team.\n\nAs the quarterly financials were released, the group got a call from the CFO. He was disappointed. He had expected to not only achieve the 5% incremental revenue growth but do much better given the results of the experiment. The working group was stunned. Revenue growth only increased from 5% to 9%. Despite missing the mark, the executives agreed to continue sponsoring the working group as it did produce tangible results and business outcomes. \n\n**Financials**\n\n| | Previous | Current | \n| --- | --- | --- |\n| Revenue | 50.9 | 55.4 |\n| COGS | 4.6 | 5.0 |\n| **Net Sales** | **46.3** | **50.1** |\n| *Margin* | *91%* | *91%* |\n|  |  |  |\n| R&D | 22.7 | 23.1 | \n| Marketing | 7.4 | 8 |\n| Sales | 21.2 | 22.9 |\n| G&A | 2.9 | 3.8 |\n| **Total OpEx** | **54.2** | **57.8** |\n| |  |  |\n| **Net Profit** | **-7.9** | **-7.3** |\n| *Profit Margin* | *-16%* | *-13%* |\n\n\u003Cbr>\n**Value Delivery Economics**\n\n| | Previous | Current |\n| --- | --- | --- | \n| Items Delivered | 3,131 | 5,002 |\n| Lead Time | 88 | 37 |\n| Cost Per Item Delivered | 7,250 | 4,617 |\n\nThey combed through the financials and productivity metrics to figure out what had happened. The value delivery system had become much more efficient. The cost per item delivered decreased. What went wrong? They circled back to where they started with the sales team to ask for an update. The sales team reported that they still felt like speed on delivering customer requests was moving slowly. It had gotten a little bit better in a few areas, but overall it was having a nominal impact on fixing the problem of declining up-sell rates. Determined to get to the bottom of things, the working group decides to dig a bit deeper into the kind of work flowing through the R&D value stream. \n\nThis is what they discovered upon further investigation into the breakdown on types of work items R&D was delivering:\n\n| | Q2 | Q3 | Q4 | Q1 |\n| --- | --- | --- | --- |\n| Feature | 63% | 46% | 38% | 44% |\n| Defect | 13% | 22% | 29% | 36% |\n| Technical Debt | 24% | 32% | 33% | 20% |\n\nA light bulb clicked. No wonder the sales team felt like things were moving slowly. The amount of new value being delivered to customers was near an all-time low. It was certainly better than the previous quarter thanks to the improvement in lead time, but this trend was concerning to the group. \n\nAs they followed the same path from the first time around, they connected with the R&D team to ask questions. After presenting the work item breakdown by type, R&D was not surprised at all. They look at this ratio quite often. When the group asked them why they hadn't surfaced this earlier, they explained that the pace of constant feature prioritization over the last several years gave the team little extra time to refactor parts of the codebase that were critical to the value delivery system. \n\nIn continuing to listen intently, the team conveyed that they had brought this up on several occasions. Upper management insisted that the team focus on creating new value to fuel the growth targets set by the company leaders. The team also explained that because the company has SLAs with customers, they have had to prioritize severe defects over new features; leaving little time for anything else. \n\nA product manager stepped forward and openly confessed that he regularly makes pacts with his engineering manager to repay technical debt in short order if they would do whatever it took to get the feature shipped. He said they rarely followed through to go back and clean up technical debt. Everyone in the meeting confirmed that this was an acceptable practice. The working group started to see the long term consequences that are so easily masked by strong financial growth. As the R&D team cut corners to ship features, technical debt mounted. Unable to take the time to address it, defect count had started increasing at a consistent rate. The R&D team now had to split focus across three different areas - trying to keep the lights on, delivering new value, and trying to find time to repay the technical debt. \n\n![Technical Debt Diagram](https://about.gitlab.com/images/blogimages/optimizing-the-value-exchange/tech-debt.jpg){: .center}\n\nDespite this revelation, the working group was still responsible for delivering on their target condition. Now it was on them to convince leadership to take technical debt seriously. As they mulled over ways to present this, they asked for a report of all the inventory that was waiting to be produced. For this to resonate with the executives, they needed to communicate it in a way that provided justifiable business reasons to throttle back on feature development. They calculated the cost per quarter that the company was spending on work that did not add intrinsic customer value or positively reinforce the Value Exchange System.\n\n| | Type | Quarterly Cost |\n| --- | --- | --- |\n| Features | 38% | 8.8 | \n| Defects | 22% | 5.1 | \n| Tech Debt | 40% | 9.2 | \n\nThe company was investing $14.3 million in activities that did not have any tangible impact on driving revenue growth. This was something the leadership team could not ignore. As they presented their case, the executives were concerned that slowing down on feature development in any way would adversely impact revenue growth. The working group talked around this by helping them understand that slowing down for a little bit right now would enable them to accelerate much faster as they would be able to more than double their investment in value-adding features once defects and technical debt were in check. The leadership agreed to the working group's proposal. \n\nOver the next few quarters, R&D invested heavily in paying down technical debt and reducing defects. They also made optimizations to the value stream and were able to replace both the staging and canary environments by deploying directly to production behind feature flags. Product managers were able to test things out in production and as soon as a feature was ready for release, they clicked a button and it would become incrementally available to everyone. New defects were at an all-time low and the team was confident all of the efforts would produce huge dividends. \n\nThe working group double-checked the value stream in anticipation of reviewing the quarterly financials that contained the first full quarter where the team spent 80% of their time on value-adding features. \n\n![Acme's Value Stream Map - 80% features%](https://about.gitlab.com/images/blogimages/optimizing-the-value-exchange/d4.png){: .center}\n\n**Value Delivery Economics**\n\n| | W/ Tech Debt | 80% Features |\n| --- | --- | --- | \n| Items Delivered | 5,002 | 11,467 |\n| Lead Time | 37 | 11.9 |\n| Cost Per Item Delivered | 4,617 | 2,014 |\n\nThey were pleased with their value stream. It was fine-tuned, waste-free, and incredibly efficient.  Leads times were hovering around 286 hours (11.9 days). The company had continued to invest in scaling the R&D team and they were able to churn out an astonishing 11,467 work items. But then came the financials. \n\n| | W/ Tech Debt | Paying It Down | 80% Features | \n| --- | --- | --- | --- |\n| Revenue | 55.4 | 57.7 | 68.04 |\n| COGS | 5.0 | 5.2 | 6.1 |\n| **Net Sales** | **50.5** | **52.5** | **61.9** |\n| *Margin* | *91%* | *91%* | *91%* |\n|  |  |  | |\n| R&D | 23.1 | 25.3 | 27.2 | \n| Marketing | 8 | 8.9 | 9.6 |\n| Sales | 22.9 | 24 | 25.6 |\n| G&A | 3.8 | 5.1 | 6.4 |\n| **Total OpEx** | **57.8** | **63.3** | **68.8** |\n| |  |  | |\n| **Net Profit** | **-7.3** | **-10.8** | **-6.9** |\n| *Profit Margin* | *-13%* | *-19%* | *-10%* |\n\nThey only improved revenue growth rates by 9 points, from 9% to 18%. The working group was surprised the CFO hadn't called. When they checked their calendars, a week-long onsite meeting had been scheduled with heads from every division at the company. It wasn't all too surprising. They continued to fail to meet the mark, but could not figure out why. The cost per work item had been reduced by 72%. Investment in value-adding work items was at an unprecedented all-time high. The value delivery stream was a well-oiled machine, but growth rates weren't responding.\n\nAs the working group prepared to meet their fate, they decided to conduct a retrospective to try to gain an understanding of where they went wrong. During the conversation, one of the group members reminded the team of the earlier conversation with R&D discussing success metrics; where R&D recounted that Acme had never really figured out how to more formally tie business outcomes to the output they were creating. The group didn't know the answer, but they agreed to spend some time before the on-site to explore the topic.\n\nThe group talked to peers at other companies and read as many books as they could get their hands on. One of them was [The Goal](https://www.amazon.com/Goal-Process-Ongoing-Improvement/dp/0884271951/ref=asc_df_0884271951/?tag=hyprod-20&linkCode=df0&hvadid=312175933381&hvpos=1o1&hvnetw=g&hvrand=10925659369993672127&hvpone=&hvptwo=&hvqmt=&hvdev=c&hvdvcmdl=&hvlocint=&hvlocphy=9031326&hvtargid=pla-434522630098&psc=1&tag=&ref=&adgrpid=60258871817&hvpone=&hvptwo=&hvadid=312175933381&hvpos=1o1&hvnetw=g&hvrand=10925659369993672127&hvqmt=&hvdev=c&hvdvcmdl=&hvlocint=&hvlocphy=9031326&hvtargid=pla-434522630098) by Eliyahu Goldratt, a story much similar to the journey that the working group had been on. As the group discussed the book, they finally realized the fatal flaw they had made -- they had spent all of their time fixated on trying to achieve the local maximum of R&D instead of aligning the feedback loops between the value delivery system and the value capture system. \n\nThe most effective way to do this is by focusing on the efficiency of the Value Exchange System as a whole. As the group learned, this can be accomplished by measuring the Value Exchange System's throughput. It's important to note that this kind of throughput should not be confused with the definition in Kanban (work output), but rather the rate at which a system generates revenue through sales. The group had achieved a new level of alignment...almost. To start measuring this, they first needed to learn more about throughput accounting. \n\n## Throughput Accounting 101\n\n> “The most marvelous characteristic of some complex systems is their ability to learn, diversify, complexify, and evolve…Like resilience, self-organization is often sacrificed for purposes of short-term productivity and stability. Productivity and stability are the usual excuses for turning creative human beings into mechanical adjuncts to production processes.” -- [Thinking In Systems](https://www.amazon.com/Thinking-Systems-Donella-H-Meadows/dp/1603580557)\n\nIn stark contrast to cost-based accounting, [throughput accounting](https://en.wikipedia.org/wiki/Throughput_accounting) prioritizes value creation over cost-cutting. It consists of [three primary metrics](https://www.amazon.com/Scaling-Lean-Mastering-Metrics-Startup/dp/1101980524):\n\n- **Throughput** is the rate at which monetizable value is generated from a company's customers over their lifetime minus any variable costs such as the cost of customer acquisition. \n- **Inventory** represents all of the money a company has invested into the Value Exchange System including current product capabilities, unfinished goods (requirements, features, job stories, etc.), equipment, and infrastructure. \n- **Operating Expenses** are the costs expended turning inventory into throughput. This includes research and development, sales, and general expenses. \n\nThese metrics are then built up to determine the financial health of a given Value Exchange System:\n\n- **Throughput (T)** = Lifetime Value (LTV) - Cost of Customer Acquisition (COCA)\n- **Profit (P)** = Total Throughput (T) - Operating Expenses (OE)\n- **Return On Investment (ROI)** = Profit (P) / Inventory (I)\n\nHere is where the differences between cost-based accounting and throughput accounting start to manifest. As Ash Maurya, points out in Scaling Lean:\n\n> Cost-based accounting places more emphasis on the right-hand side of the profit equation — decrease operating expenses. It focuses on scalable efficiency and squeezing out costs — especially labor costs. This typically manifests itself as policies requiring detailed weekly timesheets broken down by task, as well as downsizing, outsourcing, and other cost-reducing measures.\n\n> It is much more powerful to try to affect the left-hand side of the profit equation — increase throughput — because cost-cutting has a theoretical limit of zero. Increasing throughput has no theoretical upper limit. You can find ways to add more value to an existing product, build more add-on products, or expand the market — provided, of course, that these efforts lead to a positive return on investment.\n\nWhat's also important about looking at the Value Exchange System through the lens of throughput accounting is that it provides the preferred order for where to focus optimization efforts:\n\n1. Throughput\n1. Decreasing inventory\n1. Reducing operating expenses\n\n## The Universal Goal\n\nAs Eliyahu Goldratt, creator of throughput accounting and author of [The Goal](https://en.wikipedia.org/wiki/The_Goal_(novel)), frames it:\n\n> The universal goal of every business is to increase throughput while minimizing inventory and operating expenses provided doing that doesn't degrade throughput\n\nThe working group was starting to think differently. Throughput accounting provides a simple, but powerful decision support framework. Every target condition should map to one of three desired outcomes in the Value Exchange System - increase throughput (T), reduce investment (I), or reduce operating expenses (OE). It's also important to point out that throughput accounting does not use the balance sheet to [stash deferred expenses](https://www.scienceofbusiness.com/throughput-accounting/) to be realized later. This is a practice that many companies use to manipulate the income statement depending on the kind of story they want to tell. This ultimately masks the true health of the Value Exchange System. \n\nWhile excited about what they are learning, the working group was still a bit stumped about how to implement it within Acme in a way that would help drive improved revenue growth and profitability. They weren't able to formalize a proposal in time for the onsite, but they were now equipped with some better data points around why they had failed and planned to share the news with the executive group.\n\nContinue reading:\n\n- Part 1: [A Gentle(ish) Introduction](/blog/optimizing-the-value-exchange-a-gentle-introduction/\n)\n- Part 2 : Reduce Waste To Increase Flow\n- Part 3 (Next): [The Compounding Value Of Shorter Feedback Loops](/blog/optimizing-the-value-exchange-the-compounding-value-of-shorter-feedback-loops/)\n\n***Giving credit where it is due:** Throughput Accounting and the Theory Of Constraints was originally pioneered by Eliyahu Goldratt. Introduced in his book, [The Goal](https://en.wikipedia.org/wiki/The_Goal_(novel)), these concepts have revolutionized the way businesses operate.*",[1017,9,1018],{"slug":1038,"featured":6,"template":685},"optimizing-the-value-exchange-reduce-waste-to-increase-flow","content:en-us:blog:optimizing-the-value-exchange-reduce-waste-to-increase-flow.yml","Optimizing The Value Exchange Reduce Waste To Increase Flow","en-us/blog/optimizing-the-value-exchange-reduce-waste-to-increase-flow.yml","en-us/blog/optimizing-the-value-exchange-reduce-waste-to-increase-flow",{"_path":1044,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":1045,"content":1051,"config":1056,"_id":1058,"_type":13,"title":1059,"_source":15,"_file":1060,"_stem":1061,"_extension":18},"/en-us/blog/optimizing-the-value-exchange-the-compounding-value-of-shorter-feedback-loops",{"title":1046,"description":1047,"ogTitle":1046,"ogDescription":1047,"noIndex":6,"ogImage":1048,"ogUrl":1049,"ogSiteName":669,"ogType":670,"canonicalUrls":1049,"schema":1050},"Shorter feedback loops compound value in development","Part three of a pragmatic, business-driven guide to help teams transition from fixating on output to optimizing the value exchange with their customers.