What this covers

This is a conversation between host Patrick and Jason Citron, who traces the path that led to Discord. The earlier ventures get as much attention as Discord itself. Citron describes testing demand for OpenFeint with a landing page of faked Photoshop screenshots and a TechCrunch mention, which produced a couple hundred developer sign-ups and funded an MVP. He explains why Fates Forever missed: the team picked the wrong platform, iPad, and arrived somewhat early, even though the hunch about mobile was right. He also covers why cofounder Stan pushed to start on PC with a browser-based voice app that anyone could join by clicking a link.

The middle of the conversation covers early traction and purpose. A friend's post in the Final Fantasy XIV subreddit, answered with a link to a live server where the team was present, brought in a couple hundred sign-ups in a day. Applying Five Whys to the original mission, both cofounders landed independently on belonging and rewrote it. Citron also recounts the end-of-2018 game store, launched with about 1,000 games, and the exit from games when users wanted better communication and Microsoft and Epic Games were turning exclusives into a spending contest.

The last stretch is about building the company. OpenFeint's unplanned culture convinced him culture must be intentional. At about 20 employees Discord froze hiring for three months to train managers, a choice VCs thought was crazy. He screens for humility, curiosity, and compassion, including a lunch to watch how candidates treat waiters. He links slow hiring to N-squared communication costs and small and mighty teams, and warns against over-focusing on competitors.

Sharpest takeaway

Citron argues Discord became a digital third place because its founders built for belonging and low-friction hanging out, validated product bets through fast signal-seeking experiments, and chose premium features over advertising so that monetization stayed aligned with users.

  • Discord's third-place feel was an intentional design goal for low-friction, low-commitment togetherness, though the founders only named it later after reading Oldenburg
  • Hypothesis-driven, minimum-signal experimentation drove product decisions, and earlier games failed partly because feedback came too late
  • Paid Nitro features were chosen over ads so the business would not degrade the free user experience

The argument · threads9 threads · 66 claims
0.82

Discord's value lies in third-place belonging and community, not in games.

10 pointscentrality 5/5
  • Discord's always-on voice channels within invite-only servers let friends see who is present and drop in and out, creating a conference-call-like experience that made voice chat while gaming far easier than existing tools.

    “So it was kind of an always-on conference call experience.”

  • A 2019 survey found about 30 percent of respondents said they did not primarily use Discord for gaming, which forced the company to consider a much bigger mission than games.

    “people self-reported that actually about 30%, it was 30% of the people who took our survey self-reported that they did not primarily use Discord for gaming.”

  • Discord launched a PC game store at the end of 2018 with about 1,000 games, but users wanted a better communication service, a signal that the team had overemphasized games in its mental model, and the company then pivoted away from it.

    “We launched a video game store on PC at the end of 2018 with about 1,000 games. And very quickly what we learned was that people actually just wanted us to build a better communication service.”

  • Bringing people together around games was ultimately about meeting a deep social need for belonging, which Citron links to the third tier of Maslow's hierarchy.

    “Maslow's hierarchy, belonging, I think, is the third one.”

  • Three features combine to create a third-place experience: an invite-only space containing only chosen people, easy drop-in hanging out, and server-level permissions and roles that let communities set their own rules and norms, like hosts, regulars, and VIP areas.

    “these three things: Have an invite-only space, Easy to hang out, and You can control the rules and norms, really creates this experience of having a place that you can go to.”

  • People judge a place as cool when it is full of people, so Discord's presence indicators and always-on voice channels make the service feel alive, which is the busy restaurant phenomenon.

    “it gives you this sort of, I call it the busy restaurant phenomenon. You look at a restaurant and it's full of people. You're like, "That place is cool, it's fun, it's alive."”

  • Go Live recreates sitting on a couch next to a friend by streaming games with ultra-low latency to up to 50 people, with free 720p at 30 FPS and paid 1080p at 60 FPS or source quality.

