🎧🍌 $50M ARR with 30 Employees: Inside Gamma's Journey to 50 Million Users | Co-founder and CEO Grant Lee
Taking a horizontal vs vertical product approach, rebuilding as an AI-native product, importance of design, hiring generalists, and building a culture with zero employee attrition
Grant Lee started Gamma with his co-founders in 2020 to build the anti-PowerPoint, changing slides from design to narrative-driven. As someone who’s made a lot of decks, it’s a refreshing approach.
They’ve put up impressive metrics too, growing to $50 million in ARR and 50 million users with only 30 employees. Even more impressive, they’ve had zero employee attrition, and have a negative lifetime burn rate, with more cash in the bank than they’ve raised.
We talk about building a horizontal product instead of for a specific vertical, why Grant likes hiring generalist’s, how a quarter of the team is designers, why they’ve never raised large funding rounds (and what that means for employee's equity value), and how they run the company with an efficient team while not subscribing to 996.
Grant also shares how Gamma rebuilt their entire product to be AI-native in the months after ChatGPT launched, and how every department at Gamma uses AI internally.
Shoutout to Anamitra and Gaurav at Afore, Shiyan at Hustle Fund, Evan at South Park Commons, and Vas at Accel for helping brainstorm topics for Grant.
Presented by Ramp
Time is your most valuable resource. And its impossible to get more of it.
Unless you’re one of 40,000 companies using Ramp.
Ramp is the easy-to-use corporate card, bill payments, accounting, and expense management platform. Issue cards, get spend visibility and control policies, manage approvals, vendors, travel, and close your books 8 times faster, all in one place.
The product is simple enough for startups, powerful enough for large enterprise, and the Ramp team is known for being one of the fastest shipping in tech.
Stop chasing receipts. Make your expenses submit themselves.
Time is money. Save both with Ramp.
Try it here.
Numeral: The end-to-end platform for sales tax and compliance. Try it here.
To inquire about sponsoring future episodes, click here.
Timestamps to jump in:
3:46 Gamma: The anti-PowerPoint
5:56 How to be efficient with a small team
7:58 Importance of full-stack generalists
12:15 How to hire problem solvers
15:57 Changing slides from designs to narratives
20:13 Gamma’s freemium AI business model
22:36 Ignoring conventional wisdom with a horizontal product
28:47 Why Gamma started with Slides
32:21 Raising a Pre-Seed for a horizontal product
38:25 Why Gamma avoided hyped funding rounds
40:57 How fundraising impacts recruiting
43:40 Liquidation preferences and employee equity
47:08 Gamma’s zero employee attrition
49:54 Working in-person during COVID
52:15 Using waitlists to batch new user cohorts
56:08 Re-building the product to be AI-native
58:17 How to improve your onboarding
1:00:46 Benefitting from AI models getting better
1:05:10 Growing from 60k to 50 million users
1:07:30 How to stand out as an AI company
1:09:23 Creating a Gamma API
1:11:37 How Gamma uses AI internally
1:15:06 Why Gamma doesn’t do 996 working hours
1:19:54 4-month product sprints
1:22:16 Parenting hacks: sleep, exercise, nutrition
1:24:26 Brand and community lessons from Nike and Apple
Referenced:
Find Grant on X / Twitter and LinkedIn
👉 Stream on YouTube, Spotify, and Apple
Transcript
Find transcripts of all prior episodes here.
Turner Novak:
Grant, welcome to the show.
Grant Lee:
Hey, thanks for having me.
Turner Novak:
Yeah, I'm excited to do this. I kind of wanted to start off, just really quick for people who don't know, what is Gamma?
Grant Lee:
Yeah, so I guess the easiest way to describe it is we're trying to build the anti-PowerPoint. PowerPoint's been the language of business. People use it to present, share their ideas, and we're trying to build a tool that fundamentally changes how you might approach sharing your ideas.
Turner Novak:
So this is maybe a rhetorical question for somebody who knows what Gamma is. For somebody who doesn't know and is not familiar, why is that a big deal? Because you think PowerPoint, it's just some slides and pretty straightforward.
Grant Lee:
Yeah. So, I think with PowerPoint and many of the sort of traditional tools, you get dropped into this canvas, and you as a person, you're forced to be a designer. You're kind of moving pixels around, resizing boxes, aligning formatting, and I think the vast majority of us aren't visual designers.
And so, we want to take a fundamentally different approach, which is rather than being designed first, what if it could be design later or design last? And much more lean into content first. You start with your ideas, we help you shape them. And so, we want to just change that approach because we think it actually aligns better with how most people actually want to create content.
Turner Novak:
And it sounds like people are liking it and using it. I know there's some public stuff that you've kind of talked, about how well the company's doing, but what's current state of the business?
Grant Lee:
So we passed 50 million users, 50 million ARR, been profitable. I think a lot of what we pride ourselves is in trying to scale the company really efficiently, too. And so, that inherently means you need to have some pocket of users that love your product, that are pulling you into the market or pulling the market towards you. And we're so lucky that we're still early, but I've been able to achieve some great early traction.