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749672726/Blog/Hero%20Images/post-3-cover.jpg","https://about.gitlab.com/blog/optimizing-the-value-exchange-the-compounding-value-of-shorter-feedback-loops","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"Optimizing The Value Exchange: The Compounding Value Of Shorter Feedback Loops\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"Gabe Weaver\"}],\n        \"datePublished\": \"2019-12-16\",\n      }",{"title":1052,"description":1047,"authors":1053,"heroImage":1048,"date":1014,"body":1054,"category":789,"tags":1055},"Optimizing The Value Exchange: The Compounding Value Of Shorter Feedback Loops",[1013],"If you're new to the series, checkout Part 1: [A Gentle(ish) Introduction](/blog/optimizing-the-value-exchange-a-gentle-introduction/\n)\n\n*Reading time: 12 minutes, 8 seconds*\n\n## The Problem Of The Tree Swing\n\n> “The most powerful ways to influence the behavior of a system is through its purpose or goal. That’s because the goal is the direction-setter of the system, the definer of discrepancies that require action, the indicator of compliance, failure, or success toward which balancing feedback loops work.” -- [Thinking In Systems](https://www.amazon.com/Thinking-Systems-Donella-H-Meadows/dp/1603580557)\n\nThe working group sat down with the leadership and department heads; things were tense. The CEO announced some major news -- the board planned to start cutting R&D budget in an attempt to drive profitability. A hiring freeze would be put into effect if profitability wasn't achieved within the next fiscal year. The board would also be considering a massive round of layoffs as an extreme measure. Everyone in the room understood the stark reality of the situation. There was a fundamental problem in how the company was operating.  \n\nAs discussions progressed, the team reviewed the dozens of major initiatives that had shipped over the last several quarters. As they went through these one-by-one, it was not immediately clear how they impacted business outcomes. Sales said that they were all the things customers requested. Product said they talked to the customers and had a validation process for work items during the planning stage. Someone from the working group asked a focusing question -- how does Acme understand how a product capability is used once it has been delivered to a customer? The roadmap was supposedly organized by needs that customers had expressed, but nothing was linking them to quantifiable changes in customer behavior that would result in specific business outcomes.\n\nProduct said they measured certain kinds of activities, but it didn't help them get a clear sense of whether something was working or not. Sales said that account managers regularly checked in with customers and kept track of the feedback. Customer support mentioned they share tickets with product from time-to-time. To grasp where the root issues were in the product, the team decided to call customers that requested features but hadn't fully adopted them. After dozens of interviews and ethnographic observations, everyone started to see a common pattern emerge.\n\n![Insert Tree Diagram](https://about.gitlab.com/images/blogimages/optimizing-the-value-exchange/tree-swing.jpeg){: .center}\n\nAcme wasn't building what customers needed. How had this happened? Each team interpreted things differently; each with a slight bias based on their functional area. They were never fully aligned. Even worse, they didn't understand what their customers needed to move their own businesses forward. They just built what they thought customers wanted. This is the essence of the tree swing problem. \n\nThe working group shared their revelation about optimizing for the wrong thing within R&D and suggested reviewing how success was being measured across the company. Each department pulled up their objectives and key results. Nearly all of them were aimed at each department seeking local maximum. Sales had their quotas to meet. R&D had specific output targets. Customer support was focused on triaging and closing tickets as quickly as possible. The company had grown rapidly, but that just amplified the underlying problems. The signal had been completely lost in the noise.\n\nAs the systemic problems became clear, the working group presented the concept of throughput accounting and how it could help align the organization around a single measure of success. The other team members were skeptical. Nothing else was working and the worst-case scenario was already upon them. Acme finally agreed to give it a try. They spent the next several days discussing what the transition would look like and the mechanics of how it could be measured. They also tackled ways to create tighter alignment across functional areas of the company. Together, they came up with a list of core things they would implement - all to shift the focus from measuring output to measuring outcomes.\n\n## Adopt throughput accounting over cost-based accounting\n\nThe first thing they tackled was converting their traditional cost-based accounting model to the throughput accounting model. They still maintained the traditional models for public reporting, but internally, they would make decisions based upon ROI as defined in throughput accounting.\n\nAcme's original cost-based approach for the previous quarter:\n\n| Income Statement | Previous Period |\n| --- | --- |\n| Revenue | 68.0 |\n| COGS | 6.1 |\n| **Net Sales** | **61.9** |\n| _Margin_ | _91%_ |\n|  |  |\n| R&D | 27.2 |\n| Marketing | 9.6 | \n| Sales | 25.6 | \n| G&A | 6.4 |\n| **Total OpEx** | **68.8** |\n| |  |\n| **Net Profit** | **-6.9** |\n| _Profit Margin_ | _-10%_ |\n\nAcme's original balance sheet for the previous quarter (simplified of course):\n\n| Balance Statement | Previous Period |\n| --- | --- |\n| Cash | 201.2 |\n| Accounts Receivable | 3.1 |\n| Goodwill (Product/Intangibles Last 2 Qs) | 52.4 (25.2 + 27.2) |\n| | |\n| **Total Assets** | **284.7** |\n|  |  |\n| Current Liabilities | 16.9 |\n| Long Term Liabilities | 21.5 | \n| **Total Liabilities** | **38.4** | \n|  | |\n| Retained Earnings | -6.9 |\n| Shareholder's Equity | 253.2 |\n| |  |\n| **Total Liabilities & Equity** | **284.7** |\n\nThe working group explained how this approach artificially rewarded the company by converting expenses in R&D to an intangible goodwill asset. There is no feedback on how that asset generates and captures value -- throughput. It also came to light that the financial team was deferring certain R&D expenses as a way to bolster performance on the income statement in the short term. They had planned to realize the expenses when Acme hit the inflection point as a tax mitigation strategy, but that never happened. \n\nOnce they were finished converting their financials to throughput accounting, they were overwhelmed by the reality of their situation.\n\n| TA | Previous Period |\n| --- | --- |\n| Revenue | 68.0 |\n| Total Variable Costs (COCA) | 13.6 |\n| **Throughput** | **54.4** |\n| _Efficiency_ | _80%_ |\n|  |  |\n| R&D | 34.2 |\n| Marketing | 2.1 | \n| Sales | 25.6 | \n| G&A | 6.4 |\n| **Total OpEx** | **68.3** |\n| | |\n| **Profit** | **-13.8** |\n| | |\n| Inventory: Goodwill (previous quarter R&D) | 25.2 |\n| Inventory: 6 Months of Requirements | 120,868 |\n| **Total Inventory** | **120,893.2** |\n| | |\n| **Return On Investment** |  **-.01%** = (-13.8/120,893.2) |\n\nSince thoughput accounting (TA) defines inventory as completed and uncompleted assets within the Value Exchange System, Acme's ROI was tanked by all of the work waiting to be completed. The powerful thing that this helped Acme understand was that creating more inventory drains the overall return on investment. It also learned that by treating some of the variable costs related to COCA as an operating expense, they missed critical opportunities for increasing the efficiency of marketing operations.\n\nLooking at the financials through this perspective, Acme struggled to determine where to focus their efforts. The ongoing conversation about this exposed another critical flaw in how Acme had been approaching the Value Exchange System. Due to the speed at which they had scaled, 60,000+ work items piled up in their backlog. They struggled to understand their potential impact and effectively map them to customer and business outcomes. A product manager shared how they used they had successfully used the RICE model when they were smaller, but it had become highly inaccurate at scale given the quality of the input data and variability across product teams. She continued to explain that often, the loudest voice in the room would ultimately influence when things were worked on. In listening to her describe the flow of information between sales, customer support, and R&D, it was clear that everyone was working hard, but with such a high volume of communications, it was nearly impossible to take into account all the variables across the disparate contexts when prioritizing the roadmap.\n\nA data scientist overheard the group talking about their challenges and interjected with a suggestion -- Acme was already using machine learning and advanced statistics in their products. Why not use the same approach to optimizing their own Value Exchange System?\n\n## The first iteration towards intelligent decision support\n\nThe data scientist explained that by using pre-existing data from sources across their toolchain, it was possible to build a decision support framework using [natural language processing (NLP)](https://medium.com/analytics-vidhya/automated-keyword-extraction-from-articles-using-nlp-bfd864f41b34) to extract topics and sentiment from text-based data across all of the core tools that Acme used to sell, deliver, and support their products. Teams would not have to adopt new tools or processes, and there would be no disruption to existing business functions. Everyone put their heads together and came up with the scope for the first MVC of their intelligent decision support framework.\n\n![Decision Support Framework - Iteration 1](https://about.gitlab.com/images/blogimages/optimizing-the-value-exchange/d5.png){: .center}\n\nOnce they achieved this level of visibility, they used the topics to create alignment across teams. Additionally, instead of continuing forward using traditional project-based Gannt charts for communicating progress on timelines and deliverables, they adopted an Opportunity Tree map that better reflected the relationship between business outcomes and solving customer needs. Based on the number of customers and the sentiment of a given topic, Acme prioritized where to focus their efforts to yield the biggest impact. Looking through all of a topic's related texts across their various communication channels within a single interface, they were able to more efficiently and effectively synthesize opportunities and connected them to outcomes.   \n\n![Acme's Opportunity Tree](https://about.gitlab.com/images/blogimages/optimizing-the-value-exchange/d6.png){: .center}\n\nTo ensure they were building what the customers needed and that the solutions captured the opportunity, they started using a [continuous innovation](https://leanstack.com/library/categories/fundamentals/courses/what_is_continuous_innovation/lessons/fundamental_shift) framework to systemically validate their assumptions and solutions through [lean sprints](https://blog.leanstack.com/the-lean-sprint-bc3f9f8caafd). This did not require the introduction of a new tool, but rather a change in mindset and process among the teams.\n\n![The Lean Srint Diagram from Scaling Lean by Ash Maurya](https://about.gitlab.com/images/blogimages/optimizing-the-value-exchange/lean-sprints.png){: .center} \n\nEach sprint established a target condition, hypothesis, experiment, and measurable outcome to validate or invalidate the hypothesis. This new mental model and approach to lean product development helped Acme reduce risk and establish a repeatable cadence that accomplished the following:\n\n1. **Goal** — Break a big vision into smaller time-boxed goals.\n2. **Orient** — Align the team around problems versus solutions.\n3. **Leverage** — Source/rank a wide diversity of possible solutions or strategies for achieving the goal.\n4. **Experiment** — Test these strategies additively using experiments.\n5. **Analyze** — Reevaluate learning against the goal.\n6. **Next Action** — Decide what’s next. \n\n![The Lean Cycle Diagram from Scaling Lean by Ash Maurya](https://about.gitlab.com/images/blogimages/optimizing-the-value-exchange/lean-cycles.png){: .center} \n\nIt wasn't just product teams that adopted this process -- marketing, sales, and customer support all learned how to leverage this process to drive throughput. Before any initiative was funded, the team had to justify the effort by stating a quantifiable hypothesis that explained how it would increase throughput (T), decrease inventory (I), or reduce operating expenses (OE). By being disciplined, lean, manually improving processes, and having a rudimentary decision support framework powered by their existing data, they started to see enormous gains in their throughput. Acme was just starting to scratch the surface of the possibilities for optimizing their Value Exchange System. \n\n## The Importance Of A Timely Decision Support Framework\n\n> “You can make a system work better with surprising ease if you can give it more timely, more accurate, more complete information” -- [Thinking In Systems](https://www.amazon.com/Thinking-Systems-Donella-H-Meadows/dp/1603580557)\n\nAcme continued to invest in enhancing their intelligent decision support tool. Their ultimate goal was to model a new value stream that incorporated both the value delivery system and the value capture system. In their next iteration, they added new dimensions and types of features that were extracted from their various operational systems. They also connected their behavior analytics collection service to more effectively map how customer behavior drove specific outcomes and experiments. \n\n![Decision Support Tool - Iteration 2](https://about.gitlab.com/images/blogimages/optimizing-the-value-exchange/d7.png){: .center}\n\nBy pulling in financial data from their CRM, they derived the potential value for every single opportunity and how each mapped to the key business drivers in the Value Exchange System -- acquisition, activation, adoption, up-sell, and retention. Additionally, they automatically calculated the RICE score across all work items in a consistent manner. As a result, they removed over 99% of their inventory that did not map to valuable opportunities and adopted a \"just in time\" approach for requirements definition, drastically reducing inventory and improving ROI.\n\nAcme went a step further and augmented its Opportunity Tree used for strategic planning with a dynamic value stream map to report on the health of the Value Exchange System. Since the software code and features were integrated into the decision support framework, they consolidated all of their analytics, logging, tracing, and reporting tools that required complex queries to surface meaningful insights with a simple visual diagram of the critical paths customers took through their product. This effectively tied the Opportunity Tree to their actual product in a way that was easy for anyone at Acme to understand. It also surfaced low hanging fruit that enabled Acme to increase the value capture rate of the assets they already had.