    “For free, you can do it at 720p, 30FPS. And if you pay, you can do 1080, 60 FPS or go all the way to source.”

  • On Discord, text and voice usage are roughly equal, with video as an emerging mode, because each medium requires a different level of attention and energy that users select according to the moment.

    “It turns out, text and voice are about even, and video is kind of an up and comer.”

  • Discord's roles, permissions, channels, and invite-only spaces replicate how real-world gathering places organize hosts, regulars, and entry areas, giving users a sense that they must behave.

    “these three things: Have an invite-only space, Easy to hang out, and You can control the rules and norms, really creates this experience of having a place that you can go to.”

  • Citron read Ray Oldenburg's 1980s book The Great Good Place in 2019, finding that it described the third-place concept that Discord had built in digital form.

    “I read the book, The Great Good Place, by Ray Oldenburg, which was written in the '80s”

0.71

Product success depends on hypothesis-driven loops, early platform timing, and disciplined fast feedback over overreaching.

11 pointscentrality 4/5
  • Unlike individual entertainment content that loses novelty, the need for more games is durable, so a game distribution mechanism could serve as a durable core for a business, though this shift only became clearer with games as a service in the last decade.

    “what's durable about entertainment is the need for more of it. Whereas most pieces of entertainment content have novelty”

  • In 2012 Citron's plan was that long-form cooperative multiplayer games would move to tablets, which would create a need for a group communication service, and that the communication service would later anchor a distribution business for games.

    “build a competitive multiplayer game with team dynamics to bootstrap a group communication service. And then once we amass scale, build a business by distributing games on top of making our own.”

  • Game designs are reflections of how human brains work, so new platforms tend to produce popular reinterpretations of existing genres, and genuinely new fundamental game mechanics are rare.

    “game designs are actually sort of a reflection of the way that human brains work. There aren't that many new genres”

  • Citron approaches product development as a scientific process: form a hypothesis, build the minimum needed to get a real signal, learn from it, and repeat, running this loop as a flywheel.

    “I treat it almost like science. You express art through the framework of the scientific process, which is like, you have a hypothesis, you come up with the minimum amount of things you can do to actually get a real signal back on that hypothesis.”

  • Fates Forever was not a hit because the team picked iPad, which was the wrong platform, and was somewhat early, since mobile phones later became where people played these games constantly, even though the hunch about the market direction was right.

    “the game ultimately wasn't a hit because I think we picked the wrong platform. We picked iPad and I also think we were a little bit early.”

  • A product aiming for massive scale needs very high early usage frequency and longevity, because as it grows it attracts less passionate users, so the most rabidly obsessed early users are the best indicator of eventual success.

    “if you're trying to build something that's going to be at scale, frequency of usage and longevity of usage of a product in the early days needs to be astronomically high”

  • Citron's biggest product mistakes came from going too long before getting signal back on a hypothesis, because the test structure required too much work to yield feedback.

    “The biggest mistakes that I've made building this company has been where we went too long before getting signal back”

  • Stan proposed starting Discord on PC, where players were already using competitive multiplayer apps, rather than bootstrapping through a mobile game, so the team could be ready when mobile took off for the next billion users.

    “if we actually started with PC where people were already playing competitive multiplayer games a lot, including us, that we could build a better version of the apps”

  • Word of mouth starts slowly and requires people who are excited enough to share a product, so founders must generate initial momentum through targeted distribution before organic spread can happen.

    “word of mouth starts slowly when there's no mouths to have words. You need people to be excited to share it, to generate momentum.”

  • Competition for exclusive games between Microsoft and Epic was becoming a contest over who could spend the most on developers, which was not a prudent fight for an unscaled startup, so Discord exited the games business.

    “it was very clear that it was going to become a battle of who can throw the most money at game developers to get exclusive content.”

