How to Build an AI Business from Scratch

But an investor told me this is the worst idea he had ever heard. There's no way we would ever succeed. >> This is [music] Grant Lee, co-founder and CEO of Gamma, the AI tool that turns one prompt [music] into a finished presentation. He started at the end of 2020 before the AI boom. Today Gamma has 100 million users and a 2.1 billion dollar valuation. A lot of people are trying to build an AI startup these days. What is the one thing they should copy from your playbook? >> You can build anything, whether you're vibe coding or just agentic coding, all the capabilities feel like they're there. You can build anything, the question is should you? So, how do you know? One thing you'll learn is friends will lie to you. They don't want to hurt your feelings, they'll tell you the product is amazing. If your product is valuable, they'll put their credit card down to pay for the product. Then you'll know you have product market fit. >> In this episode, Grant breaks down the three things that decided how to make people fall in love with your product, how to know if an idea is worth a year of your life, [music] and how to pitch it so investors say yes. Can you give me their first 30-days plan, step-by-step? >> First 30-days plan is >> Thanks to HubSpot for sponsoring this video. A lot of people are trying to build an AI startup these days. What is the one thing they should copy from your playbook when building in 2026? >> Honestly, the number one thing is to figure out who you're going to build with. Um, you know, there's obviously a lot of speculation around the single person billion-dollar startup. I still think, you know, even if that were possible, why do it alone? It's so much more fun building with a team. If you're going to build a team, make sure you choose the right team. So, a team needs to look like a set of individuals that maybe share the same ambition, share the same values and principles, but should be very complimentary in skill sets. So, for us, I chose two co-founders that, again, we worked together at a company called Optimizely, became friends, but we're also so different. We bring a lot of different skills to the table. And so, if you're starting to tinker today, find people that are very complimentary. You don't all need to be, you know, developers. You don't all need to be, you know, product people. Come up with something where each of you bring something and then start tinkering on a bunch of different ideas. Find a problem space that feels uh uniquely sort of positioned where you as a team have the best chance of actually going out there and create something that's compelling for the market. >> Yeah, now that you're saying it makes total sense, but back then you had to go through so many difficulties. I read the story where an investor hung up on you saying it's the dumbest idea. Then you were fundraising in your small apartment. Your wife just had a second baby. Talk to me about that time. How did you know you should keep pushing? And by the way, was that investor right about anything at all? >> So the investor, yeah, basically, you know, um this was early on in pitching, uh told me this is the worst idea he had ever heard. Um said that, you know, not only were we going up against massive incumbents, but they had ultimate distribution. So there was no way we would ever succeed. Um and I think, you know, he was absolutely right. You know, in this market, especially what we're trying to build, there's going to be, you know, incumbents. They do have distribution. So as a product, you really have to think about growth from day one. You have to think about ways that your product itself can distribute itself just by use of the product. So leaning heavily into things like product-led growth and how do you build organic virality? These are things you can't afford to wait until many years later. Like as you're building the foundation of the product, you're trying to integrate that into the philosophy of building the product. So he was right and I think for better or worse, we we took that sort of painful feedback early on and sort of internalized it and used that as motivation to keep building. Um for for me, you know, I think trying to start the company when peak pandemic, you know, uh my daughter was just born. There was all the reasons not to start the company. It sounds [laughter] like the worst time, but then in many ways it was the best time because you also start feeling the sense of what if we didn't do something? Like the the regret you'd have, you know? The the sort of life is short mentality of hey, you know, like why not just try something in the hardest of times, you know, if if we come out through the other end, we're going to feel all the more sort of like we you know, sense of accomplishment. And so I sort of avoided the sort of you know, the trap of saying hey, too much going on. I'm just going to wait till later, wait for perfect timing. I think the reality is anybody that's starting a business realizes there's no such thing as perfect timing. There's you with an idea that you just cannot help but like think about night and day. And hopefully if you have a team that's also equally excited and and energized by idea, sort of like you're just hungry to to go after it. We were ready at the end of 2020 to to commit to this full-time. And so for us it was a no-brainer. We were going to go after it no matter what. >> What about your wife? How did she take that decision? Because you quit your job, right? >> so I I was in between jobs. I The last startup I was at was acquired. So I was, you know, exploring what was next. Um I think this is another part about anybody that's building companies like obviously you have your partners internally at the startup, your founders, your founding team. And then you have your life partner. And you have to be equally transparent with both. And so we had a conversation around hey, if I'm going to do this, obviously I need, you know, the full support of of you and and being able to actually, you know, take a big swing at this. And it means like, you know, probably not taking a meaningful paycheck for a period of time. It means many late nights. It means you know, sacrificing some weekends to get work done. And so we all we you know, we sat down and we felt like yes, this was right. It felt like the right thing to do. And and obviously, you know, the the benefit of building something special as a founder means like you share that with your family. If you build a meaningful business, you bring that to your family as well. >> Okay, let's actually test your product. You know, a lot of people who are building today are going to be fundraising. So