\n\n![Acme's New Value Stream Map](https://about.gitlab.com/images/blogimages/optimizing-the-value-exchange/d8.png){: .center}\n\nTo put their journey succinctly -- Acme transitioned from sensing and responding to predicting and acting. They did not become sales-led, engineering-centric, or product-first. They became throughput driven. They still have a long way to go before becoming ROI positive due to the inefficiencies of the past, but the progress thus far indicates an upward trajectory. Because of their successful transition to profitability, the board pushed Acme to continue scaling R&D and operating expenses. The CEO politely declined; showing the board their intelligent decision support tool that highlighted hundreds of millions of dollars in potential revenue just by improving what was already there. He explained that more people won't solve the problem or help them go faster. Better decisions will. \n\n![Insert Sense and Respond](https://about.gitlab.com/images/blogimages/optimizing-the-value-exchange/predictive-analytics.png){: .center} \n\nThe biggest challenge they face moving forward is staying disciplined and not falling into old habits -- something most companies are notorious for doing. It's one of the primary reasons why there are very few truly great companies out there today. Luckily, the investment in making better decisions through shorter, more effective feedback loops will let them know the day that starts happening. \n\n\u003Cbr>\n**Acme's Final Set of Reported Financials**\n\u003Cbr>\n\n| TA | Without Decision Support | Starting The Journey | Fully Integrated |\n| --- | --- | --- | --- |\n| Revenue | 68.0 | 88.4 | 138.8 |\n| Total Variable Costs (COCA) | 13.6 | 11.2 | 7.6 |\n| **Throughput** | **54.4** | **77.2** | **131.2** |\n| _Efficiency_ | _80%_ | _87%_ | _95%_ |\n|  |  | | |\n| R&D | 34.2 | 37.1 | 40.5 |\n| Marketing | 2.1 | 3.2 | 4.5 | \n| Sales | 25.6 | 28.6 | 31.9 |\n| G&A | 6.4 | 8.0 | 9.5 |\n| **Total OpEx** | **68.3** | **76.9** | **86.4** |\n| | | | |\n| **Profit** | **-13.8** | **0.3** | **44.8** |\n| | | | |\n| Inventory: Goodwill | 25.3 | 59.5 | 96.6 |\n| Inventory: Requirements | 120,868 | 117,567 | 1,176.2 |\n| **Total Inventory** | **120,893.3** | **117,626.5** | **1,272.8** |\n| | | | |\n| **Return On Investment** |  **0%** | **0.0%** | **3.5%** |\n\n\u003Cbr>\n\n## Ending On A Personal Note\n\n> “Remember that hierarchies exist to serve the bottom layers, not the top. Don’t maximize parts of the systems or subsystems while ignoring the whole. Aim to enhance total system properties, such as growth, stability, diversity, resilience, and sustainability - whether they are easily measured or not.” -- [Thinking In Systems](https://www.amazon.com/Thinking-Systems-Donella-H-Meadows/dp/1603580557)\n\nI acknowledge Acme's story and financials were maybe extreme and, at times, overly exaggerated. I did so to help illustrate the main themes that nearly every company I've been a part of has struggled with to varying degrees. Ultimately, the longer we continue to rely on a Value Exchange operating structure and toolchain that consists of dozens of siloed point solutions and people organizations, the longer we will be fixated on the local optimization of its parts. We need to shift our focus towards achieving the global maximum of the Value Exchange System. This will be an ongoing challenge for nearly every company as tranditional organizational theory is over [250 years old](https://aboutleaders.com/10-traditional-organizational-culture-problems/#gs.ij731t) and was designed to replicate factories and machines, not dynamic systems and subsystems that are constantly evolving and changing.  \n\nAs we wrap up, you might be thinking the intelligent decision support tool I described is far fetched and a pipe dream. Topic extraction, cluster analysis, sentiment analysis, Bayesian inference, and NLP does not rely on Strong AI or necessarily adopting a whole new suite of tools. It is all achievable by a capable team of data scientists and engineers using currently available open-source packages, ML libraries, and advanced statistics leveraging data most companies already have on hand. As DevOps continues to take hold and organizations shift from a project to product mindset, it's just a matter of time until someone capitalizes on the tremendous opportunity there is in optimizing the feedback loops from sales, marketing, product, engineering, UX, and customer support at scale.\n\nWhen I first started at GitLab, I thought the idea of building a single application for the entire DevOps lifecycle was insane and nearly unachievable. That challenge is part of what drew me to the company. Now that I have more perspective, I'm starting to realize that maybe we've been thinking too small. Anyone up for some dogfooding?   \n\nAll posts in the series:\n\n- Part 1: [A Gentle(ish) Introduction](/blog/optimizing-the-value-exchange-a-gentle-introduction/\n)\n- Part 2: [Reduce Waste To Increase Flow](/blog/optimizing-the-value-exchange-reduce-waste-to-increase-flow/)\n- Part 3: The Compounding Value Of Shorter Feedback Loops\n\n_**Giving credit where it is due:** The Lean Sprint and Continuous Innovation Framework were developed by Ash Maurya, founder of [LeanStack](https://leanstack.com/)._\n\nCover photo by Ryan Hafey on [Unsplash](https://unsplash.com/photos/PdJjT-zYg_E)\n",[1017,9,1018],{"slug":1057,"featured":6,"template":685},"optimizing-the-value-exchange-the-compounding-value-of-shorter-feedback-loops","content:en-us:blog:optimizing-the-value-exchange-the-compounding-value-of-shorter-feedback-loops.yml","Optimizing The Value Exchange The Compounding Value Of Shorter Feedback Loops","en-us/blog/optimizing-the-value-exchange-the-compounding-value-of-shorter-feedback-loops.yml","en-us/blog/optimizing-the-value-exchange-the-compounding-value-of-shorter-feedback-loops",{"_path":1063,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":1064,"content":1070,"config":1076,"_id":1078,"_type":13,"title":1079,"_source":15,"_file":1080,"_stem":1081,"_extension":18},"/en-us/blog/pick-your-brain-interview-brandon-foo",{"title":1065,"description":1066,"ogTitle":1065,"ogDescription":1066,"noIndex":6,"ogImage":1067,"ogUrl":1068,"ogSiteName":669,"ogType":670,"canonicalUrls":1068,"schema":1069},"Pick Your Brain interview with CEO Sid Sijbrandij","Brandon Foo, co-founder and CEO of Polymail (YC S16), recently sat down with GitLab CEO Sid Sijbrandij.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749680453/Blog/Hero%20Images/pick-your-brain-interview.jpg","https://about.gitlab.com/blog/pick-your-brain-interview-brandon-foo","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"Pick Your Brain interview with CEO Sid Sijbrandij\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"Brandon Foo\"}],\n        \"datePublished\": \"2017-06-02\",\n      }",{"title":1065,"description":1066,"authors":1071,"heroImage":1067,"date":1073,"body":1074,"category":703,"tags":1075},[1072],"Brandon Foo","2017-06-02","\n\nI sat down for a “[pick your brain](/handbook/eba/ceo-scheduling/#pick-your-brain-meetings)” meeting with GitLab’s CEO and Co-founder, [Sid Sijbrandij](/company/team/#sytses), to learn about his approach towards different aspects of building a successful startup. Here are some highlights of the conversation.\n\n\u003C!-- more -->\n\n**Brandon: When you were an earlier company around your seed stage, what were your most effective growth strategies?**\n\n**Sid:** GitLab got started as a [Show HN of GitLab.com](https://news.ycombinator.com/item?id=4428278). We’ve always tried to see where our users were and talk with them there.\n\nWhen you find people who have a need for your product, you start by trying to bring it to their attention. Then you enter a phase where they care about your product, and they start asking you for more — that’s easy, that’s the honeymoon phase. Now we’re getting to the phase where people think of GitLab as a given, and that it should be perfect, so they tell you the things that could be better.\n\n**Brandon: How do you think about product strategy with respect to building new features versus improving or increasing adoption of existing features?**\n\n**Sid:** It’s kind of a pendulum that swings back and forth. We focused a lot on new features for a while to accomplish our [idea to production vision](https://www.youtube.com/watch?v=PoBaY_rqeKA), and now this quarter [we’re focusing](/direction/) on increasing adoption of existing features. Mostly this is necessary for newer features, but that’s not the same as increasing the features’ scope, it’s more a question of how we can increase adoption for the features we already have, and seeing which functions are missing. When we release features and have the suspicion few people are using them, we evaluate to make sure those features are things that people can really use. Most recently in [9.2](/releases/2017/05/22/gitlab-9-2-released/) we added the framework to translate GitLab into any language, and allowed users to specify multiple assignees to better track shared ownership of an issue. [In 2017](/direction/#2017-goals), we’ll continue to ship features tailored for enterprise development teams, and make it easier to build, deploy, and monitor applications within GitLab.\n\n\u003Cfigure class=\"video_container\">\n  \u003Ciframe src=\"https://www.youtube.com/embed/WBf_DA0FF9k\" frameborder=\"0\" allowfullscreen=\"true\"> \u003C/iframe>\n\u003C/figure>\n\n**Brandon: How do you balance building visionary features that people aren’t necessarily asking for vs. building in direct response to customer requests?**\n\n**Sid:** We do both. We started off doing just version control and code review, and now GitLab delivers the entire DevOps pipeline, everything from chatting about an idea and planning it, to getting it out in production and monitoring. We envision enabling everyone to collaborate on digital content, so they can work together and achieve better results. No one asked for that — it’s something we did, it’s the future of the company now. We’d have been in a bad spot if we hadn’t done that.\n\nAt the same time, don’t lose track of what your customers are asking for. Balancing that is the hard part. The natural result is too little visionary stuff; if you build the right company, then everyone will be listening to your customers and screaming, “Let’s build the things customers want!” So the leadership’s task is focusing on what we need to do in order to be a better company in five years.\n\n**Brandon: Since you bootstrapped for some time, how did you decide when it was the right time to raise institutional funding?**\n\n**Sid:** One big reason is the talent we wanted to attract. While we were in YC, we tried to hire a good sales leader, but everyone we approached wanted stock in the company. We hadn’t raised any outside money so stock was all mine and my co-founder [Dmitriy’s](/company/team/#dzaporozhets) — he started GitLab and I started GitLab.com.\n\nThis made clear that if we were unable to give out stock, we were not going to hire the best people; if we’re not getting the best people, we’re going to lose in the marketplace. If you give people stock while not taking outside money, you’ll still grow but very slowly, which is not the kind of deal these executives were expecting. They expect that after 6-7 years the stock is worth something and they can get liquid. The only way to get there is to attract external capital.\n\n**Brandon: Is there anything that you would change in retrospect that you think might improve the outcome of where GitLab is today?**\n\n**Sid:** In hindsight, I’d rather have started GitLab.com a bit later. We’ve grown so fast since then that we’ve been behind in making a great experience for our users.\n\nI would focus on people running GitLab self-managed, and start GitLab.com when we were ready for it. I’d rather have people not use our product than using the product and not being absolutely happy about it. It’s not about users, it’s about happy users.\n\nIf not 100% of the users are happy, we’re not doing a good enough job.\n\n## About the Guest Author\n\nBrandon Foo is the Co-founder and CEO of [Polymail](https://polymail.io/), an email productivity platform designed for modern teams and companies.\n",[682,9,705],{"slug":1077,"featured":6,"template":685},"pick-your-brain-interview-brandon-foo","content:en-us:blog:pick-your-brain-interview-brandon-foo.yml","Pick Your Brain Interview Brandon Foo","en-us/blog/pick-your-brain-interview-brandon-foo.yml","en-us/blog/pick-your-brain-interview-brandon-foo",{"_path":1083,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":1084,"content":1090,"config":1097,"_id":1099,"_type":13,"title":1100,"_source":15,"_file":1101,"_stem":1102,"_extension":18},"/en-us/blog/pick-your-brain-interview-vincent-jong",{"title":1085,"description":1086,"ogTitle":1085,"ogDescription":1086,"noIndex":6,"ogImage":1087,"ogUrl":1088,"ogSiteName":669,"ogType":670,"canonicalUrls":1088,"schema":1089},"Key decisions for building successful startups","Vincent Jong of SaaS.CEO sits down for a 'pick your brain' meeting with GitLab CEO Sid Sijbrandij.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749680253/Blog/Hero%20Images/pick-your-brain-interview-thrive.jpg","https://about.gitlab.com/blog/pick-your-brain-interview-vincent-jong","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"Saas.CEO and Sid Sijbrandij talk key decisions, influential connections, and strategic vision when building a startup\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"Vincent Jong\"}],\n        \"datePublished\": \"2018-01-26\",\n      }",{"title":1091,"description":1086,"authors":1092,"heroImage":1087,"date":1094,"body":1095,"category":955,"tags":1096},"Saas.CEO and Sid Sijbrandij talk key decisions, influential connections, and strategic vision when building a startup",[1093],"Vincent Jong","2018-01-26","\n_GitLab CEO Sid Sijbrandij occasionally sits down for a \"[pick your brain](/handbook/eba/ceo-scheduling/#pick-your-brain-meetings)\"\nmeeting with people seeking advice on open source, remote work, or discussion of other things related to GitLab._\n\n\u003C!-- more -->\n\n**GitLab has become a leading provider in software development solutions, but it didn’t start out like that. Looking back, what were the one or two decisions that really made the company to the success it is today?**\n\nThe first one is the decision to build a company around it, because GitLab started as an open source project without a company. As such a project gets bigger, you will have to pay people to keep the quality high.\n\nAnother thing was my co-founder Dmitriy tweeting \"I want to work on GitLab full time,\" which led me to contact him and hire him, which was a great change.\n\nThis may be atypical advice on a SaaS CEO interview series, but one thing we did right was not to focus on SaaS. The demand for GitLab was coming from the self-managed side much more than from the SaaS side, so we decided to focus on that first.\n\nThe final one was the decision to apply to Y Combinator. This changed our ambition level from just running the project to being a market leader.\n\n**Would you say that your focus on the self-managed product also allowed you to focus on a different market segment than where players like GitHub were already capturing market share?**\n\nWhen we started, GitHub and Atlassian were already there in that market and it should have been locked up. But they left an opening in the self-managed market and at the bottom of the market.\n\nIn the beginning our software wasn’t very good, but we were able to rapidly make it better and grow upmarket. This is a great thing because I think today most of the revenue is coming from those large accounts.\n\nThe way I see it, source code management is one of the last things to leave self hosted for SaaS. Where this happened much earlier for CRM for example, I think source code for various reasons is transitioning later. We still see that for companies with more than 5,000 employees, 95 percent is still self hosted.