  • Companies should pay attention to competition but not focus too heavily on it, because people and businesses move toward whatever they focus on, so over-focus on rivals leads to copying them rather than serving customer needs.

    “I think it's human behavior that you will go where you focus. So if you focus too much on your competition, you just end up copying them”

0.66

Founders must plan distribution, reaching users where game communities already gather.

9 pointscentrality 4/5
  • The first games failed because building something people want is very hard; one shut down on launch day, another never shipped, and the one with traction earned only about $30K and could not fund the team.

    “What I learned really quickly from that is it's really hard to make something that people want. I went through probably four projects”

  • Console launch windows give early titles little competition and distribution, so the App Store's launch should work the same way, letting a studio that ships on day one establish itself on the new platform.

    “if you can be a launch title, you're going to make money. You're going to be able to get distribution because there's not a lot of stuff to pick from.”

  • The founders tested demand for OpenFeint before building it by creating a landing page with fake Photoshop screenshots and getting TechCrunch coverage, which produced a couple hundred developer sign-ups that funded a bridge round and an MVP.

    “I made fake screenshots in Photoshop of what the system would look like. And we convinced TechCrunch to cover it. We got a couple 100 developers to sign up”

  • Building something people want is not sufficient; founders must also plan how to cut through noise and get the word out, since distribution determines whether an early product gets seen.

    “So what I really learned from that was distribution is really important. And you have to think about how you're going to cut through the noise and not only just build something that people want, but also get the word out.”

  • Early in the tablet game project, the team posted concept art on Reddit asking gaming communities for feedback on the art style, and the feedback led to iterations that improved the final art.

    “We actually did concept art and put it up on Reddit and said, "Hey, we're a gaming studio making mobile, like, what do you think of this, it's art style?"”

  • A friend posted asking about the new voice app in the Final Fantasy XIV subreddit, and the team replied with a link to a live server where they were present, so Reddit visitors could click in, talk to the developers, and register, producing a couple hundred sign-ups in one day.

    “We ended up getting a couple 100 people to register that day.”

  • Asking friends to try Discord produced only about 20 daily active users, so growth at scale required going to where game communities already gathered.

    “we probably had, I don't know, 20 daily active users, 20 people maybe using it. To actually get the word out and grow at scale, we basically went to where people who play games hung out and congregated.”

  • Discord's early users spent an extraordinary amount of time in the app, which Citron treats as the strongest indicator of the product's objective utility.

    “the amount of time that people were spending using Discord in the early days was just totally ridiculous”

  • Stan proposed building Discord on web technologies using WebRTC so users could join by clicking a link without installing anything, then download a native app later if they liked it.

    “you could join it by just sending someone a link, you don't have to install anything”

0.52

Slow hiring, humility, and long-term horizons yield returns for small, mighty teams.

9 pointscentrality 3/5
  • At about 20 employees, Discord froze hiring for three months to build a management training program and selected who wanted to become managers, a choice that VCs thought was crazy but that Citron believed was right for long-term returns.

    “we froze hiring for three months, created a management training program, and asked who wanted to be a manager.”

  • Hiring for mission fit, meaning people with personal experiences connected to the product, leads employees to care more about details and work harder because they care about the change being created.

    “when people really care about the purpose of your mission, they just naturally care more about everything, about the details.”

  • Investing in training and development may have poor returns within a month but produces incredible returns in two, three, or four years, and making people more productive through better tools or making them smarter is central to Discord's small-team approach.

    “investing in your people, while in a month may not have great returns, in two, three or four years have incredible returns.”

  • Because communication scales like an N-squared network problem, hiring slowly keeps coordination costs low, and Discord organized around small and mighty teams of smart people making high-quality, compounding decisions.

    “as you add more people, everything gets harder because you have to communicate more and it's like an N-squared problem because it's a network. So one of the original insights that I had which to this day holds true is if you can hire slower, everything gets easier.”