Turner Novak:
Yeah, I think I saw that you're profitable, I think this was in an article, that you have more in the bank than you've raised. That's always a good, like a humblebrag in VC-backed startup world.
Grant Lee:
Yeah, it's a stat that doesn't get maybe shared very often, which is lifetime negative net burn, more money in the bank than you've ever raised, which yeah, feels good.
Turner Novak:
Yeah, I know we're going to talk a little bit more about that, but I think the thing I thought was most interesting is you have about 30 employees, and how do you be so efficient with such a small team? 'Cause I feel like everyone, that's kind of come back in vogue. I feel like '21, the metrics that KPISP would lead with is, "We have X employees." And you have pretty small team for kind how far you've gotten it. What do you think is the secret to doing that?
Grant Lee:
Yeah, I attribute it a lot back to just how the company was formed. So, we started off with three co-founders, but then we had a founding team of seven. And we all came from a company called Optimizely. We had worked together for five-plus years then, and now it's been basically almost a decade working together. And really just having that foundation of trust, but also very complimentary skill sets. So, three co-founders, my two co-founders, John and James. One was the VP of product at Optimizely, the other was a senior engineer, and then myself coming more from an operating background was basically leading finance and strategy there.
And so, I think each of us kind of brings different skill sets to the table. And then we look at the broader team, we had amongst our first seven, we had our head of design, we had a front-end engineer, we had two backend engineers. And so, we had everything we needed to get the early versions of the product moving in the right direction.
And then since then, we've continued to layer on, and I think that just means that when you think about having the sort of leanest team possible, you need all the positions filled, and each one of those need to be filled with someone that's obviously highly capable and can move fast. And then the more that you can bring together that trust factor, people that work together, that have that muscle memory, that allows you to also just navigate much quickly than you could otherwise.
Turner Novak:
So I saw specifically the first hire you made, I don't know if this was a part of the seven person founding team or just the fourth after the co-founders, was you hired a designer, which I guess is maybe you'd say that's ironic with you're an AI-driven design company. There should be no designers, maybe there should be a bunch of designers. So why hire a designer as the first person?
Grant Lee:
Yeah. I feel like others, when they've looked at the makeup of our company have always said that it seems like we've over invested in design, 'cause almost a quarter of our team is our product designers.
Turner Novak:
Oh, wow.
Grant Lee:
And when you compare that to many other organizations, that's just not the same makeup. And we always believe that was the right investment, especially early on. Zach, who is, to your point, who you're mentioning here, our first sort of design hire, he's our head of design. He is so gifted in many different ways, but what he brings to the table is this ability to really explore many different ideas and help us understand the different directions we might pursue.
And I think only through prototyping and iterating with customer feedback, and doing that relentlessly over a long time horizon, can you finally get to a point where the product can be useful, it can be valuable. And we knew that we needed that on day one. We couldn't make progress without that sort of core component as a strength of the team. And so, yeah, it's been awesome having him on board since almost basically day one from the company foundation, formation. And then since then, having him be able to build out a team around him.
Turner Novak:
I think there's this concept of a full stack designer, where they can not only design but also code and build stuff. And has that been a big piece of it, too?
Grant Lee:
Yeah, 100%. And this is becoming maybe more obvious now just in this era of AI and all the advancements here, but I do think in many ways this is the rise of the generalists. And whether it's full stack designers or a PM that knows how to code, or a marketer that can do both the research as well as going in and actually synthesizing that, coming up with a strategy and actually then deploying that through different campaigns. All that's becoming more and more important for the average person to be able to play a variety of roles, and not to be necessarily siloed as a specialist.
And so, yeah, absolutely that sort of full stack design has been a designer like someone like Zach who can code, who is also visual, and can also sit down with a customer and deeply empathize with that feedback. That's a superpower. And bringing that all together just gives you as a company a huge advantage.
Turner Novak:
Is there a trick for just hiring people like that? How do you just screen and make sure when you're adding to the team, that like, "Oh, this person is good at understanding the customer can also build the stuff for them, can make it look good, is a good boss and manager, and can help us recruit"? Just how do you find these just full stack generalist people?
Grant Lee:
Yeah, there's this intangible that you sort of look for, which is someone that's just continuously learning. They're hungry to learn. Oftentimes, what you'll find is someone, as they progress in their career, their desire to learn new things quickly diminishes.
Turner Novak:
Oh yeah, I definitely get caught in that rut sometimes. That's part of the doing this podcast is like, "Okay, who can I meet just a new industry?" The episode, by the time this comes out, it'll be a couple of weeks ago, but I had Sanjit Biswas at Samsara, and they do kind of internet of things, physical sensors and devices for transportation. And it's kind of interesting topic, know nothing about it. I feel like I learned a lot from it. So just like you just got to expose yourself to it.