when you do help me create a perfect deck with Gamma >> Yeah. >> for a seed stage company and you've been investing, you've launched this $10 million for >> Oh, yeah. You raised a four, yes. >> Yeah. So, you know what a good deck is, right? Let's imagine cuz I have a business where we help people uh learn languages uh specifically English for universities. Let's make a deck for a an app that helps people master their English. >> I want to hear your prompts. All right. And this would be for a services business or >> it's an app that I'm trying to raise for. It's >> Okay. Um so, I'm actually going to use this new um uh agent feature down here that we have. So, let's just say um >> You can also use this. >> build a presentation uh pitch You want to do a pitch presentation, right? >> Yeah. >> Well, let's just start there. And so, it's going to do some thinking in terms of, okay, what are the things uh that would be important if you were actually pitching this sort of app. And um this will uh I'll start with a a question here, which is what's your app's name and maybe key differentiator. Um would you like to describe it? >> Yeah, let's say it's a This is like in general general English language learning app. Uh content is generated on the go for you based on students' current level and interests. >> All right. Very specific. Uh what what would you like to cover? Problem, market. So, there's some options here. Um >> What I I guess for fundraising >> let's do all all of the above. >> Yeah. We can ask it to come up with the number. >> we're pitching VCs in in this case or >> And angels, yeah. >> Angels, all right. >> 10 slides. >> 10 slides. >> That's a perfect rate for >> All right. So, we got our basic scaffolding in place and then we'll hit continue. In this case right now, we were uh leveraging more of like web search to kind of parse together, okay, what are some interesting things that would be relevant to an investor. If we were doing this for real, we'd also probably include maybe if you already have like, you know, a one-pager or like, you know, a notion doc with a lot of things that, you know, describe like the the actual, you know, product requirements or the the, you know, the specific features you intend to launch with. All of that can be sort of included. In this case, we'll we'll, you know, keep it to basically what Gamma can search out in the wild with a more generic prompt here and see what sort of can be created. So, we'll choose right now it's kind of looking through. It's adding some of the sort of content things that are relevant. It created an outline. We'll choose a theme and then basically from here we can basically generate the first draft of the deck. >> How many pitches did you have to do to get your first check? >> I probably did over 100 pitches in that >> In how how many days? >> It was over 2 weeks. So, I was >> 100 >> I was living in London at the time. And so, my my version of this was after the kids, you know, put the kids to bed 8:00 p.m. around 8:00 p.m. I would pitch between 8:00 p.m. and 2:00 a.m. basically for almost 2 weeks straight. And and this was, you know, the ability to kind of, you know, do this really quickly. Fortunately during the pandemic you actually the benefit was you could do all do it over Zoom. Today if you had to drive from, you know, investor to investor actually you probably couldn't squeeze in that many. Zoom you can do back-to-back. And so, I had I I I just crammed it in. I would say even in those 100 pitches less than 50% success rate. So, of course, you know, you you have to figure out how do you pick yourself up after hearing no many many times. But I think one thing you do also learn is that in hearing yeses you can always ask an investor, okay, what made you want to invest in Gamma? Like what are the things that actually caught your attention? You fold that into much earlier on in the pitch. You kind of adopt your pitch. And then you can also ask, you know, if you're pitching to angels, angels you can ask, okay, the ones that said yes, you know, who else should I talk to because many of these angels co-invest with other investors. So, tapping into their network, getting a warm intro, these are things that just made it easier even though it ended up being, you know, 100 plus, I got to the point where near the end you're starting hearing many more yeses. So, you start off with like many Os, a few yeses, and you start shifting it to like snowball into people that give you warm intros and you can build towards actually getting many more yeses by by by the end of it. >> Is there a mistake that you're seeing a lot of AI founders are making when fundraising? >> I I'd say part of it is just they, you know, you need to get into the mindset of this is this is a game. There's a two-week period or like again, going back to time boxing, time boxing matters a ton. You don't want to get in position where fundraising drags on indefinitely. Many times people will say, oh, I'll just give myself, you know, a couple months to figure it out. Then there's no urgency for investors and there is no sort of process that that you're running, you're just letting it drift. So, one, define the timeline. Two, out of the founders you should have one person running point. Uh, it's very distracting to have multiple founders all kind of being dragged into hearing all the nose, for instance, it could be demoralizing. You need to have one person that has the right mindset going into it. And then, uh, yeah, number three is like when you design the process in a way that can help allow you to snowball in the end, that just allows you to close, you know, stick to your timeline, be able to actually, uh, you know, wrap things up in a really an efficient manner. And go into you should always go into a fundraise with as much momentum as possible. So, if you're early, that means like maybe having a prototype you've really thought through and it like really feels complete. If you're later stage, obviously it means like you're building towards, hey, we're going into this fundraise with a ton of revenue momentum or ton of sign-up momentum. Go into it with momentum, set that sort of time boundary for yourself two or three weeks, get it done in that process, have real urgency. Uh, and I think that's where, you know, the best founders treat this as like a short-term sprint, get it done with, move on to what's more important, which is building the business, and not be overly distracting to the rest of the team. >> What happens if in 2 weeks you haven't raised what you wanted? Do you just stop and go back to it in 6 months? >> It depends on what the what you're hearing from the market, right? Like if if it's like a