  • Citron screens for humility, curiosity, open-mindedness, and compassion, and uses a lunch with three non-interviewers to observe how candidates treat waiters, assuming that treatment of people without power predicts how they treat others.

    “we watch for how they treat the waiters. And if they don't respect the waiters... then we assume that they don't respect people who don't have power over them.”

  • Humility is a key hiring criterion for Jason because without it leaders drink their own Kool-Aid and get carried away, and humility is also a precondition for being curious and open-minded.

    “You got to be humble because if you're not humble, I think it's too easy to sort of drink your own Kool-Aid and get carried away with things.”

  • With about 280 employees and over 100 million monthly active users, Discord's employee-to-user ratio is unusually lean for a service of its scale, which Citron attributes to small and mighty teams.

    “I think we're around 280 people now. And given that we have over a hundred million monthly active users”

  • Discord has rejected VP candidates who spoke only to senior people or dismissed junior staff during virtual lunches, treating these as hard disqualifiers.

    “we've rejected a couple of VP candidates because of stuff like that, it's a hard line for us.”

  • Citron argues that enduring companies compound value over time, so decisions should be judged on a 10 to 20 year horizon even when they look suboptimal in the short term.

    “Having a 10, 20-year view on what we're doing, allows us to make decisions that can result in incredible yields in the future.”

0.47

Discord succeeded by making voice communication easy, accessible, and oriented toward belonging.

7 pointscentrality 3/5
  • Discord was built as a better version of the voice and communication tools that Citron and co-founder Stan used when playing online games, because communication tools had barely changed in roughly 15 years.

    “The communication tools that existed actually hadn't changed that much from when we were teenagers to then”

  • Discord's early retention and time-spent metrics were unusually high, indicating that people really liked the product, which contrasted with the game that failed even by standards of good games.

    “the amount of time that people were spending using Discord in the early days was just totally ridiculous because people really, really, really liked it.”

  • Discord rewrote its mission from bringing people together around games to giving people the power to create belonging in their life, because users were using the tool for all kinds of shared experiences and the team wanted to be valuable to everybody.

    “we could just go more fundamental with our mission and expand the aperture to what we're doing to be valuable to literally everybody.”

  • Using the Five Whys technique on the original mission of bringing people together around games, both co-founders independently arrived at belonging and connection with humanity as the root need, which led them to rewrite the mission as giving people the power to create belonging in their life.

    “at around the fourth or fifth why, we both said, "To find belonging in the world," or something to the effect of belonging”

  • Discord did not invent gaming voice chat; predecessors such as TeamSpeak, Mumble, and Roger Wilco already had voice features, and Discord's contribution was to make the experience easy and accessible.

    “This concept though, we didn't actually invent it. There were apps that came before us, TeamSpeak, and Mumble”

  • Discord's third-place character was intentional in the sense that the team deliberately built low-friction, playful, low-commitment communication, though they only recognized it as a third place after reading about the concept years later.

    “The third-place idea, it's funny your question asks if it's intentional because it was intentional, but we didn't know that we were making a third place.”

  • Discord ran inside a web browser using WebRTC, so people could join by clicking a link without installing anything, which removed the friction that held back rival voice apps such as Ventrilo-style IP-address workflows and paid tools.

    “the way we ended up solving this friction issue is by using WebRTC and building the app in web technology frameworks. So we could make the whole application run inside of a web browser so that you could join it by just sending someone a link”

0.36

Culture must be intentionally designed around intrinsic motivation from the start.

3 pointscentrality 3/5
  • At OpenFeint the culture was not deliberately designed, and when an early employee said it sucked to work there, the company was too far along to fix it before its sale, which convinced Citron that culture must be intentional from the start.

    “it's hard to fix culture when you haven't been intentional about it and you have 100 people.”

  • Citron says Dan Pink's book Drive helped codify his ideas on intrinsic motivation, which rest on autonomy, mastery, and purpose, and he adds compassion as a fourth element.