Grant Lee:
Yeah, you got to expose yourself. And some people just are constantly trying to stretch their own sort of thinking and imagination. And that comes with learning new tools, new techniques, new frameworks. And when you talk to someone, the energy they express, the way they talk about things, if they're expressing these new things with a level of excitement, and you can even ask them to teach you about some of these concepts, like how deep can they go, I think that those are all signs of someone that is just continuously learning.
No matter where they're in their career, whether they're fresh out of college or they're someone that's 10 plus years, you want to feel that energy, and someone that really just enjoys teaching. If they know the concept well enough to teach it, it means that they're probably armed with the right tools to be able to take a ton of ownership, high agency, move fast. And so those are absolutely the things you're trying to suss out during the interview process.
Turner Novak:
How would you structure that? I know we're getting way off script here already, but how would you suss that out in an interview, figuring out what kind of things would you ask and figure out where to dig deeper on?
Grant Lee:
Yeah, there's maybe a few common techniques. One, is just asking them to literally teach you something, something that they're passionate about. It could be completely not even work-related, just have them teach you about something they're passionate about and see how they articulate those things. Is it well-structured? Have they been thinking about this for a long time? Do they do a good job of getting you excited and understanding and appreciating the topic?
And then the other is to ask them to talk about a challenge that they've experienced and overcome. And so, oftentimes, you'll have somebody that can describe the situation with some level of humility and be able to actually go deep into understanding the problem. If someone rushes immediately into the solution, it's usually a flag. It's like, "Okay, they're able to really quickly jump to something and highlight how great they were."
But a lot of what greatness looks like actually is you encounter a problem, you realize that's actually the surface level problem and you dig 10 levels deeper. And someone that can actually articulate how they understood what the true problem was and then went about trying to fix the problem, and then realizing, "Hey, actually the initial solution maybe wasn't the right solution. I went back and iterated." Those are the things you really want to hear about.
And especially if it's something that's super challenging in that person's career, they should be able to go into depth. And if they aren't able to, it just means that, again, they probably are a surface level thinker. They haven't gone to the extent of really learning and trying to go deep on anything.
Turner Novak:
Have you as a candidate, I don't know if you've ever been through that in your process, but how do you do that without, I don't know, setting yourself up in the wrong way or setting yourself in the right way, if that makes sense?
Grant Lee:
Yeah, the tricky thing with interviews in general, some people end up becoming really polished 'cause they practice a ton, they become good interviewers. I think in general, the more I've done interviews myself, it's just when you talk to people, you want it to feel authentic. And you can kind of start getting a glimpse of how much are they trying to have the PC version of the story or trying to make themselves look good.
And we all know the truth is going to be somewhere in the middle, and you want to be able to hear some texture to the story. It's not just all always great, there's some areas where there's wrinkles, and you got to go in and be able to have someone that's willing to appreciate that and go into detail. Otherwise, you know they're just holding back a bunch. And if that's the case, it's hard to trust how much of it is being honest and authentic.
Turner Novak:
I feel like a lot of people kind of forget, maybe if you're just a younger student maybe and you're interviewing for your first job or something, you kind of forget this whole interview thing, we make a big deal about it at certain maybe phases of your career, but at the end of the day, 99% of the time is just sitting there working, talking with people. And no one knows the answer. I'm sure you don't sit there with the team like, "Oh, here's my polished research response to this thing." You're like, "I don't know, let's just figure it out" or something.
Grant Lee:
Totally, yeah. Yeah, a lot of it is passing the elevator test, like, could you stand here with them? And is this someone you gain energy from? And so, if it's a person that all they say is about how great they are, it's hard to imagine you wanting to spend a lot of time with. But if it's someone that is genuinely curious about the world, that's great because that means there'll be an infinite number of things you can work on with this person, and you're never going to end up being in a place where there's actually not enough work for this person to work on.
Turner Novak:
And I kind of want to talk a little bit more about the Gamma product, AI, some of the stuff you're kind of thinking about and using. So what's been maybe the most interesting use case that you've seen for the Gamma product? I assume you have some stuff off the top of your head, but.
Grant Lee:
Yeah, the initial wedge for the product was just a way for people to create presentations very simply and quickly. And so, what you'll end up seeing is a lot of people creating a lot of external-facing content. So if they need to share something with a customer or client, a partner, and that could be in the form of a pitch deck, a proposal, some sort of onboarding or training. Those are all things that traditionally are a pain in the butt to try to create in traditional tools like PowerPoint.
And not only just create, but also then edit and refine, and then actually share and get feedback on. And so, we're just trying to make it that simple for people to create those forms of content where all that could be done very fluidly. It could be shared, you can iterate on it, editing is in a pain in the but, and get your stuff out there quickly.
Turner Novak:
Yeah, I think going back to the beginning, I know your initial insight was basically it's not design, it's like writing. I was listening to some other podcasts you'd done and I was reflecting. I was like, "Yeah, you're right. Whenever I make PowerPoint, I just kind of write it out first." Even if it's in the first slide, I'll just make some bullet points that are the title of each slide, and then maybe I'll expand on the bullet point and go a little bit deeper. But that's historically, I've always done it.