\n\n**Alright. Looking at a more general perspective, what would you say you understand about building a (SaaS) company that is often overlooked or underestimated by other founders?**\n\nWhat we do differently is that we write things down. We’re a remote-only company of 200 people working from 200 locations. We try to work as asynchronously as we can and we write down what we do. The output of that is a [company handbook](/handbook/) with over 500 pages of our processes.\n\nFor a fast growing company, it is important that new people know the customs and values of the organization. Spending a lot of time to verbally communicate this is time consuming and dilutive, because you are never going to be able to tell person 100 as well as you’ve told the first. However, when you write it down, which is very painful in itself, person 100 will have an even more detailed version than person 1. So it gets better over time.\n\n**Then let’s talk about the people you’ve worked with. For startups, connecting with the right people can be a game-changer. One person can provide a connection that changes everything. If you look at people who are not employed at GitLab – which person provided essential additional value and how did you get in touch with this person?**\n\nJoining Y Combinator has been essential for us. It opened up lots of doors that would otherwise have been closed. For example, the seed round of investors we have with people like Ashton Kutcher and Michael Arrington. I don’t think they would have even looked at us if it wasn’t for Y Combinator.\n\nThen your board members are just very important. We got lucky with our first board member, Bruce Armstrong, operating partner at Khosla Ventures, who was very thoughtful with us and very hardworking in helping us every step along the way. That felt very empowering and it’s not always the case with venture capitalists, so that was awesome.\n\nSometimes it’s just reaching out. Like Matt Mullenweg who joined our board. He is the CEO of Automattic, the makers of WordPress. I just sent him an email saying “Hey, can we talk?” If you show you’ve done your homework, like mentioning why you want to talk and reference a blog post or something they tweeted, people are more likely to respond.\n\n**One of the things we do at [SaaS.CEO](https://www.saas.ceo/) is ask our audience beforehand if they have any questions for the CEO who is being interviewed. This time two questions came up. The first is coming from Michael Kamleiter, CEO of Swat.io and Walls.io. He asks \"How do you go about positioning towards other players like GitHub, especially when you were still a smaller company?\"**\n\nI don’t think we’ve figured it out yet. Where our competition was sometimes more focused on the needs of open source projects, we focused on those large customers and their requirements. For example, our competition has two levels of authorization and we have five, because our customers need more granularity.\n\nPositioning to me is mostly marketing and I think we have lagged in that regard. Actually, the last two days I have been in a workshop to figure out our positioning. What we’re going to do is articulate that GitLab is an end-to-end tool. Where all the other applications are about assembling a toolchain and orchestrating that toolchain, we want to be \"toolless.\"\n\nIf you have a toolchain, you end up having all these handoffs that create delays from working in serial. We want people with GitLab to be able to work in parallel. I think that will be a big enabler of our future growth. But it’s a really hard thing to determine, to get everybody aligned on, and then to roll it out on all your channels, from product to sales to marketing.\n\n**The second question we received is from Florian Dorfbauer, CEO of Usersnap. His question is: \"With the latest investment round, you've also revealed the bigger vision of GitLab: providing a complete DevOps experience. How much time do you spend on strategic vision building and what does the process look like to work on such strategies?\"**\n\nI consider myself a Product CEO and spend most of my time on our product. The way I spend time on this is first of all by talking with customers. My call before this was with a potential customer, to answer their questions. It’s great to be able to talk directly with customers.\n\nI also keep an eye on our issue tracker and [Hacker News](https://news.ycombinator.com/), which are important channels for me. Apart from that I work a lot with our product managers where we try to get the best out of each other.\n\nIt’s all driven by what you know about where the market is – what are the trends, what are the analysts saying, what are customers saying, what are users saying. All these things come together and you reflect on it with each other and choose a direction.\n\n**By sharing your experiences, you have given valuable input other to SaaS CEOs out there. Therefore, I want to give you the opportunity to ask something in return. Is there something our listeners can do for you?**\n\nI think it would be great that those who read this reach out to you to be interviewed so you will have more content and we can make this a bigger thing. Then when this becomes a famous podcast I can claim to be the first one ever to be interviewed.\n\nSecondly, I would like to take the opportunity to say that GitLab.com is becoming a great product now, so I hope that in 2018 people will give it a shot and try it out.\n\n**Sid, thank so much for sharing your insights. I’m very happy to have had you as our first interviewed CEO and we do hope many of the readers and listeners will follow your request.**\n\n### About the guest author\n\nVincent is a Dutch serial entrepreneur excited about advanced technology and Software as a Service solutions. While building his company, he noticed how many founders are trying to get in touch with the same people: CEOs who have already walked the path they are going. Facing the same challenge, he founded [SaaS.CEO](https://www.saas.ceo/), a platform to make successful SaaS founders more accessible. His own company [Thrive for Email](http://www.thrive.email/) is an AI-driven sales automation solution that helps sales reps increase their capacity by automatically entering all data into the CRM.\n\n[Cover image](https://unsplash.com/photos/kRnkqSKZODQ) by [Federico Beccari](https://unsplash.com/@federize) on Unsplash\n{: .note}\n",[682,9],{"slug":1098,"featured":6,"template":685},"pick-your-brain-interview-vincent-jong","content:en-us:blog:pick-your-brain-interview-vincent-jong.yml","Pick Your Brain Interview Vincent Jong","en-us/blog/pick-your-brain-interview-vincent-jong.yml","en-us/blog/pick-your-brain-interview-vincent-jong",{"_path":1104,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":1105,"content":1111,"config":1118,"_id":1120,"_type":13,"title":1121,"_source":15,"_file":1122,"_stem":1123,"_extension":18},"/en-us/blog/remote-enables-innovation",{"title":1106,"description":1107,"ogTitle":1106,"ogDescription":1107,"noIndex":6,"ogImage":1108,"ogUrl":1109,"ogSiteName":669,"ogType":670,"canonicalUrls":1109,"schema":1110},"How remote work enables rapid innovation at GitLab","At GitLab, remote isn’t a business operations risk, it’s a competitive advantage.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749678666/Blog/Hero%20Images/paper-lanterns.jpg","https://about.gitlab.com/blog/remote-enables-innovation","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"How remote work enables rapid innovation at GitLab\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"Victor Wu\"}],\n        \"datePublished\": \"2019-02-27\",\n      }",{"title":1106,"description":1107,"authors":1112,"heroImage":1108,"date":1114,"body":1115,"category":955,"tags":1116},[1113],"Victor Wu","2019-02-27","\nI’m a Product Manager here at GitLab, primarily contributing to the [Plan stage](/direction/plan/)\nof the [DevOps lifecycle](/stages-devops-lifecycle/). I joined in November 2016 and I’ve witnessed incredible\ngrowth in GitLab the product as well as GitLab the team. Many\nnew hires have asked me during [coffee chats](/company/culture/all-remote/#coffee-chats)\nabout GitLab culture and remote work in particular, since we're an [all-remote](/company/culture/all-remote/)\ncompany. My view has evolved over this time and I wanted to share specifically why I think\nremote is _not_ a challenge to overcome, but actually a _competitive advantage_, at least for GitLab.\n\n## A remote journey\n\nWhen I joined GitLab, I thought remote was a challenge to overcome or at least\nto manage. It was a risk to be mitigated. For example, I really wanted daily standup\nmeetings with the engineering team I was working with. Silicon Valley-style tech\ncompanies and product management books tell us that frequent, synchronous, face-to-face\ncommunication is necessary for building successful products efficiently and to win\nin the marketplace. To my dismay at the time, we never had in-sync standups (and\nmy team today still doesn’t have them). But curiously, we nonetheless had immense\ncollaboration and continued to ship product at a high velocity. Something really\nweird and unexpected was going on.\n\nLater on, as I started getting comfortable [doing product the GitLab way](/handbook/product/),\nI started to think that remote wasn’t really a risk, but that there were just a\nfew negatives, and that the overall effect was net positive. See the [advantages and disadvantages of remote](/company/culture/all-remote/#advantages-for-employees).\n\nToday, I realize that even a positive-negative accounting of remote is insufficient\nto articulate what remote means at GitLab. I think that remote\n(along with a few other key crucial GitLab ingredients) gives us a differentiated\nand competitive advantage, in particular allowing us to innovate at a rapid pace\nthat is truly unique. Here's why:\n\n## Interdependent ingredients\n\nThere are a several crucial and interdependent GitLab ingredients that make remote\ntruly work in our favor:\n\n### Async communication\n\nRemote implies geographic diversity (since we hire all over the world),\nand because most folks work during the day, that further implies time zone diversity.\nConsequently, we prefer **[Async communication (primarily with text)](/handbook/communication/)** as we scale our organization in\nspace-time. Async demands everything be written down and that it be clear and concise.\nYou can’t afford a prolonged back-and-forth conversation because every round-trip\ntransaction is possibly 24 hours in the worst case. In particular, we prefer text\nbecause the internet and modern apps (for example [GitLab issues](https://docs.gitlab.com/ee/user/project/issues/)) has allowed text\nto be easily organizable, searchable, and even hyperlinked. Text is easy to parse\nand thus consume. It is a highly efficient form of communication, especially for\ntransactional collaboration.\n\n### Transparency\n\nThe async communication we reference is also digital, making it infinitely\nscalable. Unlike the printed page in a physical office, anybody should\nbe able to access a digital message. So, rather than re-erecting the walls and silos\nthat plague traditional organizations and inevitably block collaboration, we\nmake communications and work **[transparent](https://handbook.gitlab.com/handbook/values/#transparency)** by default.\nAdding a layer of permissions is necessary sometimes, and in those cases it becomes an overhead cost to manage\nand use (for example fixing a security bug.) The transmitter of communications\nneeds to figure out who should receive, and set the appropriate permissions. The\nreceiver themself needs additional work to access the content. It’s more pain. It\nadds up. So we try to avoid it when we can.\n\n>Because you know everything you write down will potentially be viewed by anyone – inside or even outside the company – simply telling the truth is the optimal and most efficient strategy\n\nTransparency also makes it really easy to tell the truth, and disincentivizes dishonesty.\nTelling the truth is simply the right thing to do, but it’s also a great strategy\nto grow a long-term sustainable business. In particular, because you know everything\nyou write down will potentially be viewed by anyone in the company or even outside\nthe company, simply telling the truth is the optimal and most efficient strategy\nand you will thus adopt it with little friction. You don’t have to make up slightly\ndifferent versions for different stakeholders. You don’t have to keep track of all\nthese versions. And you only need a single artifact to document that one source\nof truth, which will never be out of sync, because there’s only one! For\nus, that single source of truth is typically the description in an issue.\n\n### Everyone can contribute\n\nWith a single source of truth that is consumable by anybody, it allows **[everyone to contribute](/company/mission/#mission)**.\nEveryone has information parity. And so anyone is welcome to contribute. In fact,\nremember I mentioned above that the transmitter of information typically has an intended receiver\nin mind? In this case, oftentimes somebody who they didn’t expect can even participate\nand add value. This isn’t possible if there’s no transparency because artificial\nbarriers pre-close the opportunities of potential collaboration. Also, everyone\ncan contribute means future folks can participate too. You may start a conversation\non an idea that turns out to be suboptimal in the current circumstances. But it\nmight end up being just a timing issue. And so posterity might be able to recover\nthe old idea and ship a feature later on, taking advantage of all the discussions\nthat were had and made available publicly.\n\nEveryone can contribute also means that the diversity of ideas skyrockets. And so\nat GitLab, people often cross departments and offer some of the best ideas to solve\nbig challenging problems. But we still have [directly responsible individuals](/handbook/people-group/directly-responsible-individuals/)\nto make decisions in order to avoid analysis paralysis.\n\n### Iteration\n\nFinally, how can all this communication and collaboration truly function if the\nmechanisms are so transactional, distributed, and unstructured? It works because\nit forces us to be **[iterative](https://handbook.gitlab.com/handbook/values/#iteration)**. Most people think they understand iteration (myself\nincluded) before joining GitLab. But I’ve discovered over and over again that new\nfolks are surprised that this concept is taken to an extreme. Product\nand code are shipped in the absolute smallest piece possible in an effort to get\nfeedback and momentum. Implementing programs and processes at GitLab means breaking\noff the smallest chunk and then putting it into action right away. We still make\nbig, bold plans and big bets on the future. But we don’t obsess over extended analysis.\nInstead we find the smallest thing that we can do now and we do it. We believe that\nwaiting until tomorrow is an opportunity cost. Doing something small today is low\nrisk and results in immediate feedback. We have a [bias for action](https://handbook.gitlab.com/handbook/values/#bias-for-action).\n\n>We believe that waiting until tomorrow is an opportunity cost. Doing something small today is low\nrisk and results in immediate feedback.\n\nAnd so if all our communication and collaboration is focused on small iterations,\nthe scope of a typical  problem is small and manageable. And it turns out (unsurprisingly)\nmore people are willing to participate in a small problem if it literally takes\nthem a few moments to voluntarily glance at an issue description, instead of being\nforced to attend a two-hour slide presentation explaining a big problem.