definitive no across the board, everybody you've talked to, then um it probably means your product's not ready, or there's something about the market that is out of your your business. >> team. >> Or could be your team, yeah. So it's like you need to almost then be the most you you have to be your own worst critic in that sense. You have to self-diagnose like, okay, what is the root cause? And for every company it's going to be probably slightly different. You know, it could be that the team's not, you know, the right team to go after this market. Could be the market has too many of these they've heard too many of these pitches, so there's not appetite. Or it could just be you've spoke to the wrong investors. Maybe you pitched a bunch of like fintech investors, where in fact your business is a completely different type of business. So each sort of, you know, uh areas like one area to like to potentially diagnose. And then if you can find the root cause, then go back to the drawing board and probably reset the clock and say, "Okay, I'm going to give myself another month to prep for this new process." And then again, go into that 2-week sprint. I would say more often than not, it's because you're going into the process without enough momentum. If you're going into with momentum, in most cases you should be able to find a clear path to to be on the raise. The hard part for most founders isn't knowing that. It's actually doing it when you're buried in operational work. Rewriting the same pitch deck for different investors, researching leads one by one, starting every LinkedIn post from scratch. So much of it is repetitive work AI can already handle. I learned this running two companies and a content business with 18 million followers and a team of 35. For a long time, the honest answer to how to keep up was barely. So my team and I started asking one question on every task. What should stay human and what should AI own. We turned that whole system into a free bundle we made together with HubSpot for startups and we called it the workplace automation kit. Inside are five Claude skills we actually use internally. One helps write investor updates and LinkedIn posts in my voice. Another turns messy brain dumps into structured reports. Plus two creator skills my content team runs on every script including the voice filter that catches AI sounding copy before it goes live. And my favorite is the automate or hire roadmap. It's a one-page framework for what to delegate to AI, what still needs a human in the loop, and what should never leave you as a founder. The whole point is not replacing people. You will be running the same system my team uses every day. You don't need a huge team to use any of this. You can be a single founder or you can be a knowledge worker automating their workplace. Drop the skills into Claude, takes 2 minutes. It's free. The link is in the description. Thanks to HubSpot for startups for partnering with me on this. And we're back to the interview with Grant. Let's look at the deck. Let's see what Gamma built. Can you, as an investor, can you rank rate it? >> Yeah, so this is talking about the language app that learns you. Um, it's talking about a billion uh English learners, zero personalization. So it's saying that, you know, a lot of people go into uh um actually trying to learn, but it in no way has it has this opportunity been personalized. And so, I mean, you know, obviously this is generic. So the question is are you building a product that actually addresses this? Is there true personalization in in the way you're you're dealing with this? Of course, these are this is intended to be a placeholder. So, you know, you'd have to go back and say is this, you know, are any of these things relevant to how I'm thinking about building uh you know, um my app and uh and be able to flesh that out quite a bit more. So, um I think what we found is that, you know, often times when you're starting off with something that is a little bit more generic in this case, you're using it as more of like a thought partner. >> Yeah. >> You're saying, "Okay, these are the potential questions or things that could come up." And then you as the builder should say, "Is any of this actually relevant? Do I believe in any of this?" And if not, then I'm going to go ahead and replace it. And of course, this is where you start iterating. >> Cuz I remember when I was building an a deck, it was just me and like looking at some like Airbnb decks. >> Yeah. >> [laughter] >> Like what's good? >> And um, you know, I think one big thing so traction of obviously is important and to plug in your real num real thing here. I think one big thing that's actually missing here and and where I would have probably um changed the order is to talk about your team way earlier on. >> Mhm. >> Um especially if you're an early stage company, I always think about as like when you're just pitching, you know, your your product and you're just getting going, an investor is really just looking at the team and the dream. It's really just those two things. And you're answering you're trying to answer for them why you and and why now, right? There's probably been many different startups that have attempted building, you know, in this category. So what makes your team unique and special? What what sort of insights do you come that actually offer something that feels differentiated? And then why now? Is there something else that has happened such that there's an opening in this market for for your team to go after? These would be things I would actually, you know, looking at this deck, I would actually push all the way to the front is like start off with that before you start talking about like TAM or like the market opportunity. Just talk a little bit about like what is it about you your your team that feels truly differentiated? >> Yeah, I feel like cuz so many people have so many ideas ideas change so much. You can ship a new product every week like an MVP. >> Yeah. >> New MVP every week. It all comes down to to the team. You were picking ideas based on energy. Is that right? >> Yeah, early on I think for us the the the true sort of North Star was was energy. Was this an idea where again, you wake up in the middle of night sometimes thinking about oh my god like tomorrow I'm going to work on these things cuz it's going to be so impactful and I think that energy gets you to point where you can create the early versions of the road map, you can start talking to early users, but you really need that energy to be the fuel. And we actually had when we first kind of raised money, we actually the ambition was to sort of help people change the way they communicate. So we actually had