    “it boils down to basically three things, autonomy, mastery, and then purpose, which is the mission fit thing I talked about. And I add compassion to it”

  • The kindest acts are consistent small gestures, such as making coffee, rather than grand ones, and that consistency is what makes relationships and the world work.

    “so much of life is about the little things and about the consistency of the little things that people do.”

0.33

Scaling and selling OpenFeint taught Citron leadership, though the exit left him feeling he had failed.

2 pointscentrality 3/5
  • After OpenFeint was acquired, the acquirer asked Citron to step aside, and he left after three or four months; although the financial outcome was decent, he felt he had failed and cried in his car for 15 to 20 minutes.

    “And I sat in my car and cried for about 15, 20 minutes when that happened because I thought I had failed.”

  • Scaling OpenFeint to about 100 employees, with roughly 7,000 games on the service before its sale, taught Citron how to move from being an engineer to being a CEO.

    “two years later, at 100 employees, there were like 7,000 games in the service and we sold it for a decent chunk. And I really learned how to go from being an engineer to being CEO”

0.27

Monetization should add personalization and paid extras without degrading or excluding free hangout users.

5 pointscentrality 2/5
  • Selling ads would require building ad technology that does not improve the end-user experience and actively subtracts from it, so Discord avoided ads to keep incentives aligned with users.

    “selling ads would require us to spend a tremendous amount of energy investing in building all this ad technology that actually doesn't make the end-user experience directly better, it actually subtracts from the end-user experience.”

  • Discord keeps server-level custom emoji free but charges for using custom emoji in direct messages and for animated emoji, creating a second tier of personalization that users are willing to pay for.

    “What we charge for is if you want to use those custom emojis in your direct messages”

  • Discord's paywall design keeps the free tier fully functional and puts exciting but non-essential extras behind payment, so nobody feels left out and the experience of hanging out is preserved.

    “So we try to make sure that the things that are premium on discord are exciting and fun but don't feel necessary because we don't want people to really feel left out”

  • Discord Nitro launched in 2017 as a $5 subscription for chat perks, initially framed as a donation-like support mechanism, and covered about a third of burn in early 2019.

    “It was a $5 subscription service and it was modestly successful, it was covering I think about a third of our burn rate in early 2019”

  • Server boosting lets members pool payments to unlock community perks like banners and animated icons, which people enjoy because it lets them contribute to their community and earns a visible badge, creating a human dynamic that drives sales.

    “boosting is you can choose to boost the server, depending on how many times the server gets boosted, it unlocks a different set of perks”

0.27

Earlier game ventures failed under fragile market and pricing conditions.

5 pointscentrality 2/5
  • Working for independent game developers was difficult because companies depended on a hit-driven market, which made the business model fragile and led Jason to change studios frequently.

    “you were really dependent on this sort of like hit-driven nature of the industry, that was challenging.”

  • In 2012 Jason bet that long-form cooperative multiplayer games from PC would migrate to tablets, because tablets were accessible and used for extended sessions, a thesis that proved too early for the 2014 game.

    “I thought that perhaps the kinds of long-form cooperative multiplayer games that I love playing on PC would become popular on tablets”

  • Citron's free-to-play game with an $8 premium multiplayer sequel earned only about $30,000 because buyers would not pay $8 for games and coin-pack exploits pushed the market toward $1 pricing, so the model could not fund the team.

    “no one was buying games for eight bucks. ... And we made, I don't know, 30 grand on that game.”

  • Discord's team spent six to eight months prototyping game concepts with gray blocks, which generated too little conviction and led them to pivot to a port of a popular PC genre.

    “we spent, call it, six or eight months doodling around on game ideas”

  • In the early App Store, games could either be free or sold outright with no in-app purchases, so developers used a free-then-premium model and some developers exploited coin packs sold as separate apps.

    “In the early days of the iPhone, there were no in-app purchases, you could either make something free or you could sell something.”