Gamma is the anti-PowerPoint. Grant Lee and two former Optimizely colleagues set out to kill the blank page problem, and five years later they’re at $100M ARR, 50 million users, and 600,000 paying subscribers. Profitably. With a team of 50.
They did the bulk of that with zero sales and zero marketing spend. Growth came from word of mouth. Here’s how he built it, in his own framing, plus the mistakes he admitted to along the way. And that he wishes even with all that … he’s added sales earlier.
Top 5 takeaways from the session:
- Build word of mouth before you spend a dollar on marketing. It’s the only channel that amplifies everything else, and it can’t be bought.
- The first 30 seconds of your product have to be magical. Not good enough. Magical enough that users tell their friends unprompted.
- Become a creator yourself before you run creator marketing. You can’t coach creators through a motion you’ve never done.
- Dogfood to build conviction, not just to catch bugs. Gamma ran two entire products in parallel for six months and let the energy decide which one to kill.
- Revisit pricing and packaging constantly. Set-it-and-forget-it pricing is dead. The seat vs. usage vs. API question never stops being live.
The origin: solving the blank page problem
Lee had just come off an acquisition and was advising early-stage startups. One late night he was fighting Google Slides before a meeting, spending his time on formatting and design instead of the actual content. That felt backwards to him. What if you could rebuild the format from the ground up so people could shape and share ideas without the busywork?
The early prototype was a hybrid document-slash-deck. It got good enough that it became their pre-seed pitch deck. Lee pitched with the product itself instead of Google Slides.
He was living in London, doing all his pitching in the middle of the night to line up with U.S. hours. He’d tuck his kids into bed, borrow their little desk, move it into the corner between the kitchenette and the laundry room, put up a fake Zoom background, and pitch.
Third pitch in, twenty minutes deep, he finished and waited for questions. Long pause. Then the investor told him it was the worst idea he’d ever heard: massive incumbents, ultimate distribution, no way to win. Then he hung up and left the call.
Lee took part of it to heart. In this category you can’t just build a great product. You have to think about distribution from day one. That lesson shaped everything after.
Lesson 1: Word of mouth beats everything, and you can’t fake it
Two years later Gamma launched its public beta on Product Hunt and won product of the day, week, and month. The team felt great. Signups spiked.
Then they plateaued.
That plateau was the real signal. No organic growth. No product people were telling their friends about. Product Hunt buzz is not product-market fit, and Lee is direct that they nearly fooled themselves.
So they hit the fork every founder hits: pour more into marketing, or go back and rebuild the product. Most founders read “do things that don’t scale” as permission to brute-force the top of the funnel. Gamma went the other way. They tore the product back down and asked what it would take to build something with real organic pull.
The bet: make the first 30 seconds feel magical. Twelve people, crammed into a converted two-bedroom apartment in San Francisco, three months, all hands on deck, rebuilding the entire onboarding experience. Type a prompt, get a first draft of your presentation, keep editing with AI from there.
Then they had to announce it. The first draft of the launch tweet was safe and conservative. They scrapped it for something more provocative: “The most valuable skill in business is about to become obsolete.” A little clickbaity, and it worked. Paul Graham replied with some shade, the tweet went viral, and the product went with it.
5,000 signups a day. Then 10,000. Then 20,000. Then 50,000. Zero marketing. Zero sales. People telling each other.
His advice to anyone building right now: before you spend money on marketing, build a product with strong word of mouth. It’s the only form of distribution that amplifies everything else, and it’s hard-earned. It can’t be bought. If you have it, don’t worry about anything else yet.
Lesson 2: Do creator marketing yourself before you scale it
That viral tweet was Lee’s first taste of being a creator, and it made him realize creator marketing could be a real channel for Gamma. Problem: he’s a finance person and had no idea how creator marketing worked.
So instead of writing a check to a big creator and calling it done, he went through the journey himself. Posting regularly. Learning the fundamentals. His phrase for it: you have to go through “cringe valley” for a while, putting out stuff that feels terrible, because that’s how you learn and build an audience.
Once he understood the process, Gamma’s creator program looked nothing like the transactional influencer playbook. Every creator got manually onboarded. Lee would sit with them personally. A lot of them had struggled to make presentations before, so they’d fall in love with the tool first, then tell their audience about it in an authentic voice because they were actually using it.
The payoff was a halo effect. Every time creators amplified Gamma, word of mouth multiplied. They weren’t just tapping the creator’s audience, they were feeding the same organic flywheel. But you don’t earn that until you’ve invested in building the creator community first.
The rebrand nobody wanted to do. At 50 million users, Gamma’s original brand was a placeholder with thin DNA. Too few assets for creators or their own content team to work with. Rebranding at that scale is risky because it confuses the market. They did it anyway, because the bigger risk was staying small. They picked a completely different art direction and open-sourced the whole thing at brand.gamma.app, so any creator can pull assets or build their own Gamma-branded content and amplify the message.