\nAnd since the problem is made transparent by default, the pool of contributors is\nvery high, as mentioned earlier. Personally, I am actively involved\nin at least 20 to 30 parallel problem conversations on a daily basis. It is impossible\nfor anyone to achieve that level of productivity if all to those conversations required\ndedicated, ongoing, synchronous meetings. This results in an incredible rate of collaboration\nfor myself. Multiply that by all team members at GitLab, and then also all GitLab\ncommunity members further still, and you can see now why GitLab’s pace of innovation\nis ridiculously high.\n\nRemote is not a challenge for GitLab to overcome. It’s a clear business advantage.\n\n## Ending caveat\n\nThe picture I’ve painted here is one of constant messaging and wild ideas. And\nthat’s intentional because it’s true. New folks joining GitLab often are inundated\nby the number of discussions they find themselves involved in after several weeks\nin. This is indeed an ongoing risk for GitLab especially as we scale and the level\nof ideation grows exponentially in relation to headcount (since communication links\ngrow exponentially as nodes in a people network grow). I’ve observed that GitLab\nteam members usually figure out a way to cope soon enough, and typically become\nmore selective in their communications over time. I think this is a good general\nstrategy overall, because good ideas tend to get more attention, and we essentially\nrely on the wisdom of the crowds to surface them. Of course we still have well-defined\nroles and responsibilities that serve as guardrails too, that allow subject matter\nexperts and directly responsible individuals to strategically guide our innovation\nin the right general direction.\n\nHow are you making remote work work? Let us know in the comments or tweet us [@gitlab](https://twitter.com/gitlab).\n\n[Cover image](https://unsplash.com/photos/TaXPogWdzR0) by [amseaman](https://unsplash.com/@amseaman) on Unsplash\n{: .note}\n",[1117,705,706,9,1018],"collaboration",{"slug":1119,"featured":6,"template":685},"remote-enables-innovation","content:en-us:blog:remote-enables-innovation.yml","Remote Enables Innovation","en-us/blog/remote-enables-innovation.yml","en-us/blog/remote-enables-innovation",{"_path":1125,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":1126,"content":1132,"config":1138,"_id":1140,"_type":13,"title":1141,"_source":15,"_file":1142,"_stem":1143,"_extension":18},"/en-us/blog/sids-top-advice-for-startups",{"title":1127,"description":1128,"ogTitle":1127,"ogDescription":1128,"noIndex":6,"ogImage":1129,"ogUrl":1130,"ogSiteName":669,"ogType":670,"canonicalUrls":1130,"schema":1131},"Sid’s top advice for startup CEOs","Learn more about the advice Sid gives to other startup CEOs","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749680855/Blog/Hero%20Images/sid-top-advice-startup.jpg","https://about.gitlab.com/blog/sids-top-advice-for-startups","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"Sid’s top advice for startup CEOs\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"Clement Ho\"}],\n        \"datePublished\": \"2019-11-01\",\n      }",{"title":1127,"description":1128,"authors":1133,"heroImage":1129,"date":1135,"body":1136,"category":789,"tags":1137},[1134],"Clement Ho","2019-11-01","\n\n{::options parse_block_html=\"true\" /}\n\n\n\nThis past week, several startups came to Sid - GitLab's CEO and cofounder - asking for advice. In light of wanting to share advice to a wider audience, so here is some of his top advice.\n\n## Grow usage or revenue\n\nA startup that has no usage or revenue will die. It is important to be laser focused on this. By setting specific goals pertaining to this area, you can set yourself up for better success. Most startups aim for 10% week over week growth. Use that as a guidance for your own metric.\n\n## Have users switch to your product\n\nIt is better for people to switch completely to your product rather than just using parts of your product for your existing workflow. This makes your product more sticky and allows you to get better feedback from users about which features to continue investing your time and money into.\n\n## Iterate\n\nStartups tackle big problems but it’s easy to get paralyzed by big problems. It’s also easy to miss the value that your product could bring to users if you spend too much time making the big solution. By iterating, your startup should aim to build things that pay off in a week. Scoping down solutions to the bare minimum is a great start for this area.\n\n## Do things that don’t scale\n\nThis is a common phrase with startups but this is critical to do. Be willing to spend money, do tasks that clearly don’t make sense for your future scaled company because the faster you are able to find ways to grow your product with your users, the better your product will be. Airbnb has some great examples of this (detailed in the video interview below).\n\n## Interview with Sid\n\n\u003Cfigure class=\"video_container\">\n  \u003Ciframe src=\"https://www.youtube.com/embed/HklxGOXFoPQ\" frameborder=\"0\" allowfullscreen=\"true\"> \u003C/iframe>\n\u003C/figure>\n\nCover image by [Samson](https://unsplash.com/photos/ZGjbiukp_-A) on [Unsplash](https://www.unsplash.com)\n{: .note}\n",[9],{"slug":1139,"featured":6,"template":685},"sids-top-advice-for-startups","content:en-us:blog:sids-top-advice-for-startups.yml","Sids Top Advice For Startups","en-us/blog/sids-top-advice-for-startups.yml","en-us/blog/sids-top-advice-for-startups",{"_path":1145,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":1146,"content":1152,"config":1158,"_id":1160,"_type":13,"title":1161,"_source":15,"_file":1162,"_stem":1163,"_extension":18},"/en-us/blog/the-single-application-to-accelerate-your-startup",{"title":1147,"description":1148,"ogTitle":1147,"ogDescription":1148,"noIndex":6,"ogImage":1149,"ogUrl":1150,"ogSiteName":669,"ogType":670,"canonicalUrls":1150,"schema":1151},"The single application to accelerate your startup","GitLab Ultimate and Gold now free for select YC Startups.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749679046/Blog/Hero%20Images/startup.jpg","https://about.gitlab.com/blog/the-single-application-to-accelerate-your-startup","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"The single application to accelerate your startup\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"Borivoje Tasovac\"}],\n        \"datePublished\": \"2019-11-25\",\n      }",{"title":1147,"description":1148,"authors":1153,"heroImage":1149,"date":1155,"body":1156,"category":298,"tags":1157},[1154],"Borivoje Tasovac","2019-11-25","\nGitLab started as an idea that was [accepted by the HackerNews community](https://news.ycombinator.com/item?id=4428278), joined the [Y Combinator family](/blog/gitlab-is-part-of-the-y-combinator-family/), and recently was valued at $2.75 billion during the [Series E funding round](/blog/gitlab-series-e-funding/).\n\n## Giving back to the startup community\n\nWe've been supporting [Open Source projects](/solutions/open-source/join/) and [education](/solutions/education/) with our top tiers at no cost for more than a year, but in addition to that, we decided to extend this offer and also grant free access to YC companies. We understand that startups face many obstacles trying to bring value to their customers while also trying to find their place among competitors.\n\nHiring suitable candidates, making partnership decisions and trying to manage your finances can all be overwhelming, especially at an early stage of a startup. We believe that startups should set a good foundation as soon as possible and that adopting a single application for the entire DevSecOps lifecycle will allow them to focus on their customers instead of the tooling. If done right, it could also help them [avoid complex toolchains](/topics/devops/reduce-devops-costs/), [reduce cycle time](/blog/strategies-to-reduce-cycle-times/) and [secure their apps](/solutions/security-compliance/).\n\nEligible startups can choose between our [top tiers](https://about.gitlab.com/pricing/) (self-managed Ultimate or cloud-hosted Gold) free of charge. Both of them support [every stage of the DevOps lifecycle](https://about.gitlab.com/stages-devops-lifecycle/) and enable them to ship their products to the market faster.\n\n## Who can qualify?\n\nWe realize that many growing companies would benefit from this offer, but we decided to narrow the initial launch to YC companies that are members of the current or two most recent [YCombinator batches](https://www.ycombinator.com/companies) and that raised less than $3M in funding.\n\nWe might also assess other portfolios in subsequent iterations.\n\n## The offer and the application process\n\nWe are offering our most comprehensive offerings for a year for free, with optional support at a discounted rate (95% off, $4.95 per user per month).\n\nQualifying startups can submit the application form on our [Startups page](/solutions/startups/). For additional questions regarding this offer, please see our FAQ section or feel free to reach us at [startups@gitlab.com](mailto:startups@gitlab.com).\n\nCover image by [Marvin Meyer](https://unsplash.com/photos/SYTO3xs06fU) on [Unsplash](https://www.unsplash.com)\n{: .note}\n",[266,727,9],{"slug":1159,"featured":6,"template":685},"the-single-application-to-accelerate-your-startup","content:en-us:blog:the-single-application-to-accelerate-your-startup.yml","The Single Application To Accelerate Your Startup","en-us/blog/the-single-application-to-accelerate-your-startup.yml","en-us/blog/the-single-application-to-accelerate-your-startup",{"_path":1165,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":1166,"content":1172,"config":1179,"_id":1181,"_type":13,"title":1182,"_source":15,"_file":1183,"_stem":1184,"_extension":18},"/en-us/blog/we-need-to-talk-no-proxy",{"title":1167,"description":1168,"ogTitle":1167,"ogDescription":1168,"noIndex":6,"ogImage":1169,"ogUrl":1170,"ogSiteName":669,"ogType":670,"canonicalUrls":1170,"schema":1171},"We need to talk: Can we standardize NO_PROXY?","Subtle differences in proxy setting implementations led to surprising problems for a GitLab customer. Here's how we got to the root of it.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749659507/Blog/Hero%20Images/AdobeStock_623844718.jpg","https://about.gitlab.com/blog/we-need-to-talk-no-proxy","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"We need to talk: Can we standardize NO_PROXY?\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"Stan Hu\"}],\n        \"datePublished\": \"2021-01-27\",\n      }",{"title":1167,"description":1168,"authors":1173,"heroImage":1169,"date":1175,"updatedDate":1176,"body":1177,"category":677,"tags":1178},[1174],"Stan Hu","2021-01-27","2025-06-09","If you've used a Web proxy server before, you're probably familiar with\nthe environment variables `http_proxy` or `HTTP_PROXY`. You may be less\nfamiliar with `no_proxy`, which provides a way to exclude traffic\ndestined to certain hosts from using the proxy. While HTTP is a\nwell-defined standard, no standard exists for how clients should handle\nthese variables. As a result, Web clients support these variables in\nsubtly different ways. For one GitLab customer, these differences led\nto a weekend of troubleshooting that uncovered why certain services\nstopped communicating.\n\n## What is a proxy server?\n\nA proxy server acts as an intermediary between your computer or network and the internet. When you send a request to access a website or other online resource, that request first goes to the proxy server. The proxy server then forwards the request to the actual destination and delivers the response back to you. Proxies can serve various purposes, including improving security, enhancing privacy, and controlling internet usage.\n\n## Proxy server environment variables\n\nLet's now look at what proxy server environment variables are, and how to define exemptions and handle exclusions with `no_proxy`.\n\n### Understanding proxy server environment variables \n\nToday, most Web clients support connection to proxy servers via\nenvironment variables:\n\n- `http_proxy` / `HTTP_PROXY`\n- `https_proxy` / `HTTPS_PROXY`\n- `no_proxy` / `NO_PROXY`\n\nThese variables tell the client what URL should be used to access the\nproxy servers and which exceptions should be made. For example, if you\nhad a proxy server listening on `http://alice.example.com:8080`, you\nmight use it via:\n\n```sh\nexport http_proxy=http://alice.example.com:8080\n```\n\nWhich proxy server gets used if troublesome Bob also defines the\nall-caps version, `HTTP_PROXY`?\n\n```sh\nexport HTTP_PROXY=http://bob.example.com:8080\n```\n\nThe answer surprised us: it depends. In some cases, the Alice proxy\nwins, and in other cases Bob wins. We'll discuss the details later.\n\n### Defining proxy exemptions with `no_proxy`\n\nWhat happens if you want to make exceptions? For example, suppose you\nwant to use a proxy server for everything but `internal.example.com` and\n`internal2.example.com`. That's where the `no_proxy` variable comes into\nplay. Then you would define `no_proxy` as follows:\n\n```sh\nexport no_proxy=internal.example.com,internal2.example.com\n```\n\n### Handling IP exclusions in `no_proxy`\n\nWhat if you want to exclude IP addresses? Can you use asterisks or\nwildcards? Can you use CIDR blocks (e.g. `192.168.1.1/32`)? The answer\nagain: it depends.\n\n## How did we get here?\n\nLet's dig into the evolution of proxy variables, and how they are used today.\n\n### The origins of proxy variables\n\nIn 1994, most Web clients used CERN's `libwww`, which [supported `http_proxy` and the `no_proxy` environment variables](https://courses.cs.vt.edu/~cs4244/spring.09/documents/Proxies.pdf).\n`libwww` only used the lowercase form of `http_proxy`, and the [`no_proxy` syntax was\nsimple](https://github.com/w3c/libwww/blob/8678b3dcb4191065ca39caea54bb1beba809a617/Library/src/HTAccess.c#L234-L239):\n\n```\nno_proxy is a comma- or space-separated list of machine\nor domain names, with optional :port part.  If no :port\npart is present, it applies to all ports on that domain.\n\nExample:\n\t\tno_proxy=\"cern.ch,some.domain:8001\"\n```\n\nNew clients emerged that added their own HTTP implementations without\nlinking `libwww`. In January 1996, Hrvoje Niksic released\n`geturl`, the predecessor of what is now `wget`.  A month later,\n`geturl`, [added support for `http_proxy` in v1.1](https://ftp.sunet.se/mirror/archive/ftp.sunet.se/pub/www/utilities/wget/old-versions/).\nIn May 1996, `geturl` v1.3 added support for `no_proxy`. Just as with\n`libwww`, `geturl` only supported the lowercase form.\n\nIn January 1998, Daniel Stenberg released `curl` v5.1, which [supported the `http_proxy` and `no_proxy` variables](https://github.com/curl/curl/blob/ae1912cb0d494b48d514d937826c9fe83ec96c4d/CHANGES#L929-L944).\nIn addition, `curl` allowed the uppercase forms, `HTTP_PROXY` and `NO_PROXY`.\n\nPlot twist: In March 2009, [curl v7.19.4](https://github.com/curl/curl/releases/tag/curl-7_19_4) dropped support for the\nuppercase form of `HTTP_PROXY` [due to security concerns](https://github.com/curl/curl/blob/30e7641d7d2eb46c0b67c0c495a0ea7e52333ee2/lib/url.c#L2250-L2261). However, while `curl` ignores `HTTP_PROXY`, `HTTPS_PROXY` still works today.\n\n### State of the variables today\n\nFast-forward to today. As my [colleague Nourdin el Bacha researched](https://gitlab.com/gitlab-com/support/support-team-meta/-/issues/2991),\nwe can see that how these proxy server variables are handled varies, depending\non what language or tool you are using.\n\n## Current implementation of proxy variables across languages\n\nKnowing how proxy variables are handled across languages allows you to set them so that they work properly. Here’s a quick rundown.