two different ideas. One was well, you know, presentations are one form of communication. People spend so much time formatting, trying to design it rather than the idea themselves. That could be one category of an idea. The other was this is again peak pandemic and so so much work was shifting to like more hybrid and virtual work. We thought what if we created a virtual office environment where you know, instead of all being in the same room like we can collaborate in a on a canvas and like really share the work that's going on and be able to communicate in a way that almost emulates being in an office. We sort of parallel path like building both for about 6 months. >> Oh, wow. This is a lot of work. >> A good amount of work and we were dogfooding both products at the same time. So we would, you know, you know, create presentations, you present internally to each other. We would sometimes present in the virtual office and like there's some sort of like cross-pollination of the ideas there. And after 6 months, we realized that the presentation idea each one of us could like have like this road map of all the things we wish we could build. Felt like even after 6 months we were barely scratching the surface on like the possibilities of what would the future of presentations look like if we could just invest fully in it. Whereas like the the virtual office, we almost hit this sort of ceiling and part of the ceiling was part just due to our own ability to imagine in the space. It was like, well, we really felt like we, you know, the virtual office no matter how good it got, it could never really replace the sort of magic that you get within real life work and like being able to work shoulder to shoulder with a colleague. And so that was almost a limitation of our own imagination. And so we were like, okay, well, if we're having trouble even like imagining the next few months worth of road map for this idea, let's go back to the one where we can't stop thinking about it. Let's go back to the presentations because this is the one where every one of us lights up when thinking about the problem. >> Do you use the same metric cuz I also read this amazing story how you rebuild the whole onboarding process? Especially when you get or you were getting good traction but you decided to rebuild it and you started getting an enormous amount of users. And in general, when you rebuild a product, when do you know when to stop iterating on the features? >> Yeah, I mean probably never need to stop innovating today especially given how fast things are moving. But when you're early on, the sort of, you know, one question every founder will ask is how do I know if I have product market fit? And I think the the answer for many types of businesses, maybe not like infrastructure or B2B, but most like consumer post-consumer businesses, there's basically two things you're looking for. One is does your product have organic growth? Is it so good people are telling all their friends, family, colleagues about because if they if it is, then you don't even need to spend any money in sales marketing. You'll just have more and more users come to your door like actually try to use your product. So that's step one. If people are using your product, telling all their friends about it, you have organic growth. Step two is is there willingness to pay? Um you know, if you're a consumer product, maybe you can monetize via ads or different formats. But if you're you know, a true sort of productivity or creative tool, people need to pay for your tool if it's really valuable and payment is always an exchange of value. And if your product is valuable, they'll put their credit card down to to pay for their product. So if you have those two ingredients, I think then you'll know you have you know, product market fit and you'll know you have something where you can put more fuel on the fire by obviously, you know, investing in marketing and sales in different ways. But without that, any sort of sales and marketing is is premature because the product itself isn't good enough for it to actually spread on its own and you're always going to be fighting an uphill battle to kind of retain customers. >> Quick pause. My team and I are celebrating one year of the Silicon Valley Girl podcast. I know the channel has existed for a while, but a year ago I made a decision to focus on amazing conversations with people who are building our future with AI. And for a whole year, we worked to bring you the biggest founders and builders in AI so you can keep up in this era and get inspired. If you love what we're doing, please subscribe and hit the notification bell. It's what helps us bring you the biggest minds of our lifetime. Thank you so much for being here. Talk to me about getting those first thousand users. How did you get them and is there any advice on getting them in 2026? >> Yeah, so when we first launched, um, you know, may maybe like many founders, we we tried to recruit all of our friends to to, you know, use the product. Anybody that created a presentation. And, you know, we we were able to get, you know, several thousand people just to sign up through our network, extended network. And of course, you ask your friends, like how do you like the product? And what one thing you'll learn is like friends will lie to you. They don't want to hurt your feelings. They'll tell you the product is amazing. The real truth is then going back and looking at the usage, right? Like are people actually using the product? And so then we'd go in and, you know, many of the friends that said they love the product weren't even coming back to the product, not using it. So then you start honing in, okay, where is the real usage of the product? What kind of use cases are people actually, you know, leaning into? And we saw early on a lot of, you know, freelancers, solopreneurs, small business owners that had to create a lot of content using Gamma as the way that they were presenting and sharing with their clients and customers. So that's where we understood, okay, there's a pocket of users here. We should just be spending much more time with them, talking to them. And then of course, over time, then you can find out ways to like message to them specifically, whether it's on social media or, you know, if you're working with creators and influencers, like really tap into the personas where your product is the most useful, and start getting that flywheel to going of they're finding the product, they're finding real value, they're sharing the product with their network, with their friends. And that's where you can really start building kind of the the initial core of your your your kind of core users in the