Grant Lee:
Totally. Yeah. I think that's most of us, we want to have some general structure or outline that we have in mind, maybe the story that you're trying to tell so that there is some cohesion between what is the main message and how do you then build upon that? And so yeah, it becomes challenging when the medium you're working in forces you to instantly have to design at the same time, because then those two things are at odds. The storytelling, the structure are competing against, "How do I visually format and put this in a layout that puts it in a way that's consumable by the actual reader or the person you're presenting to?" And so those are things we're trying to remove that sort of friction or that tension and making it much more fluid in the process.
Turner Novak:
Yeah, 'cause I think about when I do the deck for my fund when I'm fundraising, my designer made it in Figma a year or two ago, and so I just kind of use what she gave me. And I'll try to add a new slide that's sort of in between some things, and I'm kind of redesigning it and kind of swapping what's on what slide, but I'm trying to fit it into the structure that's already there.
And so, the way I'm just thinking about how this might play out with Gamma, do I basically type in a chat box and I can just say, "Hey, can you split this from two slides into three? Or can you change the color? Can you add a chart?" How does that usually play out when I'm interacting with it?
Grant Lee:
Yeah, totally. I think we want eventually to be just so many different entry points, because sometimes it's like you have just a topic and you just want, "Give me something so I have a rough template, I can start with, or a really good first draft. That's fine." Other times, maybe you've done some research, maybe it's in ChatGPT or a different tool, and you've actually crafted the end to end narrative you want. And then how do you visualize that? You can drop that entire thing into Gamma.
And then there may be an instance where there's maybe existing documents like a Word Doc, a PDF that you're working off of, something that was long form that you want to make more visual. And we want you to be able to upload those documents and say, "Hey, take this a 100-page document and help me turn this into a five pager because I need to get in front of an audience and communicate this more effectively. And I know a lot of the information, but I don't know the best way to turn that a 100-page document into something that is a little bit more visual and consumable."
And so obviously many different entry points, we want all those to feel effortless. You should be able to do that no matter what the occasion is. And Gamma can help you then shape the rough ideas into something that's more refined.
Turner Novak:
And this is just a question I had, thinking about, as you said, generating about 50 million in revenue, I think it said 100,000 or a couple of hundred thousand customers. Is that right?
Grant Lee:
Yeah, a few hundred thousand customers.
Turner Novak:
So what are people paying for? When you're building this, you're like, "This is free, this is paid." How did you decide maybe those processes doing that, but how did that kind of all evolve over time?
Grant Lee:
Yeah, so we basically have a freemium model. It's based on today AI usage. So, all new users get a bucket of AI credits, and then when they use that all, they can either stay on the free plan, which allows you to still use Gamma, but you have to basically manually update the content and do it the quote, unquote, "Old-fashioned way" by as a human still entering and playing around the building blocks. And even that has been useful for a lot of users, because those core primitives still allow you to manually create stuff much faster than you would in other tools.
And then if you need more AI guidance or assistance, then you can upgrade to one of our paid plans. We think of AI as being the sort of expert design partner for you, as if you're sitting right next to a great designer and you can actually have them help you craft and shape your content quickly. And so, we want that to, again, feel effortless. And anytime you need that, you can just tap your design partner, in this case Gamma, and they'll help you really refine your content faster than you could do it on your own.
Turner Novak:
So it's like instead of just emailing my designer back or texting them back and saying, "Hey, can you change this thing, change the colors, I want to go this different direction?" the product just kind of does it for you?
Grant Lee:
Exactly. Yeah. I think today, a lot of the bottleneck is either you don't have a design designer on staff at all, or maybe sometimes you outsource it to an agency. Or you just don't have much of your own design skills, and so the capabilities are limited, and so what you have to do is try to outsource it. We want to replace the need for that for the vast majority of use cases where you could just get by leaning on Gamma a bit more.
And so many people we talk to, especially in a little more corporate environment, their bottleneck is like, yeah, even if they have a design team, they oftentimes don't have enough of their time to be able to do a lot of the iteration back and forth. And so, that time is reserved for the most important super high-polish outputs, but the day-to-day presentations or decks, or things that need to be created, there's nobody there to help with that, and we want to be able to step in and serve those users right away.
Turner Novak:
And this is something maybe you did unconventionally. So, you talked about a pretty horizontal approach. I feel like generally in startups and the VC kind of conventional wisdom is go super niche, super vertical on a specific thing. And you didn't do that. I'm just curious why, because most people the advice would say, "Oh, pick management consultant specifically," or founders, raising their pre-seed round specifically. So why did you go so horizontal instead of super niche?
Grant Lee:
Yeah, I think there's two parts to this answer. The first is really just trying to think about our own founder ambition. We set out to create a new standard. We want this tool to be ubiquitous. We know that's super challenging. It's not going to be easy to create something that even has a chance at replacing a daily active tool like PowerPoint that's everywhere around the world, but that was our ambition. We believed we can create a better way. We want to create a new standard.
