Sid Sijbrandij on GitLab’s Remote Work, Feedback and Boards

Sid Sijbrandij says he saw promise in the open-source GitLab project created by his co-founder Dmitriy and believed it needed full-time support beyond volunteer contributions. He describes GitLab as a platform covering several stages of software development and delivery.

Sijbrandij says GitLab's remote model developed organically: he was in the Netherlands, an early hire was in Serbia and his co-founder was in Ukraine. Even nearby employees sometimes worked with him briefly in person and then worked from elsewhere.

The company initially planned to reconsider remote work if it stopped scaling. Sijbrandij instead argues it can work better at larger scale, though he acknowledges that a small team may benefit from being in the same room when practical. This is his management view, not a universal finding.

He says distributed teams must intentionally create informal interaction. GitLab normalizes optional coffee chats, including several during onboarding, so colleagues can talk without a work agenda; participation should not be compulsory for everyone.

He also values weekly team social time and occasional in-person gatherings. In his account, remote work and meeting in person are compatible; people's interest in social activities varies with their preferences and life circumstances.

Sijbrandij argues that companies should spend less time debating a complete return to offices and more time making distributed work effective. He cites low office attendance as part of his argument, but the interview provides no source or representative sample for that generalization.

He predicts growth in fully remote startups and distinguishes hybrid schedules where people coordinate office days from arrangements with some permanently remote colleagues. He worries the latter can create unequal access and prompt remote employees to leave; this was his 2022 prediction.

GitLab's own challenge, he says, is working across time zones. Although the company had people in more than 60 countries, his direct executive reports were all in the US; he wanted more asynchronous work and greater geographic diversity in leadership.

At GitLab, he says, meetings should not be used to present material that can be read beforehand; synchronous time is better reserved for questions. He gives the example of replacing a preparation call for this interview with written notes, while acknowledging that calls still occupy much of his day.

Sijbrandij names results, iteration and transparency as values he tries to embody. Over time he says he has learned to delegate, be more patient and give people enough context to make decisions themselves.

Stebbings asks whether transparency has limits, citing salaries, acquisitions and fundraising. Sijbrandij says GitLab is transparent by default but maintains exceptions including compensation and M&A; it shared fundraising updates internally when conditions were favorable, while acknowledging that choice can be difficult.

For feedback, he stresses thanking the messenger and avoiding punishment. He separates welcoming feedback from deciding whether to accept it, weighing factors such as how often he hears it, the source and that person's relevant expertise.

Public responses to feedback can show others whether it is safe to speak. Sijbrandij recalls recognizing that he had delivered a suggestion poorly and apologizing for the manner, using self-criticism to make it easier for colleagues to challenge him.

Stebbings asks how to manage anger when someone seems not to care. Sijbrandij advises against assuming intentions or labeling a colleague lazy: their circumstances or an overly high bar may be part of the explanation. He describes adapting leadership style to the person and situation.

GitLab's written handbook, which he says exceeded 2,000 pages, reduces the need to interrupt colleagues for routine answers. The drawback is that writing enough context can take time; for an urgent, important issue involving people with different context, he says a synchronous conversation may be better.

On mentorship, he suggests asking for help with a specific problem rather than immediately requesting an open-ended mentor relationship. Explain why the problem matters, what you have tried and why the person is well placed to help, then request a short conversation.

He credits CEO coaches and board members with helping him learn, while distinguishing directors from mentors because they also have governance authority. He recalls a coach helping him prepare for a quarter that appeared likely to miss its target by about 30%.

Instead of focusing on his fear of being fired, the coach urged him to understand the problem and prepare a plan. Sijbrandij says he opened the board meeting by explaining what was wrong and what the company would do, and a director cautioned against taking bad deals merely to meet the quarter.

For fast-growing CEOs, he recommends finding a coach with experience at the company's stage and treating selection as a mutual process. He says the value may eventually level off as the same advice repeats, at which point a different coach can provide new perspective.

Sijbrandij serves on the board of Meltano, a GitLab spinout in which GitLab has a significant ownership interest. His board perspective therefore comes with an explicit commercial relationship between the two companies.

At GitLab, he says, board members receive presentations in advance, including video clips, so meeting time can focus on discussion. He recommends telling directors the specific issues on which the CEO wants help rather than inviting unstructured advice on everything.

As a director himself, he tries not to take over the operator's role. He favors informed questions, concise contributions and complementary areas of expertise; he says a board member should challenge assumptions without directing the company as though they were its executive.

Stebbings asks whether a difficult strategic concern should be raised privately or in the meeting. Sijbrandij says a company strategy question belongs in the board discussion, where others can contribute, while directors should challenge evidence through questions rather than dictate a replacement plan.

He describes GitLab's then-current board format as roughly 90 minutes of open session and 45 minutes closed, with part of the closed session involving the CEO and part without the CEO. A brief private discussion before the meeting also establishes priorities and recognizes strong contributors.

When selecting investors, he looks at the firm's reputation and the individual who may join the board. He recalls discussing company strategy with a prospective Series B investor to see how the person contributed, rather than judging the relationship solely by the firm's name.

In the quick-fire round, he recommends High Output Management and says cross-functional initiatives are difficult in a company organized by function. He jokes that project structures complicate career development and matrix structures create their own complexity.

Sijbrandij argues that assembling many separate DevOps tools and maintaining custom integrations is becoming less attractive. This is the GitLab CEO's competitive view in 2022 and aligns with the integrated platform his company sells.

He describes co-founding Open Core Ventures to create companies around open-source projects. He says the venture typically supplies $2 million, recruits a technical leader and team, develops a paid offering and later recruits a CEO; he states it generally holds a majority before the first external round.

As a marketing example, he cites a GitLab billboard and a customer claim that T-Mobile increased deployment frequency tenfold. The interview does not provide the measurement method or independently verify the claim.

Asked about his own weakness, he says he could be more positive in meetings. He uses CEO shadows who attend and rate many meetings to help him observe and improve his behavior.

Looking ahead, he cites outside market forecasts and argues integrated DevOps platforms will expand their role. These are 2022 projections and his commercial assessment, not evidence of future market share or guaranteed outcomes.

Stebbings thanks Sijbrandij and closes the interview.