beginning. >> But you haven't done a lot of paid marketing at the beginning. So was it just word of mouth? >> It's just word of mouth. Yeah, word of mouth. Even today word of mouth is by far, you know, the number one sort of uh acquisition channel for us is like people telling other people to use Gamma. >> That's fascinating. And but you also do influencers, right? >> Yeah. >> And you do a lot of micro influencers over going after big ones. What did you learn in that process? Was there an like an influencer campaign that didn't work? >> So the very beginning it was actually um all organic uh influencers. People basically learning about, you know, Gamma and then just just sharing uh you know, their ideas. >> me. When I first saw it I did like a video about, yeah. >> We really appreciate that. >> amazing. It's amazing tool. >> Yeah, and so like, you know, that's where we started seeing just in the very beginning like some organic spikes of, you know, people coming and and learning about Gamma and like starting to use Gamma. And so we realized, okay, well, if this is already happening organically, what could we do to reach an even broader, you know, set of creators and influencers out there. And we started just testing a bunch. So um I also in this process started to try to figure out how could I be a creator myself because I felt like to really empathize with other creators, I I kind of needed to go through the on the journey myself and like just understand how hard it is, right? To even build an audience, come up with content, share information in a way that feels like authentic and and really genuine and like you you you need to like build that muscle, right? And so for me it was like, okay, I'm going to start by learning how to do this. And then when we do work with influencers and creators, I'm going to onboard all of them myself. And I'm going to drive >> you know how to do that, right? >> how to do it, but I'm also still learning, right? Like it's helpful for me to like go down and say, okay, rather than just saying, "Hey, go, you know, go and showcase Gamma." It's like, "Okay, let's actually sit down and think about your audience specifically." And what I actually learned was that many of these creators struggled with presentations cuz they were never learned it in school, they never, you know, learned it on the job. and so I got to actually unveil kind of the magic for them. And then specifically for their use case, like, you know, if they were targeting, you know, real estate professionals, we would think through, okay, how do we actually package this up in a way where the use cases we're showing are relevant, make sense, the way we talk about the product, like resonate with their audience, and we can brainstorm a ton. And that for me was also a learning opportunity because then I could really say, okay, what do the best creators do? How do they like write a good hook? How do they come up with like a structured story in a short amount of time? How do they capture people's attention? And you know, it was I was also sort of like, now I could teach them about the product, they learn how to use Gamma, they could talk about Gamma in a very authentic way, and then in the process I can become a better creator myself and translate that into, you know, many founders today have the chance to be their company's sort of megaphone or amplifier, right? Like anytime they want to share information, if you have your own audience, you should be able to package up that narrative in in the right way, too. And I did so by just learning from, you know, the best creators out there, like how can I actually build this muscle myself? And it definitely felt like there was some good synergy there. >> And you're active on X and LinkedIn, right? >> Try to be on active on X and LinkedIn, yeah. And and I think this is where, yeah, many, you know, customers will reach out, many partnership opportunities come to fruition because of those, you know, because of the audiences that I've built there. And again, barely scratching the surface, but more I want to do, and I think it just feels very rewarding to build to start building an audience. >> Yeah, just wanted to highlight it for everyone. Like 10 years ago, when I was talking to entrepreneurs, the sentiment was you're not taken too seriously if you're actively posting on social media. Now, I think it's such an inevitable part of being a founder because this is how you hire, this is how you learn the marketing channel, and this is how you just meet the peers. >> Totally. And I think there there is a negative stigma because, you know, sometimes, you know, this is more like 5-10 years ago, where people almost felt like they were tooting their own horn, they want to like, oh, hey, look at me, look at me. I think you need to reframe it. Like if you have an audience and hopefully your audience overlaps with your core user base, you're trying to provide as much value to them as possible. Like you're not talking about your product all the time. In fact, you should be talking about your product very little. You should be talking about things that your audience cares about, has questions about. Obviously in this day and age, everyone's curious about how to apply AI to their work lives. How do they, you know, how do they sift through all the noise? How do they maintain a sense of like productivity without just becoming insane because there's just too much to like try to do in any one given day. So, I think as a founder, you have a chance to figure out what are the, you know, with your experience and your background and what you're building, what can you instill on your audience to like inform them and keep them educated? That value in itself should be where like all the, you know, the ability to build an audience comes from. It's not about tooting your own horn or like really just talking about your product all day. Have you looked at users who are coming from LLMs when ChatGPT recommends Gamma? >> we're seeing we're tracking on a newsletter. We're just seeing crazy open rates when people find us through LLMs. >> Totally. Yeah, that's actually been one that's also trending very positively both in terms of people, you know, landing on you know, Gamma's website but also converting long-term. So, I think we're early days. It's still a relatively smaller portion of overall traffic but it's definitely growing and it's exciting to think about what this could look like obviously in a couple years and you know, every business wants obviously to diversify like how customers hear about this you know, so that you know, obviously you want to build out activate multiple different channels at once and it does feel like LLMs, you