The only way to do that is to create something that can be horizontal in nature, such that if you're an educator, you can use it. If you are in the corporate world, you're a marketer, you can use it. If you're a consultant, freelancer, solopreneur, small business owner, all of these types of users should be able to benefit from what we're creating. And so that was our ambition. From day one, we knew that's what we wanted. And so, horizontal was the only sort of mechanism of trying to fulfill that longer term vision and dream.
And the second thing is, even if that were the long-term vision, oftentimes people will say, "Okay, we'll just start with one ICP. Just try to serve salespeople," for instance. I think that ends up becoming much more challenging than sort of... It's easier said than done, right? Because what ends up happening is you try to serve that one ICP pretty well, and inevitably they have very specific things that end up needing to be tailored to their use case. And so, what you end up doing is creating a specialized tool that maybe the salespeople do get a ton of benefit out of, but when you start zooming out, none of what you've built over the past couple of years are then beneficial to the consultant or beneficial to the educator.
And so inevitably, you've already created this verticalized tool or vertical tool, even though your long-term ambition might've been to build horizontally. So when we thought about just the sequencing, we knew that both address number one, which is our own founder ambition. And to actually have a shot at that, we had to resist the temptation of just serving one ICP well. We needed to have a pretty broad swamp of users.
And the way we sort of acknowledged that, was rather than focus on personas, we focus on pain points. So not ICPs, not personas, but really the pain point that a user might have, and what are those universal pain points? And for us, those universal pain points are it's super hard to get what's in your head into a format that others can visualize. Getting what's in your head into other people's head requires this sort of visualization piece that is really, really hard, and that's a universal pain point. How do we do that well? That's what we're focused on doing.
Turner Novak:
So do you think this is applicable for everyone, any kind of category of company creation? Or do you think it only works if there's an opportunity to build a super horizontal tool that specifically has pain points like this?
Grant Lee:
No. Yeah, I definitely think this is definitely not the only approach. There's some markets out there where the TAM is probably big enough just to go after that vertical ALONE, and maybe it's just so underserved too by incumbent tools, that it's almost a flip of this, which is maybe incumbent tools are more horizontal in nature. That vertical is actually a very, very large juicy TAM, and to have a verticalized tool is actually the right approach, especially if the founders that aligns with their ambition like, "Hey, we want to serve the legal market," for instance, "And that, we believe, is a big enough market. And incumbent tools are going to be too generalized to not serve that market well, and we believe it's a massive TAM, let's actually build specifically for that."
I think there's definitely merit in doing that, and that makes sense for certain markets. Founders, they should always ask, "Is that market itself big enough? Is it growing? Is it expanding? Is a horizontal approach fundamentally limited in serving that market?" And it's a good sort of thought exercise before, just assuming you should go through the exercise of just asking, "What do you believe to be true years from now?"
Turner Novak:
Yeah, and actually super crazy on that point of legal market TAM. So I think I saw a tweet from Patrick Collison at Stripe. I believe the stat was something like, you know when you see those billboards on the side of the road for personal injury lawyers where it's like, "Oh, did you get an accident? Call this number." It's like 1% of GDP or something in the US. Just that market of suing people after getting into an accident or something.
Grant Lee:
And it's growing, and that's what you want to look for. It's already a meaningful amount and it's growing. Then it's like, "Okay, well, there's probably something there." You want to move where the markets are experiencing some sort of organic tailwinds and then try to capture that in that moment, too.
Turner Novak:
What is the tamper, just slides and presentations? I actually didn't even look this up. Do you know what it is?
Grant Lee:
Microsoft Office doesn't really share too much in terms of just the revenue, but it's in the billions. And when you look at the user base of Slides, it's in the billions. And the majority of that is PowerPoint, but you have Google Slides and the other incumbent tools. And so, even just the user base alone is in the billions.
And I think for us to be successful longer term, yes, we're going to carve out a good chunk of people that are maybe using incumbent tools, but also then create just new market opportunity. There's a lot of people that even Slides are something they're super well versed in, so they don't create them. But in the future, when digital content's going to be even more important in terms of connecting with others or customers or partners, they need simpler tooling. We can be that tooling for them.
And for those that had never even adopted, really ingrained PowerPoint into their workflows, Gamma can be that solution for them. And so it's almost like creating a brand new market that just has latent demand but hasn't yet been fulfilled by an existing product.
Turner Novak:
Yeah. How do you think about just generally this whole office tool workplace? When you just think about meeting summaries, presentations, docs, calls, there's a lot of people kind of playing in this space. And I think if you just kind of fast-forward, you think 5 years forward, 10 years forward, they're all kind of converging. I don't know if there's a strategic reason of starting with Slides or if it was just curiosity and interest, and then how you just think about it over the long term, how everything's going to play out?
Grant Lee:
Yeah, the challenging thing is things are evolving so quickly, hard to know how things will all pan out. We started with Slides, because one, was actually just trying to solve my own pain points. So the last startup I was at was acquired. I went back into more of an advisory and consulting role where I actually started my career.