\n\n### `http_proxy` and `https_proxy`\n\nIn the following table, each row represents a supported behavior, while\neach column holds the tool (e.g. `curl`) or language (e.g. `Ruby`) to\nwhich it applies:\n\n|                 | curl      | wget           | Ruby          | Python    | Go        |\n|-----------------|-----------|----------------|---------------|-----------|-----------|\n| `http_proxy`    | Yes       | Yes            | Yes           | Yes       | Yes       |\n| `HTTP_PROXY`    | No        | No             | Yes ([warning](https://github.com/ruby/ruby/blob/0ed71b37fa9af134fdd5a7fd1cebd171eba83541/lib/uri/generic.rb#L1519)) | Yes (if `REQUEST_METHOD` not in env)       | Yes       |\n| `https_proxy`   | Yes       | Yes            | Yes           | Yes       | Yes       |\n| `HTTPS_PROXY`   | Yes       | No             | Yes           | Yes       | Yes       |\n| Case precedence | lowercase | lowercase only | lowercase     | lowercase | Uppercase |\n| Reference       | [source](https://github.com/curl/curl/blob/30e7641d7d2eb46c0b67c0c495a0ea7e52333ee2/lib/url.c#L2250-L2266) | [source](https://github.com/jay/wget/blob/099d8ee3da3a6eea5635581ae517035165f400a5/src/retr.c#L1222-L1239) | [source](https://github.com/ruby/ruby/blob/0ed71b37fa9af134fdd5a7fd1cebd171eba83541/lib/uri/generic.rb#L1474-L1543) | [source](https://github.com/python/cpython/blob/030a713183084594659aefd77b76fe30178e23c8/Lib/urllib/request.py#L2488-L2517) | [source](https://github.com/golang/go/blob/682a1d2176b02337460aeede0ff9e49429525195/src/vendor/golang.org/x/net/http/httpproxy/proxy.go#L82-L97) |\n\u003Cbr>\u003C/br>\nNote that `http_proxy` and `https_proxy` are always supported across the\nboard, while `HTTP_PROXY` is not always supported. Python (via `urllib`) complicates\nthe picture even more: `HTTP_PROXY` can be used [as long as\n`REQUEST_METHOD` is not defined in the environment](https://github.com/python/cpython/blob/030a713183084594659aefd77b76fe30178e23c8/Lib/urllib/request.py#L2504-L2508).\n\nWhile you might expect environment variables to be all-caps,\n`http_proxy` came first, so that's the de facto standard. When in doubt,\ngo with the lowercase form because that's universally supported.\n\nInstead of environment variables, Java uses [system properties](https://docs.oracle.com/javase/8/docs/technotes/guides/net/proxies.html). This avoids case issues entirely.\n\nUnlike most implementations, Go tries the uppercase version before\nfalling back to the lowercase version. We will see later why that caused\nissues for one GitLab customer.\n\n### `no_proxy` format\n\nSome users have [discussed the lack of the `no_proxy` specification in this issue](https://github.com/curl/curl/issues/1208). As\n`no_proxy` specifies an exclusion list, many questions arise about\nhow it behaves. For example, suppose your `no_proxy` configuration is defined:\n\n```sh\nexport no_proxy=example.com\n```\n\nDoes this mean that the domain must be an exact match, or will\n`subdomain.example.com` also match against this configuration? The\nfollowing table shows the state of various implementations. It turns out\nall implementations will match suffixes properly, as shown in the\n`Matches suffixes?` row:\n\n|                       | curl      | wget           | Ruby      | Python    | Go        |Java |\n|-----------------------|-----------|----------------|-----------|-----------|-----------|\n| `no_proxy`            | Yes       | Yes            | Yes       | Yes       | Yes       |No*|\n| `NO_PROXY`            | Yes       | No             | Yes       | Yes       | Yes       |No*|\n| Case precedence       | lowercase | lowercase only | lowercase | lowercase |Uppercase |N/A|\n| Matches suffixes?     | Yes       | Yes            | Yes       | Yes       | Yes       |No|\n| Strips leading `.`?   | Yes       | No             | Yes       | Yes       | No        |No|\n| `*` matches all hosts?| Yes       | No             | No        | Yes       | Yes       |Yes|\n| Supports regexes?     | No        | No             | No        | No        | No        |No|\n| Supports CIDR blocks? | No        | No             | Yes       | No        | Yes       |No|\n| Detects loopback IPs? | No        | No             | No        | No        | Yes       |No|\n| Resolves IP addresses? | No        | No             | Yes        | No        | Yes       |No|\n| Reference             | [source](https://github.com/curl/curl/blob/30e7641d7d2eb46c0b67c0c495a0ea7e52333ee2/lib/url.c#L2152-L2206) | [source](https://github.com/jay/wget/blob/099d8ee3da3a6eea5635581ae517035165f400a5/src/retr.c#L1266-L1274) | [source](https://github.com/ruby/ruby/blob/eead83160bcc5f49706e05669e5a7e2620b9b605/lib/uri/generic.rb#L1552-L1577) | [source](https://github.com/python/cpython/blob/030a713183084594659aefd77b76fe30178e23c8/Lib/urllib/request.py#L2519-L2551)| [source](https://github.com/golang/go/blob/master/src/vendor/golang.org/x/net/http/httpproxy/proxy.go#L170-L205) |[documentation](https://docs.oracle.com/javase/8/docs/technotes/guides/net/proxies.html)\n\n* Java uses the `http.nonProxyHosts` system property.\n\n### The impact of leading dots in no_proxy\n\nHowever, if there is a leading `.` in the `no_proxy` setting, the\nbehavior varies. For example, `curl` and `wget` behave\ndifferently. `curl` will always strip the leading `.` and match against\na domain suffix. This call bypasses the proxy:\n\n```sh\n$ env https_proxy=http://non.existent/ no_proxy=.gitlab.com curl https://gitlab.com\n\u003Chtml>\u003Cbody>You are being \u003Ca href=\"https://about.gitlab.com/\">redirected\u003C/a>.\u003C/body>\u003C/html>\n```\n\nHowever, `wget` does not strip the leading `.` and performs an exact\nstring match against a hostname. As a result, `wget` attempts to use a\nproxy if a top-level domain is used:\n\n```sh\n$ env https_proxy=http://non.existent/ no_proxy=.gitlab.com wget https://gitlab.com\nResolving non.existent (non.existent)... failed: Name or service not known.\nwget: unable to resolve host address 'non.existent'\n```\n\nIn all implementations, regular expressions are never supported. I\nsuspect using regexes complicates matters further, because regexes have\ntheir own flavors (e.g. PCRE, POSIX, etc.). Using regexes also\nintroduces potential performance and security issues.\n\nIn some cases, setting `no_proxy` to `*` effectively disables proxies\naltogether, but this is not a universal rule.\n\nOnly Ruby performs a DNS lookup to resolve a hostname to an IP address when deciding if a proxy should be used. Be careful if you use IP addresses with Ruby because it’s possible a hostname may resolve to an excluded IP address. In general, do not specify IP addresses in no_proxy variable unless you expect that the IPs are explicitly used by the client.\n\nThe same holds true for CIDR blocks, such as `18.240.0.1/24`. CIDR\nblocks only work when the request is directly made to an IP\naddress. Only Go and Ruby allow CIDR blocks. Unlike other\nimplementations, Go even automatically disables the use of a proxy if it\ndetects a loopback IP addresses.\n\n## Why does this matter?\nDiscrepancies in proxy environment variable handling, particularly between Ruby and Go, can lead to a real-world issues where Git pushes worked via the command line but failed in the web UI for a GitLab customer. Understanding these inconsistencies is crucial for troubleshooting and configuring applications that operate across multiple languages within corporate networks utilizing proxy servers.\n\n### Challenges of defining proxy variables in multi-language applications\n\nIf you have an application written in multiple languages that needs to\nwork behind a corporate firewall with a proxy server, you may need to\npay attention to these differences. For example, GitLab is composed of a\nfew services written in Ruby and Go. One customer set its proxy\nconfiguration to something like the following:\n\n```yaml\nHTTP_PROXY: http://proxy.company.com\nHTTPS_PROXY: http://proxy.company.com\nNO_PROXY: .correct-company.com\n```\n\nThe customer reported the following issue with GitLab:\n\n1. A `git push` from the command line worked\n1. Git changes made via the Web UI failed\n\nOur support engineers discovered that due to a Kubernetes configuration\nissue, a few stale values lingered. The pod actually had an environment\nthat looked something like:\n\n```yaml\nHTTP_PROXY: http://proxy.company.com\nHTTPS_PROXY: http://proxy.company.com\nNO_PROXY: .correct-company.com\nno_proxy: .wrong-company.com\n```\n### How inconsistent proxy settings can cause failures\n\nThe inconsistent definitions in `no_proxy` and `NO_PROXY` set off red\nflags, and we could have resolved the issue by making them consistent or\nremoving the incorrect entry. But let's drill into what happened.\nRemember from above that:\n\n1. Ruby tries the lowercase form first\n1. Go tries the uppercase form first\n\nAs a result, services written in Go, such as GitLab Workhorse, had the\ncorrect proxy configuration. A `git push` from the command line worked\nfine because the Go services primarily handled this activity:\n\n```mermaid\nsequenceDiagram\n    participant C as Client\n    participant W as Workhorse\n    participant G as Gitaly\n    C->>W: 1. git push\n    W->>G: 2. gRPC: PostReceivePack\n    G->>W: 3. OK\n    W->>C: 4. OK\n```\n\nThe gRPC call in step 2 never attempted to use the proxy because\n`no_proxy` was configured properly to connect directly to Gitaly.\n\nHowever, when a user makes a change in the UI, Gitaly forwards the\nrequest to a `gitaly-ruby` service, which is written in\nRuby. `gitaly-ruby` makes changes to the repository and [reports back\nvia a gRPC call back to its parent process](https://gitlab.com/gitlab-org/gitaly/-/issues/3189).  However,\nas seen in step 4 below, the reporting step didn't happen:\n\n```mermaid\nsequenceDiagram\n    participant C as Client\n    participant R as Rails\n    participant G as Gitaly\n    participant GR as gitaly-ruby\n    participant P as Proxy\n    C->>R: 1. Change file in UI\n    R->>G: 2. gRPC: UserCommitFiles\n    G->>GR: 3. gRPC: UserCommitFiles\n    GR->>P: 4. CONNECT\n    P->>GR: 5. FAIL\n```\n\nBecause gRPC uses HTTP/2 as the underlying transport, `gitaly-ruby`\nattempted a CONNECT to the proxy since it was configured with the wrong\n`no_proxy` setting. The proxy immediately rejected this HTTP request,\ncausing the failure in the Web UI push case.\n\n### Correcting proxy configuration issues\n\nOnce we eliminated the lowercase `no_proxy` from the environment, pushes\nfrom the UI worked as expected, and `gitaly-ruby` connected directly to\nthe parent Gitaly process. Step 4 worked properly in the diagram below:\n\n```mermaid\nsequenceDiagram\n    participant C as Client\n    participant R as Rails\n    participant G as Gitaly\n    participant GR as gitaly-ruby\n    participant P as Proxy\n    C->>R: 1. Change file in UI\n    R->>G: 2. gRPC: UserCommitFiles\n    G->>GR: 3. gRPC: UserCommitFiles\n    GR->>G: 4. OK\n    G->>R: 5. OK\n    R->>C: 6. OK\n```\n\n## A surprising discovery with gRPC\n\nWe also discovered that gRPC does not [support HTTPS proxies](https://github.com/grpc/grpc/issues/20939). This again subtly affects the behavior of the system depending on how `HTTPS_PROXY` is set.\n\n### gRPC behavior with `HTTPS_proxy`\n\nNote that the customer set `HTTPS_PROXY` to an unencrypted HTTP proxy;\nnotice that `http://` is used instead of `https://`. While this isn't\nideal from a security standpoint, some people do this to avoid the\nhassle of clients failing due to TLS certificate verification issues.\n\nIronically, if an HTTPS proxy were specified, we would not have seen\nthis problem. If an HTTPS proxy is used, gRPC will ignore this setting\nsince HTTPS proxies are not supported.\n\n### The lowest common denominator\n\nI think we can all agree that one should never define inconsistent\nvalues with lowercase and uppercase proxy settings. However, if you ever\nhave to manage a stack written in multiple languages, you might need to\nconsider setting HTTP proxy configurations to the lowest common\ndenominator.\n\n1. `http_proxy` and `https_proxy`\n\n* Use lowercase form. `HTTP_PROXY` is not always supported or recommended.\n    * If you _absolutely must_ use the uppercase form as well, be **sure** they share the same value.\n\n2. `no_proxy`\n\n1. Use lowercase form.\n1. Use comma-separated `hostname:port` values.\n1. IP addresses are okay, but hostnames are never resolved.\n1. Suffixes are always matched (e.g. `example.com` will match `test.example.com`).\n1. If top-level domains need to be matched, avoid using a leading dot (`.`).\n1. Avoid using CIDR matching since only Go and Ruby support that.\n\n## Steps toward standardizing `no_proxy`\n\nKnowing the least common denominator can help avoid issues if these\ndefinitions are copied for different Web clients. But should `no_proxy`\nand the other proxy settings have a documented standard rather than an\nad hoc convention? The list below may serve as a starting point for a\nproposal:\n\n1. Prefer lowercase forms over uppercase variables (e.g. `http_proxy` should be searched before `HTTP_PROXY`).\n1. Use comma-separated `hostname:port` values.\n    * Each value may include optional whitespace.\n1. Never perform DNS lookups or use regular expressions.\n1. Use `*` to match **all** hosts.\n1. Strip leading dots (`.`) and match against domain suffixes.\n1. Support CIDR block matching.\n1. Never make assumptions about special IP addresses (e.g. loopback addresses in `no_proxy`).\n\n## Key takeaways on proxy standardization\n\nIt's been over 25 years since the first Web proxy was released. While\nthe basic mechanics of configuring a Web client via environment\nvariables have not changed much, a number of subtleties have emerged\nacross different implementations. We saw for one customer, erroneously\ndefining conflicting `no_proxy` and `NO_PROXY` variables led to hours of\ntroubleshooting due to the differences with which Ruby and Go parse\nthese settings. We hope highlighting these differences will avoid future\nissues in your production stack, and we hope that Web client maintainers\nwill standardize the behavior to avoid such issues in the first place.\n",[266,892,935,9],{"slug":1180,"featured":6,"template":685},"we-need-to-talk-no-proxy","content:en-us:blog:we-need-to-talk-no-proxy.yml","We Need To Talk No Proxy","en-us/blog/we-need-to-talk-no-proxy.yml","en-us/blog/we-need-to-talk-no-proxy",{"_path":1186,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":1187,"content":1193,"config":1200,"_id":1202,"_type":13,"title":1203,"_source":15,"_file":1204,"_stem":1205,"_extension":18},"/en-us/blog/what-i-learned-about-our-ceo-s-job-from-participating-in-the-ceo-shadow-program",{"title":1188,"description":1189,"ogTitle":1188,"ogDescription":1189,"noIndex":6,"ogImage":1190,"ogUrl":1191,"ogSiteName":669,"ogType":670,"canonicalUrls":1191,"schema":1192},"Lessons from my time in GitLab's CEO Shadow program","GitLab's CEO Shadow program gives team members insight into what the CEO does","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749666603/Blog/Hero%20Images/book.jpg","https://about.gitlab.com/blog/what-i-learned-about-our-ceo-s-job-from-participating-in-the-ceo-shadow-program","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"What I Learned about the CEO's Job from Participating in GitLab's CEO Shadow Program\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"Emilie Schario\"}],\n        \"datePublished\": \"2019-10-07\",\n      }",{"title":1194,"description":1189,"authors":1195,"heroImage":1190,"date":1197,"body":1198,"category":789,"tags":1199},"What I Learned about the CEO's Job from Participating in GitLab's CEO Shadow Program",[1196],"Emilie Schario","2019-10-07","\n\n{::options parse_block_html=\"true\" /}\n\n\n\nOver the first two weeks in August, I got the chance to participate in GitLab's [CEO Shadow program](/handbook/ceo/shadow/#rotation-schedule), during which Shadows attend upwards of 90% of the CEO's meetings over the course of their 2-week rotation.\nWe attend job interviews, 1-on-1s that the CEO has with his direct reports, investor calls, and everything else that comes up.\nDuring my two week rotation, I was asked to drop off calls twice and asked to not join a 1:1 once.