know, chats like ChatGPT, Claude are going to be, you know, meaningful channels for us in the future. >> Talk to me about pricing in general. So, you're you start with free, then it's $8 and then it goes up to $90, right? How do you come up with pricing for an AI product? And also, it's dependent on how the models price themselves, right? >> Totally. So, yeah, there's a lot of different moving parts. I think every startup today needs to be thinking about pricing as something that they're probably constantly experimenting with. It used to be, you know, you know, 5, 10 years ago, you kind of set it and forget it. You come up with pricing once and maybe revisit it every several years. I don't think companies today have the luxury of doing that. You constantly need to be thinking about, okay, the tension between like seat base versus usage base, how your product evolves, the different segments of users. So, we have users that are individuals versus, you know, B2B customers that may want to buy in a completely different way. So, I mean, the honest answer for us is this is something we're still constantly experimenting with. You also want to do so in a way where pricing, you know, pricing you're trying to align the value you provide to your user. And then the other side of the equation is you need to be able to build a durable business. So, are you actually, you know, building a business that's able to operate profitably or efficiently? The worst thing you could be doing is to be essentially selling dollars at a discount, right? Like, if you're selling dollars at a discount, feels good, you're making money, but you're doing it so very unprofitably. So, these are all things where constantly moving target. I think the best you can do is being willing to adapt and then user user base to kind of figure out what's the right way to balance both the value they're seeing and the cost you incur as a business. >> Do you have any strategies on converting customers from free into paying? Cuz you had this competitor, Zomato, that shut down and they had 25 million users, 3.5 million AR. You had 40 million users at 20 million AR. So, you're doing something there converting customers. Can you give advice to AI founders on how to do that? >> I think it will sound generic, but it just comes back to like what is the quality of the product, right? Every single category, if it's a meaningful category, there's going to be multiple competitors. And so, your ability to kind of ignore the competition and focus on your users is all that really matters. Are you delivering something to them that competitors aren't, you know, either delivering on or on focused on, right? Like even within our own category, you know, there's many people that focus on kind of the design portion of like creating a presentation. We want to focus on like the communication, the storytelling. So like how do we abstract design, the need for design completely away? And so you're going to approach it differently depending on your sort of vision for this category, and then you're specifically like your customers you're serving, what are their needs? Many different [snorts] types of customers might have different needs for presentation tool. Like you need to figure out where in that space do you actually want to play? >> And your your strategy has been product first from what I'm hearing, from what I'm experiencing. And you only have 50 people. Is it still the same? >> We're a little bit bigger than that, but like when we were crossing 100 million in ARR, we were 50 people, yeah. So it's a relatively small team. >> Ren just said he crossed 100 million in revenue with only 50 people. This is exactly the kind of thing I love to explore in my Future Proof newsletter with a community of 18,000 builders and creators. Twice a week I share the best practices from my conversations with CEOs and founders, the AI tools I'm actually using, and the skills worth learning right now. It's free, the link is in the description. >> Question that's very I think is on point for 2026. Do you think you can triple your revenue with the same amount of people? Is there technology there? >> Yeah. I think the answer is yes, and then there are the the nuance is how do you triple it, right? Because we're trying to build a global business, and so this means, you know, we have a lot of users in Europe, a lot of users in Asia. We still feel like we need to build a physical presence there, and so ignoring that for too long, just assuming hey, we'll just grow from San Francisco and not, you know, grow head count, I think would be the long wrong long-term decision. And so we are trying to grow now our footprint in these different offices, which means growing the team much bigger than it was, you know, obviously a year ago. So I think, you know, every business should think about what is their ambition as a business. If it is to be a company that has global presence, if it is to serve many different segments of users, so like we have the prosumer end and the B2B, you know, set of users that are much different in profile, and then also serving developers through our platform or API MCP part of our business. Those all require different teams and like different places of like emphasis in terms of building out the the complete product and platform. Those require head count. And so, if we were just focused purely on prosumer, growing that as fast as possible, I think we could stay pretty lean. But if our ambition is to build a global business that can operate across segments, I think we're going to need a much bigger team. >> And also the question is, is there a longevity in just growing the consumer business like trying to do everything by yourself? >> Trying to do everything. Yeah, that that is another, you know, tension I think every startup is feeling today, especially today, because when it feels like you can build anything, you know, whether you're vibe coding or just a genetic coding, all the capabilities feel like they're there. When it feels like you can build anything, the question is, should you? >> Yeah. >> And if not build everything, where are you going to focus, you know? And so, um this is this is both like the the challenge and I think the opportunity building today. Like you have the feeling of being limitless in terms of going out there, but you also need to provide your team the right level of focus and and to be kind of committed to some path of this is what we really care about and these are the things we don't and being committed to that for as long, you know, long of a time horizon as possible. >> Your principle is to hire painfully slow. Is there a role that you wanted to have in your company, but you're not