And I was staying up late at night trying to actually put together, synthesize information back to a client, back to a customer, and just dreading that moment of actually going into Google Slides, having to spend so much time formatting. And even just staring at that blank page gave me the first moment of realization, which is like, "There's got to be a better way. Slides as a format created almost 40 years ago and it's such an important medium for communication. Is there a different way we can approach this?"
So that was the initial, "Okay, something interesting." And then what we actually ended up doing was just building for several months, prototyping the initial versions of Gamma, just to give ourselves our own sort of level of convictions. We started realizing that Slides for most people isn't probably a super high MPS sort of product. It's one that people have just gotten used to the way things have always been done but never questioned, "Okay, what if the primitives are building blocks are fundamentally different?"
And when we thought about just the growing need for digital communication, we wanted something that acknowledged where things were already moving. So, Slides as a format fixed in aspects ratio or dimensions, and made for an era where we're all staring at the same projector or we needed things to be printed on pieces of paper.
And when we think about where things were already heading, much more of the communication can be async and viewed on different devices, like your phone or tablet or a monitor. And so why not have something that was more mobile-friendly or mobile-responsive by default? Why not have something that was more interactive and multimedia rich? And all the things that PowerPoint you might be able to hack together, but it wasn't built for that.
"What if we lean more into where we think things were already going to be going?" And we just saw such a massive opportunity that it just felt like, yeah, even if we didn't know necessarily all the answers on day one, it was going to be a fun journey to go on and it'd be plenty to build off of, and we'd learn and kind of iterate along the way.
Turner Novak:
Yeah, I think to your point of how the format has kind of evolved. It's always been this kind of landscape view rectangle thing, but if you think, probably half of emails are open on phones. I don't know if the stat is, but you somebody's looking at a vertical screen on their phone and then you basically limit the size to be a third of the screen in this little rectangle in the middle. I remember the first time I saw a vertical deck, I was like, "Oh, shit, that was actually pretty smart. That's a good idea." And this was probably like 2019-ish. And so this was probably, you were doing this around 2020-ish?
Grant Lee:
End of 2020 is when we first formed the company. Yeah.
Turner Novak:
Okay. And then I think with this whole horizontal vision, I think that kind of impacted how you raised a little bit of money in the beginning, in the early days. How did that kind of play out?
Grant Lee:
Yeah, we knew we needed to raise enough to assemble the early teams. I mentioned even just the initial seven, and then we got to a team of about 12 before we did our first bigger AI launch. We knew we needed enough to build the scaffolding for the initial product, then to build able to get that into the hands of users and iterate.
And so, we raised a decent size pre-seed round, it was a little bit under three million and that gave us the chance to form the initial team, and then we ended up raising another four million as part of our seed round. This is when we had basically just launched our private data.
And again, I think part of the founder journey, for anybody that's just getting started, is try to look at other companies that are in your category or might need to follow a similar path. So for us, that's like other horizontal platforms that have been created before us.
And so the Notions, the Airtables, even the Figmas of the world, which all took many, many years to get to a point where what they were building didn't feel like just a toy. It was something that could be used for real work. And I think we always acknowledged that upfront.
Even something as simple as presentations, there's a ton of surface area to cover. And so, to earn the right to be a tool that's actually used and depended on by somebody outside of your organization, it takes a ton of effort. And we knew it would take us some time, and with time means you need enough runway, with enough runway means you need enough cash in the bank. So we ended up raising in a way that just try to factor all that in.
Turner Novak:
How'd you get people to buy into it? 'Cause it's a horizontal tool. Most people be like, "This won't work." It's impossible. Go to market. Won't work. I don't like this idea." How'd you get around that?
Grant Lee:
Yeah, we definitely got a bunch of rejections. It wasn't everybody, but I think what the ones that believed in us, there was probably two things that maybe caught their attention. One is, when we created our very first pitch deck, we actually built it as a prototype of Gamma. So it was basically showing what the possibilities were, even if it was super rudimentary. And so it had those elements, it was hybrid sort of writing, a document/deck-like medium. It was interactive, it was multimedia rich. And so, it had all the principles we were sort of trying to lean into, and so people could squint their eyes and maybe see a little bit into what the feature we're trying to describe.
And then the other half was people that were the prototypical user, so Amy who led our seed round, she basically was product marketing manager, was at Stripe, was at Twitter early on. And so, face the bottlenecks we were describing, which is they had to create a lot of content. They don't always have design resources, and so you're just limited by what you can do on your own and then just constantly waiting.
What if, with a tool like Gamma, you could self-serve much more of that? And for the average professional knowledge worker, they can go into a tool and actually craft and shape their ideas? That was exciting to her. And so I think it was a combination of those two, where do a little bit of showing, not just telling, but then also resonating with somebody that just understands your vision and why your product deserves to be in the market.