\nShadows are welcome to join any and all meetings, except for where the guest requests otherwise.\n\nWhile I had seen Sid's calendar leading up to my Shadow rotation, I really did not know what to expect.\nI had chatted with other [Shadow alumni](/handbook/ceo/shadow/#ceo-shadow-program-alumni-and-learnings), but their words of advice did not convey exactly how to prepare.\nI knew I wanted to turn the whole experience into a learning opportunity where I could sponge up every interaction.\nI knew there would be a lot going on.\nI planned to take it all in, then leave room for synthesis later.\n\nNow, as I reflect on my notes, I think the biggest clarity I've gotten is on *what the CEO's job is* and how that is reflected in Sid's day-to-day.\nFrom what I saw, no two days are alike - the weeks even less so - but when I sat back there were three obvious themes that emerged: making big decisions, reinforcing our values, and stepping in wherever there is a leadership gap.\n\n## Making Big Decisions\n\nAt GitLab, we believe in having a [DRI](/handbook/people-group/directly-responsible-individuals/) - Directly Responsible Individual - for most decisions.\nFor example, product managers are DRIs for their areas of the product. Having DRIs is an incredible aspect of GitLab, as it helps keep [collaboration](https://handbook.gitlab.com/handbook/values/#collaboration) from devolving into decisions by consensus and helps empower us to work asynchronously.\n\nIf there's a moment where a decision needs to be made, a DRI can make a decision.\nIf another piece of information comes up, we can make another decision. Rarely are decisions [one-way door decisions](https://handbook.gitlab.com/handbook/values/#make-two-way-door-decisions).\n[Making decisions quickly](/handbook/leadership/mitigating-concerns/#loss-of-velocity) is key to how [we can ship](/releases/categories/releases/) as much as we do, [iterating](https://handbook.gitlab.com/handbook/values/#iteration) along the way.\n\nI think it probably seems obvious that a CEO makes decisions all the time, but given the way we set DRIs for things, I was curious what sorts of decisions I could see Sid making, and I wanted to understand *why those decisions* were being escalated to Sid.\nMy cool discovery is that he was doing the things described in his [job family](https://handbook.gitlab.com/job-families/chief-executive-officer/).\nThe second requirement of the CEO's job description is\n> Hire great people. Help people that are not a good fit find another job.\n\nFor many leadership roles, Sid is still involved in the hiring decision.\nSince [lowering the hiring bar is one of our biggest concerns](/handbook/leadership/mitigating-concerns/#lowering-the-hiring-bar), it makes sense that this is an area where the CEO would spend his time.\nSid makes [pricing decisions](https://gitlab.com/gitlab-com/Product/blob/master/.gitlab/issue_templates/Move-A-Feature-Down.md#L25) because pricing is the CEO's job.\nSid makes the decisions that it's his job to be making.\nThere's no secret process behind the curtain.\n\n## Reinforcing our GitLab Values\n\nEvery interaction with Sid is an opportunity for him to reinforce the company values.\nWhat may look to some like *handbook policing* is simply him pushing us to continue to [write things down](https://handbook.gitlab.com/handbook/values/#write-things-down).\n*An obligatory breakout call* is an opportunity to [get to know each other](https://handbook.gitlab.com/handbook/values/#get-to-know-each-other).\nAnd *thinking too small* is just [scoping to the minimum viable change](https://handbook.gitlab.com/handbook/values/#move-fast-by-shipping-the-minimum-viable-change).\n\nHalf way through my first week, I was lucky enough to attend [TractionConf](https://www.tractionconf.io/) with Sid.\nThere he would be participating in a Fireside Chat with Frederic Lardinois of TechCrunch to talk about GitLab's origin story, remote work, our transparency value, and all the things that make GitLab unique.\nGiven that this was the first time I had seen Sid speak in-person at a non-GitLab event, I opened up a Google Doc and started typing.\nI didn't have a sense of what it'd be for or how it'd be useful.\nI figured better to have the notes than not and just went for it.\n\nTwo things happened that I could not have expected.\nFirst, because I shared the doc in our #ceo channel in Slack, people helped me clean it up as I was typing notes.\nI didn't have to worry about that `teh` that I mistyped because team members who were reading my notes as the talk was happening were helping me make it better.\nSecond, when someone tweeted about the session later on, Sid suggested I reply with my notes.\nWhat may look to some as *pieced together notes* is really just accepting that [everything is in draft](https://handbook.gitlab.com/handbook/values/#everything-is-in-draft) and working with a [low level of shame](https://handbook.gitlab.com/handbook/values/#low-level-of-shame).\nNow those notes can serve not just me, but anyone who wants to read about the session.\nThey're not perfect, but they are better than nothing.\nLearning to work with a low level of shame is hard - probably one of the hardest transitions about working at GitLab! - but it makes everyone's work experience better.\n\n{::options parse_block_html=\"true\" /}\n\u003Cblockquote class=\"twitter-tweet\">\n\u003Cp lang=\"en\" dir=\"ltr\">I really enjoyed the talk too! Here are my notes from the session, in case they're useful \u003Ca href=\"https://twitter.com/hashtag/TractionConf?src=hash&ref_src=twsrc%5Etfw\">#TractionConf\u003C/a> \u003Ca href=\"https://t.co/b55bQITbNv\">https://t.co/b55bQITbNv\u003C/a>\u003C/p>— Emilie Schario (@emilieschario) \u003Ca href=\"https://twitter.com/emilieschario/status/1159667866918109190?ref_src=twsrc%5Etfw\">August 9, 2019\u003C/a>\n\u003C/blockquote>\n\nBy working with a low level of shame, I made it so that everyone could contribute!\n\nWhile I like to think I'm good at working *in the GitLab way* and encourage my peers to do the same, I never go out of my way to push people to work even more-so.\nI saw most interactions of Sid's have some aspect of stewarding our values.\nWhether he was coaching his direct reports on [how we organize the handbook](/handbook/handbook-usage/#style-guide-and-information-architecture), suggesting someone stop sharing their screen in a [Group Conversation](/handbook/group-conversations/) so that it feels more like a conversation, or pointing out that a section name is not [MECEFU](/handbook/communication/#mecefu-terms), Sid was regularly stewarding our values.\n\n## Stepping in as a Leader\n\nI saw this most obviously when I first started at GitLab before we had a CMO.\nSid was acting-CMO, so the many ways he was stepping in to run marketing was really visible in the company - most obviously in running the Group Conversations for Marketing.\n\nWe see this behavior in most managers.\nWhen a team member is on vacation or leave, their managers may step in to fill the gap.\nSid does the same for his direct reports.\nThis is most visible today in how we use [PTO Ninja](/handbook/paid-time-off/#pto-ninja).\n\nLots of CEO Shadows have walked away with [different takeaways](/handbook/ceo/shadow/#ceo-shadow-program-alumni-and-learnings).\nThe biggest understanding I walked away with was this: while Sid's job is to make decisions, steward our values, and step in when needed, it's actually expected of all of us too!\nWe live GitLab's values by having DRIs, helping steward our culture in small group settings, and fostering collaboration.\nIf you're a GitLab team member, go for the CEO Shadow program, if given the chance.\nIf you're thinking about rolling out a similar program at your company, I hope our details on [how the shadow program works at GitLab can be your blueprint](/handbook/ceo/shadow/).\n",[892,705,9],{"slug":1201,"featured":6,"template":685},"what-i-learned-about-our-ceo-s-job-from-participating-in-the-ceo-shadow-program","content:en-us:blog:what-i-learned-about-our-ceo-s-job-from-participating-in-the-ceo-shadow-program.yml","What I Learned About Our Ceo S Job From Participating In The Ceo Shadow Program","en-us/blog/what-i-learned-about-our-ceo-s-job-from-participating-in-the-ceo-shadow-program.yml","en-us/blog/what-i-learned-about-our-ceo-s-job-from-participating-in-the-ceo-shadow-program",{"_path":1207,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":1208,"content":1214,"config":1220,"_id":1222,"_type":13,"title":1223,"_source":15,"_file":1224,"_stem":1225,"_extension":18},"/en-us/blog/why-we-pay-local-rates",{"title":1209,"description":1210,"ogTitle":1209,"ogDescription":1210,"noIndex":6,"ogImage":1211,"ogUrl":1212,"ogSiteName":669,"ogType":670,"canonicalUrls":1212,"schema":1213},"Why GitLab pays local rates","Our compensation structure is known to spark controversy, so we want to give an update on our latest iteration on team member salaries.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749680461/Blog/Hero%20Images/local-rates.jpg","https://about.gitlab.com/blog/why-we-pay-local-rates","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"Why GitLab pays local rates\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"Aricka Flowers\"}],\n        \"datePublished\": \"2019-02-28\",\n      }",{"title":1209,"description":1210,"authors":1215,"heroImage":1211,"date":1217,"body":1218,"category":955,"tags":1219},[1216],"Aricka Flowers","2019-02-28","\n\nOur [compensation calculator](/handbook/total-rewards/compensation/compensation-calculator/) is a regular [hot topic on places like Hacker News](https://news.ycombinator.com/item?id=18441768#18443167) – pretty much any thread about GitLab has a comment about us paying local rates. As with everything GitLab does, we continue to [iterate](https://handbook.gitlab.com/handbook/values/#iteration) on our compensation model, and implemented a number of changes at the start of 2019. In addition to adjusting the salaries of backend developers, which were [raised considerably](https://gitlab.com/gitlab-com/www-gitlab-com/commit/9382348c3c81b92b598b0a6da0994d387bdfc404) so that we are [\"at or above market,\"](/handbook/total-rewards/compensation/#competitive-rate) according to GitLab CEO [Sid Sijbrandij](/company/team/#sytses), the location factor was also revised to better reflect the respective areas covered.\n\nBut first, let's take a step back to see how we got to here.\n\n### Why did GitLab start paying team members according to location in the first place?\n\n\"It’s something that kind of happened organically,\" Sid says. \"Every time we hired someone, we’d discuss what a reasonable compensation would be. And many times, it came back to what they were making beforehand, and that really depended a lot on where they were. So we kind of started out having local market salaries as we grew. At a certain point, we said, 'Okay, this is apparently the standard. We’re basing it not just on your function and the seniority you have in the function, but also where you live.'\"\n\nGitLab no longer uses salary history as a factor in compensation offers and does not ask candidates about their previous pay. Instead, we ask all candidates, regardless of location, for their salary expectations.\n\n### Understanding the rent index\n\nThe compensation calculator's rent index came from a noted correlation between the aforementioned local market rate salaries and rent prices in the area. Using limited data sets with more than 100 locations across the globe, an analysis was run to determine the best gauge for local rates. The rent as listed on Numbeo was found to have the highest correlation.\n\n\"When you think about it, the correlation we found made sense,\" Sid explains. \"If there’s a place where people pay high wages, it tends to attract people. And then the rents, almost by force of nature, start rising. It’s not that we want to pay you based on your rent or compensate your cost of living. We want to make sure that we pay at or above market. We found that the rent was a great way to calculate that, and it’s why there’s a rent index as part of our global compensation formula.\"\n\n### New improvements on local market calculations\n\nGitLab compensation is calculated by delving into [local market data, when possible](/handbook/total-rewards/compensation/compensation-calculator/#location-factor), to ensure that [salaries are being tabulated](https://gitlab.com/gitlab-com/www-gitlab-com/merge_requests/17460) fairly.\n\n\"Instead of using [just the rate index], what we do now is look at a number of different sources, usually four or five, to get market data for a city,\" says GitLab's outgoing Chief Culture Officer [Barbie Brewer](/company/team/#BarbieJBrewer). \"Then we find the median of that, and use it as our benchmark. That being said, you can't do this in all cities. We have a lot of employees in jobs that aren't typically available where they are located. In those instances, we fall back on the other equation. Generally speaking, it's pretty close. When we've had to go back and check those benchmarks, we found that it required very few adjustments. We were getting it right 95 percent of the time, so doing that check was good. It helped us understand that we were not that far off.\"\n\nBarbie also noted that some employees in low-income communities could fare better than expected because people in towns within 90 minutes of a large city will have their salaries calculated according to the higher metropolitan factor.\n\nNow that we know how GitLab got started with local rates, here's a look at why we have continued down this path.\n\n### Standard pay eats away at production and personnel\n\nIf everyone is paid the same role-based salary, the company would not be able to hire as many team members, and those that are brought on would not be as widely distributed, according to Sid. Ultimately, this approach would cut away at GitLab's ability to produce as well as be geographically diverse, he argues.\n\n\"If we pay everyone the San Francisco wage for their respective roles, our compensation costs would increase greatly, and we would be forced to hire a lot fewer people. Then we wouldn’t be able to produce as much as we would like,\" Sid explains. \"And if we started paying everyone the lowest rate possible, we would not be able to retain the people we want to keep.\n\n\"So you end up in a place where the compensation is somewhere in between. And that would cause us to have a concentration of team members in low-wage regions because it’s a better deal for them. They’re getting more than the market rate, so they’re more likely to apply and accept an offer. And they’re more likely to stay regardless of how happy they are, which is not healthy for them or the company.\"\n\n### Standard pay for all roles may not be as fair as it seems\n\nAnother problem with paying everyone the same salary, Sid says, comes down to how far a dollar goes in one place compared to another. If everyone is paid a standard salary, those who live in high-income areas would have less discretionary income when compared to their counterparts in lower-income communities. \n\nRemote companies using a standard pay structure are reportedly running into problems with their compensation plans. \n\n\"The most recent company I talked to has everybody getting paid the same, no matter where they're located. It's very, very different from GitLab, and it is causing problems for them,\" says Barbie. \"We have very strong communication with that company. They're hoping that we can help influence them to move away from paying everyone the same no matter their location. They're finding that it's extremely inequitable.\"\n\n### Closing the gap on local rates for distributed workers\n\nAs remote, or distributed, workplaces continue to take hold and grow across all industries, Sid hopes the location-based compensation gaps will narrow.