hiring because you think it's, you know, you're you're still like debating whether you need it or not? >> I think it goes back to like every, you know, stage of company, you need to ask yourself sort of where you're at and and what are the limiting factors for you to continue to grow. Often times the moment a startup raises a little bit of money, they think like, okay, I need to immediately open the floodgates on, you know, sales and marketing and just like hiring as many people there to to build top of the funnel, build demand." But, going back to like, is that the best way to grow an early product? Probably not. You know, it actually means going back to focusing on the core, like making sure the core product is so good, going back to organic virality is like so good people tell all their friends about. We wanted to focus there. So, rather than like open the aperture in terms of hiring across the board, when we were super early, it was all about reinvesting in the core, hiring the right, obviously, engineers, product designers. We We actually probably over invested relative to what people would say, like in product design. We had it was almost a third of the company was product designers early on. And most people were like, "Why would you invest so much in the product design?" We're like, "Well, that matters a ton to us." So, even though we were small in, you know, like the relative overall head count, we were probably over indexed on product designers because we felt like that was super important. And to get true organic virality, we needed to be exceptional in that one area of, you know, that one function. And so, uh every stage you're going to have to look at. Now that we're much bigger, we can't afford not to hire sales and marketing because we have a massive audience to go after and we have a B2B business that still requires a ton more relationship building, a much more consultative approach to to like actually, you know, expanding within these larger companies. And so, we can't just, you know, hiring people only slow at this stage doesn't mean we we never hire sales and marketing. It means now that we're at the stage, we we feel like we've earned the right. And we can gradually open up the aperture a bit more. If we had done that two or three years ago, I think that would have been premature. >> How many of your employees are generalists versus specialists? >> I feel like nearly all employees are what I would still define as generalists today, even though they spike in different areas. So, you know, for instance, our head of design, exceptionally talented designer, very visual, but he also knows how to code. And he was coding even before like, you know, vibe coding was was more accessible to to most designers, but the fact that he can now ship something like end-to-end, test it, like get early versions into hands of users, that's a superpower, right? And so, he can cross domains where, you know, traditionally it'd be a designer or UX researcher that does a little bit some of this like early, you know, sort of investigation or like early research. He can do that himself, develop the early prototype, ship the early prototype, get something into the hands of users to actually get feedback on and oftentimes then ship something into production, like almost like completely end-to-end. That just wasn't really possible before, but for us, like we try to look for that the spirit of like a generalist across every function, whether it's, you know, sales, marketing, engineering, product design. And I think that allows you to have a relatively lean team that can, again, like play across multiple domains at any given point in time. And I think in general that just means a team that can be pretty collaborative, too, because you have deeper empathy for what each function can bring to the table and and like when you're bringing something to to an engineer, for instance, you've already done so much work to think through like, okay, how do I make this right so the handoff is as seamless as possible? These things I think are hard to quantify, but when you have a team of generalists, there is some magic in like working together. >> But does this model ever break or start breaking with scale? Do you feel like there's a there will be a point in time when you will need to start hiring more specialists? >> I definitely think so, and uh but but there's also the sort of competing sort of race of like as you get bigger, you want more specialization. LLMs and AIs also getting better at being specialized in certain domains, and so, there's almost like this race now of like at what point does AI also kind of fill in the remaining gap and the generalists can go deep when they need to. TBD, I think certain functions like I think we're noticing, you know, like within like finance and legal, still pretty hard, like you still want specialization in specific types of questions that come up, like you don't want to just trust like an LLM, obviously, to give you the right answer. But for a lot of like sales and marketing where, you know, your goal isn't to replace the human, it's to augment or like to really automate as much so that the human can focus on what they're good at, which is the relationship building, again the consultative approach. You're trying to carve out, you know, if they only have a certain number of hours in a day, you don't want them to be wasting any time like fiddling with, you know, business systems or any of that. They should be spending time with customers. And so our goal is like those generalists can help automate as much of the work that is not customer facing as possible, and the rest of the time there is there special you know, their specialty is the the the work with customers. We're we're trying to carve out as much time as possible for for them to do that. >> Yeah, that makes total sense. With all the companies deploying agents, what was the last process that you completely delegated to an agent? >> I yeah, I think one thing we've learned is that you we even with agents, you're not completely delegating things away. Like a human still needs to supervise, yeah. And, you know, agents can still be fragile, they can get go off course. And so, um, I think the best cases are still where the human and agent are working hand in hand. And of course maybe over time you have multiple agents all, you know, working with the same human and they can then be subdividing into different tasks. But we're still so early, you know, we're not at the stage where yet where I think any of our teammates feels like, okay, 100% agent work now, now I can go do something else. It's more like, okay, agent can actually help me provide a ton more leverage, can