Turner Novak:
So how do you find the right investors then in that kind of situation? Did you do a lot of outbound? Did you spend a lot of time getting to know people in the years before? Is it a volume game, just getting in front of the right people?
Grant Lee:
It's definitely a volume game, but to your point, it is a game. And so, the way I think founders just structure the game, early on you want to just try to build relationships. So the best way to go into any fundraise is obviously you already have some warm relationships. The other best way is also to go in with momentum.
And so, if you're super, super early, the way to show momentum is maybe going with a prototype, something that already shows that you've been thinking about this problem, that you've built enough conviction yourself.
And then, when you start pitching early investors, you're going to get a ton of nos. The folks that say yes, you can ask them two questions. "One, what is it about your pitch that resonated with them? Was there something that really stood out? Why are they choosing to invest?"
And two, you can ask them, "Who else should I be talking to? Who else do you think this message or this pitch might resonate with?" And you're just doing that over and over again.
The funny part about our own founding during our pre-seed round, this is end of 2020, peak pandemic. I was actually living in London at the time, and so I was doing all of my pitches late into the night to 1:00 AM period.
Turner Novak:
Like Pacific Time. Yeah.
Grant Lee:
And so, I was locked in, and we didn't have a very big place in London, renting out a small apartment, so I was stuck in the sort of laundry room because that was the only place that was soundproof enough. All the rest of the family could sleep, and I would just do five, six hours of pitches back to back, just straight Zoom.
And I was fortunate, because peak pandemic, that was acceptable practices, just do back to back Zooms. And I would just rinse and repeat the strategy of meet somebody, they said, no, move on, meet somebody, they say yes, ask them those two questions. And then line up back to back basically two weeks straight worth of meetings to a point where we had enough to fill the initial round.
Turner Novak:
Amazing. Yeah, it's definitely not glamorous. A lot of people think of fundraising is this wine and dining, and it's basically just like eating shit until you get enough of the way there. Usually, that's how it goes.
Grant Lee:
Totally. I would say, I don't know, today it feels maybe slightly different, but yeah, back then that was absolutely the mindset I went into, which is, "Two weeks are going to be brutal. I've just accepted that, and I'm going to stay in this laundry room until things are done."
Turner Novak:
Yeah. Well, and then to the point, I think you guys are doing really good today, but you've never really raised big, sexy, hyped up rounds. VCs don't really talk about you. That's usually a signal that there's a lot of VCs invested, is when they're always talking about something, like hyping it up. Why did you never raise a ton of money and go down that path?
Grant Lee:
Part of this is just I think the mentality of the founding team that we've always had, this notion that raising too much money has a ton of potential downsides, especially if you do it too early. Because I think what you want early on as a founder is, one, to build good habits amongst your organization, things that are the foundation for how you want the rest of the company to operate in the future, and the mindset, what is that sort of scrappiness mindset?
I think PG talks about the cockroach mode that you can go into. And yeah, cockroaches don't need much to survive, and what can you do to make yourself super resilient as a team? And you don't build resilience out of just luck. It's a lot of day-to-day, "This is the grind." And part of that scrappiness needs to come through with just how you operate.
And I think that naturally then bleeds into the culture you have. You want your culture to be one where early on you're building the early parts of the team. And then later on, you want that culture to be solidified such that you hire the next 10, 100 people, they all have the same DNA. It's all the same. "We know what we look for, this is what Gamma represents, these are the type of people that fit in well here. And then the people that don't, that's fine, we're going to move on." And so we wanted to really invest in that.
When you're also early on, constraints, in my mind, begets creativity. You're building a tool that's horizontal, you need a ton of creativity. Forcing yourself to raise a bunch of money almost inherently means you're going to take a more lax approach and you're going to say, "Okay, money's going to help solve most of these problems. I'm going to hire the person to do the job," versus do the job yourself, and recognize that most of it is in your own control and you have probably enough of the ingredients within your core team to do a lot of the work.
And so, those are all things that we talked a lot about early on. And anytime there was an opportunity to raise more, we'd ask ourselves, "Do we really need it? Is it preventing us from achieving our long-term ambition? And if not, we're going to keep heads down and keep building," and we really haven't wanted to divert from that core mantra.
Turner Novak:
How has that impacted being able to recruit? 'Cause I feel like part of the reason people raise money is so that there's a halo around it from a recruiting perspective, but also just, "Oh, Facebook will give you a million bucks, we'll give you more," or whatever the strategy is there.
Grant Lee:
I would say it's made recruiting harder in some ways, but it has, I think, a ton of upside and value in other ways. When you are forced to play the dollar for dollar game, and you're competing against the giants of the world that can give insane packages, you end up inevitably finding people that are much more mercenaries than they are missionaries.
And so for us, we always look for people that the core product, what we're doing resonates with them on some level. And for instance, if you're coming in and doing an interview and you've already spent time in the tool and you're like, "Wow, even if I don't get this job, I may go on and use this tool." Something about it has resonated.