\n\n\"I think the core difference is there’s people saying, 'Same work, same pay.' And there’s people like us saying we should be at market,\" Sid says. \"I hope the distance between those stances becomes smaller as more companies offer remote work opportunities. I think that’s the way to fix it – just make sure the market rates become higher and consistent.\n\n\"And that’s why we will be promoting remote work a lot. We have a great [page in our handbook about running an all-remote company](/company/culture/all-remote/). Hopefully, that is the way we will contribute to having people across the world get paid the same wages. We will track with that trend; but we won't be ahead of it or behind it. If you see what remote work is doing in a country like the Ukraine, it’s a great source of income for the people there. And I want to contribute to that.\"\n\nStill have questions or thoughts on GitLab's compensation structure? Sound off in the comments below (or on HN, inevitably 😁) or ping us on Twitter [@gitlab](https://twitter.com/gitlab).\n\n[Cover image](https://unsplash.com/photos/uCMKx2H1Y38) by [AbsolutVision](https://unsplash.com/@freegraphictoday) on Unsplash\n{: .note}\n",[892,705,706,9],{"slug":1221,"featured":6,"template":685},"why-we-pay-local-rates","content:en-us:blog:why-we-pay-local-rates.yml","Why We Pay Local Rates","en-us/blog/why-we-pay-local-rates.yml","en-us/blog/why-we-pay-local-rates",{"_path":1227,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":1228,"content":1234,"config":1240,"_id":1242,"_type":13,"title":1243,"_source":15,"_file":1244,"_stem":1245,"_extension":18},"/en-us/blog/without-a-shadow-of-a-doubt",{"title":1229,"description":1230,"ogTitle":1229,"ogDescription":1230,"noIndex":6,"ogImage":1231,"ogUrl":1232,"ogSiteName":669,"ogType":670,"canonicalUrls":1232,"schema":1233},"Without a shadow of a doubt: Inside GitLab's CEO shadow program","Technical marketing manager Tye Davis did everything from joining investor meetings to battling with the flight simulator at GitLab Mission Control.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749680653/Blog/Hero%20Images/sfbaybridge.jpg","https://about.gitlab.com/blog/without-a-shadow-of-a-doubt","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"Without a shadow of a doubt: Inside GitLab's CEO shadow program\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"Tye Davis\"}],\n        \"datePublished\": \"2019-07-11\",\n      }",{"title":1229,"description":1230,"authors":1235,"heroImage":1231,"date":1237,"body":1238,"category":298,"tags":1239},[1236],"Tye Davis","2019-07-11","\n\nWalking up to the iconic Millennium tower on Monday, [I](/company/team/#TyeD19) was a bit nervous for my first day of the [GitLab CEO shadow program](/handbook/ceo/shadow/). Sometimes, our impression of the CEO is someone who is intimidating and strictly business; they only care about things work related. That persona often results from not having access to the CEO, and the fear that one mistake in their presence may cost your job. The GitLab CEO shadow program proved to be a pleasant departure from this mindset.\n\nEntering GitLab “Mission Control,” I was met with a large apartment turned into a hybrid boardroom with a touch of living space. This is a unique working environment because GitLab is an [all-remote](/company/culture/all-remote/) company that allows GitLab team members to work from their choice of location (home, coffee shop, [van](/blog/how-remote-work-at-gitlab-enables-location-independence/), shared workspaces, surfboard, etc.). So, although you are physically at \"Mission Control,\" most of the CEO shadow program is done via video conferencing. There is no need to go from physical meeting room to meeting room, you simply go from conference call to conference call (woo efficiency!). Six monitors add to the office-like feel of the living room, displaying (amazing) sales data, locations of team members in over 50+ countries, and the DevOps toolchain landscape that GitLab replaces. The boardroom also offers access to gaming systems, a [flight simulator](https://en.wikipedia.org/wiki/X-Plane_(simulator)) and readily available drinks and snacks. I was able to calm my excitement and I settle into the room with the fellow CEO shadow, [Mayank](/company/team/#mayanktahil).\n\n![Mission Control center](https://about.gitlab.com/images/blogimages/ceoshadow_graphs.jpg){: .shadow.medium.center}\nInside GitLab's \"Mission Control.\"\n{: .note.text-center}\n\n### Hitting the ground running\n\nMy first face-to-face meeting with CEO [Sid Sijbrandij](/company/team/#sytses) was on our first CEO-specific call of the day, a public live stream on \"Sid's three biggest remote work challenges\" with [Leo Widrich](https://twitter.com/leowid), co-founder of Buffer. This was the first ice breaker into the CEO shadow program and helped me understand just how inclusive the shadow program is. Sid really made us feel like we belonged on the call by incorporating us into the discussion. His inclusivity lowered my stress a few notches, and I began to understand what was to come in the next few weeks: [transparency](https://handbook.gitlab.com/handbook/values/#transparency).\n\nThe second meeting took the inclusivity of the program a step further, as we joined a group call with the executive team from across the GitLab organization (aka the [E-group](/company/team/structure/#e-group)). You might expect some hesitation in allowing someone who is not an executive to join a meeting where top-level matters are discussed, but the CEO shadow program was made exactly for these types of experiences. The program gives participants full visibility into every working part of building an enterprise company. There was no resistance from the E-group team and upon joining the meeting, I was met with an overwhelming ‘welcome’ to our working session. This alleviated most of my nervousness and truly showcased GitLab’s [collaboration value](https://handbook.gitlab.com/handbook/values/#collaboration) by displaying ‘no ego’ and ‘kindness’ from the executive team.\n\n### The feeling of welcomeness was constant\n\n There were very few circumstances where Mayank and I were not included in meetings due to the sensitivity of the subject. The most eye-opening experience for me was meeting with potential investors in GitLab that represent some of the largest and best-known investment firms in the world. These organizations discussed topics around GitLab’s vision and technology and the firms said they see the incredible upside of GitLab. If I was only able to attend one meeting during the whole program, I would choose this one. My confidence in the direction this company is taking has increased after seeing firsthand how much interest there is from investors in GitLab’s growth. Observing the amount of planning leading up to these meetings between Sid and [Paul, our CFO](/company/team/#pmachle) was a great learning experience. Investors are excited about the future of GitLab as a result of all of the hard work of every GitLab team member. My only disappointment is that the program is only two weeks long and that I won’t get to continue to be part of these developing relationships.\n\n![Shadowing the CEO](https://about.gitlab.com/images/blogimages/tyeshadowingceo.jpg){: .shadow.small.center}\nDoing my best impression of shadowing the CEO's activity in his everyday engagements\n{: .note.text-center}\n\n### Takeaways\n\nThe shadow program was an incredibly enlightening experience. Joining this program gave me an accurate and deeply intuitive understanding of the life of a CEO. Sid has mastered the high velocity of responsibility and full situational awareness that is needed to effectively lead our company as CEO. He also acknowledges that he always has room for improvement – so much so that he has a section of flaws that are listed on the GitLab [CEO handbook page](/handbook/ceo/#flaws). One big takeaway from the shadow program is listed on the CEO page. This is something I believe is the biggest factor to collaborate effectively is what Sid notes about his approach, “Not a flaw but something to know about me, I have [strong opinions weakly held](https://blog.codinghorror.com/strong-opinions-weakly-held/). Or, as someone said, I come in hot but am open to new evidence.” This is applicable across the company (and personally) as we all [iteratively](https://handbook.gitlab.com/handbook/values/#iteration) build a tool that best fits our customer needs, and we must be receptive of adjusting accordingly if new evidence corrects our product vision.\n\nBusiness aside, Sid has some pretty funny GitLab stories. If you ever get the chance to ask him about Burning Man, I promise it’ll be a good laugh. My time in the CEO shadow program was unique, educational, and inspirational. I am thankful for this opportunity and hope that one day I’ll reciprocate in a future exec role. Extra shout out to [Cheri](/company/team/#cheriholmes) who coordinates diligently so that all of us CEO shadows are set up for success. Looking back, the most stressful part of the CEO shadow program was the anxiety the X-Plane flight simulator brought when trying to land an airplane (the landing didn't go well).\n\nPhoto by [Landry Gapangwa](https://unsplash.com/@gapangwa91?utm_source=unsplash&utm_medium=referral&utm_content=creditCopyText) on [Unsplash](https://unsplash.com/@gapangwa91?utm_source=unsplash&utm_medium=referral&utm_content=creditCopyText)\n{: .note}\n",[705,9,706],{"slug":1241,"featured":6,"template":685},"without-a-shadow-of-a-doubt","content:en-us:blog:without-a-shadow-of-a-doubt.yml","Without A Shadow Of A Doubt","en-us/blog/without-a-shadow-of-a-doubt.yml","en-us/blog/without-a-shadow-of-a-doubt",{"_path":1247,"_dir":243,"_draft":6,"_partial":6,"_locale":7,"seo":1248,"content":1254,"config":1259,"_id":1261,"_type":13,"title":1262,"_source":15,"_file":1263,"_stem":1264,"_extension":18},"/en-us/blog/a-look-at-devops-salaries",{"title":1249,"description":1250,"ogTitle":1249,"ogDescription":1250,"noIndex":6,"ogImage":1251,"ogUrl":1252,"ogSiteName":669,"ogType":670,"canonicalUrls":1252,"schema":1253},"DevOps salaries in 2021: where do you rank?","Another surprise benefit of working on a DevOps platform? A higher salary! Here's why DevOps salaries are going up, and where to find the biggest paychecks.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749663397/Blog/Hero%20Images/logoforblogpost.jpg","https://about.gitlab.com/blog/a-look-at-devops-salaries","\n                        {\n        \"@context\": \"https://schema.org\",\n        \"@type\": \"Article\",\n        \"headline\": \"DevOps salaries in 2021: where do you rank?\",\n        \"author\": [{\"@type\":\"Person\",\"name\":\"Sharon Gaudin\"}],\n        \"datePublished\": \"2021-10-07\",\n      }",{"title":1249,"description":1250,"authors":1255,"heroImage":1251,"date":1256,"body":1257,"category":912,"tags":1258},[909],"2021-10-07","\n_This is the first in an occasional series of blog posts looking at DevOps salaries and careers._\n\nDespite the COVID-19 pandemic and the subsequent economic crisis that has disrupted lives and business across industries and around the world, demand for DevOps professionals remains strong and salaries continue to increase.\n\nThe IT industry, in general, fared better than many during the economic uncertainty of 2020 and 2021. With a strong IT infrastructure already in place and IT professionals accustomed to working remotely, increased demand and short supply for IT workers meant [IT salaries held steady or rose](https://rlc.randstadusa.com/for-business/learning-center/salary-insights/salary-guide/IT-technologies) in turbulent times, reported Randstad, a multinational human resources consulting firm. \n\nAnd DevOps professionals did even better than most in IT.\n\n## DevOps salaries are on the rise\n\n[DevOps, simply put,](/topics/devops/) is one of the hottest areas in the technology industry. Robert Half International Inc., a major human resources consulting firm, lists DevOps in the top 10 most in-demand jobs in 2021. Actually, in early September the firm [listed DevOps as the second hottest IT job](https://www.roberthalf.com/blog/salaries-and-skills/the-13-highest-paying-it-jobs-in-2019), just behind big data engineers, and surpassing cloud architects, security managers and database managers. And Randstad also ranked DevOps high in its [list of in-demand technology roles.](\u003Chttps://rlc.randstadusa.com/for-business/learning-center/salary-insights/salary-guide/IT-technologies>)\n\nAccording to salary watchers like Randstad, Glassdoor and ZipRecruiter, DevOps engineers, for instance, generally make approximately $100,000 to $150,000. Based on average U.S. salaries on Glassdoor, DevOps engineers are number 8 for 10 top-paying IT jobs in 2021. And DevOps developers, who Randstad calls one of the most in-demand technology roles, are doing well, too. They are in line to make $112,785 (for those with one year of experience) to $165,980 (for 10 years or more of experience). Those figures, of course, greatly depend on location and skill level.\n\nJust to drive the point home, Amanda Stansell, a data scientist at Glassdoor, said in a report earlier this year that [DevOps engineers](https://about.gitlab.com/topics/devops/what-is-a-devops-engineer/) made her list of [Top 10 Best Jobs in America for 2021.](\u003Chttps://www.glassdoor.com/research/best-jobs-in-america-for-2021/>) She bases her calculations on earning potential, overall job satisfaction, and number of job openings listed on Glassdoor. The role of DevOps engineer came in squarely in the middle at #5 - between Java developer at #1 and dentist at #10.\n\n## Demand for DevOps professionals is skyrocketing\n\nAccording to the Randstad 2021 Salary Guide, demand for DevOps developers is currently skyrocketing. “For employers… the average time-to-fill for these roles is north of 50 days,” the report noted. “That likely means many would-be employers today are instead suffering from key vacancies in their IT departments. Worse, with average annual salaries for DevOps developers at $137,830 — higher even than the average for cloud engineers — organizations should expect to spend considerable budget just to be in the running for skilled and experienced developers. It’s a pay-to-play hiring environment.”\n\n## Breaking down the DevOps salaries\n\nLet’s take a closer look at how pay for some DevOps positions breaks down:\n\n* ZipRecruiter reports that a DevOps engineer in San Francisco can take home $132,934, while the same position in Boston, Mass. would garner $113,552. In Austin, Texas, that engineer could earn $110,240 but in Boise, Idaho that drops to $102,093.\n\n* According to ZipRecruiter, the top five [highest paying cities for DevOps engineers](https://www.ziprecruiter.com/Salaries/Devops-Engineer-Salary) are Sunnyvale, Calif. (at $144,494); Santa Rosa, Calif. ($139,673); Cambridge, Mass. ($135,440); Vacaville, Calif. ($132,838), and New York City ($131,356).\n\n* The [top five best states](https://www.ziprecruiter.com/Salaries/What-Is-the-Average-Devops-Engineer-Salary-by-State) for DevOps engineers to earn the most are Massachusetts, Hawaii, Connecticut, Tennessee and Minnesota. \n\n## Read more on DevOps careers: \t\t\n\n- [Best advice for your DevOps career? Keep on learning](/blog/best-advice-for-your-devops-career-keep-on-learning/)\n\n- [6 tips to make software developer hiring easier](/blog/6-tips-to-make-software-developer-hiring-easier/)\n\n- [Four tips to increase your DevOps salary](/blog/four-tips-to-increase-your-devops-salary/)\n\n- [Have DevOps jobs to fill? Try these 3 strategies to hire and retain](/blog/have-devops-jobs-to-fill-try-these-3-strategies-to-hire-and-retain/)\n",[679,892,9],{"slug":1260,"featured":6,"template":685},"a-look-at-devops-salaries","content:en-us:blog:a-look-at-devops-salaries.yml","A Look At Devops Salaries","en-us/blog/a-look-at-devops-salaries.yml","en-us/blog/a-look-at-devops-salaries",4,[662,690,713,734,755,776,798,818,838],1753475285957]