Jason Citron on Discord’s Early Growth and Rethinking Delegation

Discord co-founder and CEO Jason Citron says scaling from 200 to 1,000 employees forced him to reconsider how he delegated product decisions. In this November 2024 conversation with Harry Stebbings, he reflects on the product’s origins, executive hiring, fundraising and the future of games. His account of Discord’s strategy reflects his role at the company.

Citron sees games as small simulations in which players solve problems, fail, learn, cooperate and take turns leading. He suspects that repeated practice may translate to company building, while acknowledging that this is an interpretation rather than proof that gaming causes founder success.

Every new stage of a company presents challenges a CEO has not seen before. Citron says he often does not know the answer at first, but trusts himself and his team to work it out through effort and learning.

Selling his prior company, OpenFeint, for $104 million gave him a financial cushion. He says it made it easier to turn down acquisition offers and pursue a larger ambition. Previous founder experience also made early hiring and management more familiar, though finding product-market fit remained hard.

His original thesis was that a durable gaming company needed a distribution advantage, potentially a group chat product for players and developers. The team first tried to build a multiplayer mobile game, Fates Forever, to bootstrap that network.

The game developed a following but did not reach the scale needed for the plan, Citron says. Building it nevertheless helped the team identify a way to launch a group chat app independently. Discord began development in 2015: the vision stayed, while the implementation changed.

The team rebuilt the voice engine several times. Within a few months it had about 20 daily active users outside the company, but still struggled to persuade other people to try another gaming chat app. Citron rejects the assumption that a good product automatically finds its audience.

The initial distribution breakthrough came when a friend posted about the product on a Final Fantasy XIV subreddit and the founders joined a Discord server to receive feedback. Citron recalls roughly 50 users arriving that day and more the next, after which the team repeated the approach for several months.

When growth slowed, the company worked on both existing users’ needs and accessibility for new users. During the pandemic it expanded into more uses, but later found that some features did not resonate as hoped and refocused on gaming in 2024.

Citron does not regret exploring beyond games: it introduced Discord to people who continued using it, even if some product efforts had less business impact than he wanted. The period also brought organizational growth from roughly 200 to 1,000 employees in about two and a half years.

He says he made some of his biggest management mistakes during that expansion. He hired executives, set a high-level vision and delegated the work in a way he now believes disconnected product decisions from a coherent customer experience.

Citron is skeptical of management language such as “alignment” when it substitutes for decisions and accountability. The warning sign for him was that products were not shipping coherently and getting things done felt harder. This is his experience, not a claim that delegation itself is always wrong.

Delegation remains important, he says, but the level at which a decision is made matters. In annual planning he now identifies problem areas and selects projects, while teams explore solutions, challenge his ideas and own the detailed work.

Teams can disagree with Citron and propose alternatives, but he wants the company-level choice of which problem to tackle made at the right level. He regards his earlier mistake as delegating some of those choices too far down the organization.

He has seven direct reports but also reviews work from directors and individual contributors. They record walkthroughs of features; he watches them, reads supporting material and often replies with a video. Meetings are reserved for questions or disagreements that need them.

The asynchronous process reduces large product-review meetings and leaves a public archive of feedback inside the company. Citron accepts that it reduces some discussion by design: he wants focused feedback rather than a ceremony in which every participant feels obliged to speak.

He still values live discussions for open-ended brainstorming or when people need to work through genuine differences. The asynchronous exchange can give everyone context before a shorter debate.

Asked whether a CEO should inject chaos, Citron says his job is to make sure the company creates great products. Chaos is useful only if it helps that goal. He considers selecting the right problems at least as important as allocating resources, because a wrong problem cannot be fixed by staffing it well.

Executive interviews can show whether people communicate and debate well, but he says they are less reliable for determining how they actually work. He now makes extensive reference calls to people who have worked with candidates over time.

For new executives, he describes a gradual onboarding process: learn the people, context and projects before making large changes. He meets them frequently early on, expects a developing point of view in the second month and a transition to leading the team by the end of the third.

His concern is that an executive who changes things immediately without context can lose credibility if a decision goes badly. During Discord’s hypergrowth, he sometimes accepted an expert’s approach even when his own intuition disagreed; he says those decisions often did not work.

He now gives more weight to his company context while recognizing the executive’s craft expertise. Citron also admits that his intuition is imperfect: he supported a PC game store in 2018 that he later shut down. The lesson, he says, is to learn where one’s judgment is more and less reliable.

Stebbings asks about an opportunity to sell Discord for a reported $12 billion. Citron describes a lengthy discussion and acknowledges that Microsoft was a valuable partner, but says the company chose to remain independent and pursue the product opportunity. The interview does not establish the final terms of a completed transaction.

He expects games to become more pervasive over the next decade despite a difficult post-pandemic period. He thinks AI could reduce content-creation costs and enable smaller teams or individuals to make more ambitious games. These are forecasts, not demonstrated economics for every game.

Stebbings asks whether a flood of new games would make discovery harder. Citron agrees discovery is already difficult and may become harder, but thinks lower production costs could also make smaller, niche audiences economically viable.

He likens AI’s future to a game’s fog of war: there are many possible paths, and confidence about the outcome is hard to justify. He imagines AI-generated characters and stories that continue to respond to a player’s choices, while acknowledging that the technology and economics are uncertain.

Such a game might use subscriptions or microtransactions, he speculates, but the business model could depend on whether the computation runs on a device or a server. These are possibilities rather than announced Discord products.

Citron says each leader should decide how much politics to discuss at work. He then recalls a first meeting with investor Danny Rimer, in which he asked detailed questions about a failed investment because he wanted to learn from someone else’s experience.

He says Discord has raised almost $1 billion. The Series A for the earlier game was difficult because investors doubted it could bootstrap the larger vision; Mitch Lasky at Benchmark supported the team. Citron chose that relationship over a slightly higher valuation elsewhere.

His fundraising lesson is that the individual investor matters, not just the firm’s name. Once Discord had strong growth and consumer enthusiasm, he found later fundraising easier; that is his account of the company’s experience.

In 2018 he met with Rockstar Games while exploring a PC game store. He concluded that competing for major titles would require costly minimum guarantees, particularly as other stores were launching, and decided the economics were not attractive.

Discord closed the store around 45 days after that meeting, soon after raising money for the initiative. Citron says he offered to return the investment; the investor encouraged him to keep going. The decision illustrates his willingness to revise a plan, although he had originally backed it.

On Discord’s future, he wants to keep building useful products with his team. Employees and investors will eventually need liquidity, he says, but he does not announce a timing or method. Being a public-company CEO is interesting to him, not the reason he started the business.

For consumer founders, his unconventional advice is to build for a problem they understand personally. He worries about civic institutions and education, and would make schooling more focused on engaging, hands-on problem solving and critical thinking.

He discloses that he is on the board of Campus, an online education company, and describes its approach. He thinks university is expensive, but says choosing education versus startup work depends on the individual rather than a rule for everyone.

Citron says he has already changed much of how Discord operates, including choosing projects more directly and reducing recurring meetings. He now looks forward to reviewing proposals from teams rather than relying on the earlier process he found frustrating.

Asked about underappreciated technology, he favors more nuclear power and believes its risks are often overstated. That is his policy view; the interview does not provide a full comparison of nuclear and fossil-fuel safety or economics.