obviously reduce a the time to complete a task, and then can help me do the things that I would just wouldn't ever gotten to cuz my time in the day is too precious. And so these there's certain aspects of that. You fast forward several years from now, then I think we're talking about maybe agents doing way more, uh, being much more self-directed and and truly being like a a partner where, um, they you can leverage the the agent sort of work force in a way that today is still seems like we're It's bit out of reach. >> What's your favorite agent? >> I mean, we build a lot of things internally to to help with um you know, things like customer research or even just understanding, you know, what what what should we be building next? Like what are pockets of of things that are going to be useful? So, like, you know, taking all the input we get, you know, we we look at obviously things like churn analysis, like what are people saying? We look at when new users sign up, you know, what are they using? How are they leveraging the product? These are all inputs and so we want to be able take those inputs to say, "Okay, now [snorts] stress test like some of the other things we're we're thinking about building. Like how useful is this to to an end user? Um where might we kind of uh add value in ways that we aren't even, you know, considering today?" Like, you know, really being a sparring partner there. I think those are things where um you can now with much more like every company now has so much more data they can actually work with. Um taking advantage of that is is is really I think where the the magic really happens. >> So, your analyst is basically data analyst meets product and they make decisions >> Yeah. >> Somebody who watches video has a startup idea, built a deck with uh with Gamma. Uh can you give me their first 30 days plan, step-by-step? >> First 30 days plan is uh spend all 30 days talking to your customers. >> [laughter] >> Really is I think it feels like, you know, if you're assembling the team, you need to be uh laser-focused customer obsessed from the very beginning and nothing else really matters in the long run. Um I would >> you wouldn't start with a co-founder? >> Well, I mean, I I'm assuming if you've raised money, you have a co-founder so >> like you just had an idea and you're like, "Okay, I'm going to start." Is it the co-founder or you start talking to clients first to figure out who you need as a co-founder? >> I personally think you shouldn't uh jump into the idea before the team because the team should help inform what the idea actually is. You might have like, "Hey, I'm again going back to our specific, you know, thing which is we think like, you know, visual communication can be improved." But we weren't going to be prescriptive into how we actually solve that. Let's as a team start tinkering. Like, what can our team deliver in the space that feels differentiated? And once we had that, that's when we said, "Okay, we feel like there is something here. We don't know exactly what the product might look like in 3 or 4 years, but there's something here. Let's go raise money so we have a chance to actually build this for, you know, over the next several quarters." Once you have that, then it's about you you have the the early team, you flip all of your energy into spending time with customers and early users and like testing like, does the my thing even deliver on, you know, uh anything that's valuable to you? And if not, let's let's course correct. Let's make sure we're not heading down a path of like delivering something that has zero value to you as a customer. Let's make sure we're uh listening to your problems, fine-tuning our own sort of understanding of the problem space, and then building the right thing over the, you know, the long haul. >> So, for someone who um came up with the idea, found the team, which is kind of hard if you are not in the ecosystem. Yeah. Like, finding the right team. But, would you say this deck will help uh that company raise money? >> I would say yes, but it it's it will help you uh the it'll help you sort of have the right conversations. But, ultimately, the investor is going to judge you based on your level of conviction and their level of conviction in in your team to go after this market. So, use this as a tool. And every great tool can help you, you know, open the right doors and obviously have some of the right conversations. >> And we'll wrap up with one final question. So, if you could send one message to the founder version of yourself in 2023, when your growth was stagnating, 6K users, just start there, what would you say? >> It'd be some version of uh just hang on. You know, like every founder will go through many ups and downs. Um if you're surrounded by a team that um again shares like that same level of ambition and and the and principles, then you [snorts] got to just lean uh you know, lean on them as much as you can and then still have belief that you can make it through. The early days are the hardest because it very much is they describe it as like the idea maze. You're like bumping into wall after wall and not knowing if you've made any sort of meaningful traction. And so, you know, for us it was like hang on for as long as possible, make sure we don't lose belief in the longer term, you know, mission of trying to help people communicate their ideas. Like we feel like that should be something that we could commit to for a very, very long time. And so, hanging on to that belief that we can make a difference, like that's what, you know, for us actually did change the the overall trajectory of the business. >> And it comes from the energy, I guess, that the product is giving you, right? Is that what's the source of that belief? >> It comes from it's it's it's the energy in knowing that what you're working on should help improve people's lives. And like the the the energy you're putting into it will hopefully result in amazing outcomes for for your customers and users. That energy obviously, you know, it needs to be built up. It's like the conviction that needs to you're constantly trying to fill up as much as possible. And And if you're building the right things, I think you'll you'll wake up every morning being excited about the things you're working on. >> Thank you so much, Grant. >> for having me. >> Thank you. Grant built Gamma to a $2 billion valuation. If you want another founder's story from the AI era, watch my conversation with Young Zhao, the CEO of Opus Clip. He gave me one of the best frameworks I've ever heard for picking a niche when there are already 50 AI tools in every category. Click right here or here. Somewhere. And also in the description.