Lesson 3: Community-led growth means treating users as people, not a metric
Lee took “voice of the customer” literally. Gamma uses tools like Voice Panel and user testing to watch real users move through the product and narrate what they’re seeing on screen. That’s where you find the true confusion and the true delight, and where you learn what to lean into.
A few programs that came out of that commitment:
- Gambassadors. A separate space for power users who get early feature access, give feedback, and are usually first to report bugs. Early on, Lee was the first one logging in every day to respond to questions and treat that group as co-shapers of the roadmap.
- Gamma Lab. They flew users from around the world into the San Francisco office to work alongside the team, see what was being built, and give live feedback on the education program so future users would learn Gamma the right way.
- Gamma Everywhere. They visited users in London and São Paulo to run workshops. Each user base needs something different from the product, and you can’t build something that feels native to a region unless you spend time there.
All of these are expensive and eat time. Lee calls every one of them a no-brainer. The bigger you get, the easier it is to treat your user base as a faceless entity, and the harder you have to resist it. They are real users with real problems.
Lesson 4: Dogfood to build conviction
Dogfooding usually means catching bugs before users do. For Gamma it’s about building conviction.
Early on they had two ideas that both fit the mission of helping people communicate ideas: reimagine presentations, or build a virtual office for a post-pandemic hybrid world. They couldn’t decide, so they built and ran both in parallel for six months and let real use decide.
With the virtual office, they kept hitting a ceiling. They couldn’t imagine it being meaningfully better than real-life work. With presentations, there was endless energy and an endless list of things they wanted to build. Six months in, they sunset the virtual office and went all in on presentations.
They still work this way. Gamma Imagine (one-pagers, infographics, design anything) and their new platform connectors both got dogfooded heavily inside the company before release. The point is intuition. You can’t build good product instinct for where things are heading unless you’re living in the product yourself.
The Q&A: agents, sales, and pricing
- Agents are becoming users too. A growing share of Gamma’s biggest customers have humans using the product and heavy automation hitting the API at the same time. Lee’s framing: every company building today needs two mindsets at once, building for the human and preparing for a world where many of your users are agents, without alienating either one. His candid read on the broader market is that it’s still very early and very fragmented. No single team or product dominates, and most users are still just getting the basics right.
- The sales team came late, by design. Gamma got to $100M with a prosumer motion and no sales team. They added one only when inbound got overwhelming and they were clearly dropping demand from people asking how to roll Gamma out to a whole team or department. Teams are much leaner and flatter than they’d have been five or ten years ago, tight feedback loops between sales, marketing, and product, all plugged into a shared context engine holding the customer transcripts. Some customers even get a forward-deployed designer or a technical resource to help with things like API setup.
- Pricing is a live problem, not a settled one. Lee is blunt that set-and-forget pricing is over. The seat vs. consumption vs. pure-usage-API question is one every startup is working through in real time, and Gamma is still on that journey. Their north star is aligning price to value while staying profitable. They deliberately don’t run negative-margin, sell-at-all-costs pricing, because you do your customers no favors by going out of business.
- On speed and models. Behind the credit tiers is model orchestration across multiple providers. Higher-stakes, higher-polish work gets routed to the best (more expensive) models. And unlike some tools that artificially slow the cheap tier to upsell speed, Lee says Gamma doesn’t do that. They run high usage costs but hold strong gross margins because they believe in building a sustainable business.
- Hire slowly, start with generalists. Strong generalists who can spike in a specific area, and a “no job too small” mindset. Lean teams give each person real ownership. Early advice he lives by: hire painfully slow to lock in the DNA, get a critical mass of like-minded people, and then the team hires for the same values on its own. As they open London and New York offices, they’re sending existing teammates as cultural carriers rather than trusting the vibe to travel by itself.
The mistakes Grant Lee owned up to
The wins get the applause. These are the misses he flagged directly, which are more useful if you’re building right now.
- They took two years to launch the beta. His own correction: compress that by 10x or more and get to market far faster. Two years felt like a lifetime and it was too slow.
- They mistook Product Hunt wins for product-market fit. Product of the day, week, and month felt like validation. Then signups plateaued because there was no real word of mouth underneath the launch buzz. They nearly fooled themselves.
- They launched self-serve with no way to pay. The credit-based product shipped before monetization existed. Chat blew up with people asking how to buy more credits, and Gamma had no way to charge them. They scrambled for a couple of weeks to figure out pricing and packaging after the fact.
- They were reactive instead of proactive on go-to-market. Lee’s words: they’ve often been reacting to the situation rather than getting ahead of it, and he’d advise founders not to repeat that. Be proactive if you can.
- They waited too long to add sales. Because they were riding inbound, they hadn’t even started engaging their self-serve base, and they were dropping real team and department demand. They were missing a huge opportunity before they finally moved.
Gamma won on product and word of mouth first, then went back and fixed the go-to-market machinery under pressure. Yes, it worked. But it would have worked faster if they’d built the second half sooner.