And those are the type of people, that we've found that once we bring them in house, there's so much love and energy that they pour into the product. And of course, we want to have super competitive offers, and we want to pay top of market and give grid equity, and those things all come packaged, but at the same time, only trying to satisfy the competing offer of a massive AI company, I think in many ways, sets you up for long-term obstacles that you may not even can even foresee.
So, what we're trying to balance now is, "How do we"... One of the biggest benefits of raising is you set the market rate. And so people can say, "Hey, what is the value of this business? How do I even think about equity?" Because equity for most people is still super opaque, hard to understand, and we try to spend a lot of time providing education.
Even this comes from my finance days, but even at Optimizely, I spent a lot of time just trying to educate people around, "How should you even value your options? What are the different considerations?" There's a lot of the fine print, that if you don't really understand the differences, then what's on paper may not actually translate into true value for you.
And so, we do the same with all of our candidates. Anybody that has competing offers, I try to provide a super objective, "Hey, this is how you might think about things." I'm not trying to overly sell them, I just want them to know how they should actually value certain things, such that they can go in and make an educated informed decision.
And I hope that many more founders will do that, which is take the onus of actually trying to educate your candidates as much as possible so that whether they decide to join you or not, many years down the road, you feel like you can sleep at night because they made the right decision with all the right information.
Turner Novak:
Yeah, I think too, there's the value of just the tangibility or the realistic chances of the equity being worth anything, just at the end of the day. If it's a public company, it's a liquid. If it's a startup, it's private. There's a little bit of uncertainty around, "What is this thing actually worth really at the end of the day? Can I get anything for it? If I needed to sell, you might have to take a discount."
But just generally, I think you guys have raised a total of $23 million, I think? So, I don't know, just generally speaking, there's usually a 1x liquidation preference on the equity that investors put in. So, if I'm somebody who's considering working at Gamma, it's like, "Okay, if the company sells, the investors get their 23 million, and then we all get our pool," that's a pretty low bar.
You're doing 50 million in revenue. I don't think people are going to be valuing you at half of ARR. Maybe you're a little bit higher, 10x. I usually think 10x ARR is just fairly reasonable number for just a healthy growing software company. So yeah, 500 million, that's pretty reasonable. You guys are in a great spot from that perspective, too.
So, I feel like not a lot of people think about that. If the company gets acquired, if you had raised a billion, and we'll just use that same number, and say you're really only worth 500 million, everyone gets nothing. The investors don't even get their money back. So there's a lot more that goes into it than that. But that's generally a high, first look is you just think, "How much has the company raised and is it even worth that much?"
Grant Lee:
The math you just described, less than 10%, less than 5% of candidates ever even get into that math. And to your point, yes, it's like when you break it down, there are things that seem to be, you can present that in a logical way and then that decision ends up becoming easier.
But the tough thing is, when you are competing against someone that's raised a ton of money, they do have that external credibility or validation that somebody believes they're worth x billion, x $100 million. And oftentimes, candidates will rely on that much more than anything else because it's just easier to understand, and it's quote, unquote, "A trusted professional investor that has done their homework and has put that money in."
I think for us, we as founders have come through a few different kind of eras of startups, and we've seen both the good outcomes, the mediocre outcomes and the terrible outcomes. And oftentimes, the mediocre and terrible outcomes, the founders and investors maybe still do fine, to your point, 1x, they'll get their money back. Founders might still get a little bit of something.
In those cases, then employees almost get nothing. And I've been both the employee and now a founder, and a lot of what I fear is I want... In the future, if Gamma has a great outcome, every employee should be there to celebrate. It should be something we are celebrating together. We are all owners of the business.
And that can only happen if you're being thoughtful about that and trying to structure things the right way from day one, not waiting until you've already made it and have huge success. It's like you need to be thinking about day one, such that in all outcomes, if the company has some meaningful outcome, everybody can celebrate.
Turner Novak:
And this is a little bit of a sidebar, but Optimizely, what was that company again? I've never heard of them before. I've actually never heard. So I'm just curious, what did they do and how did that go for you guys as a team? It sounds like you all came from there, the early team/
Grant Lee:
Yeah. So, Optimizely, it was actually one of the fastest growing YC companies back in the day. And they started off sort of in this space of A/B testing, and eventually kind of brought in that scope to be much more around just experimentation, broadly. So, we were serving companies like Nike, Best Buys, these large companies that had to do a lot of digital experimentation, and supporting them through our own platform. And so anything that needed to run experimentation at scale could lean on Optimizely to do that.
And I'd say the company, when I joined, it was series A and a little bit over 10 million in ARR. By the time I left, 100 million in ARR. So, by all means, it was a business that definitely showed that there was a market for this, and had grown and had global footprint. And I think the outcome maybe wasn't as great as what we would've all hoped.
Being one of the fastest growing YC companies at the time, I think the aspirations were to be something that super impactful, and again, a great outcome for employees, but we ended up being acquired. And I think one thing that a lot of now the Gamma early employees is we hope that one day if we do have a great outcome, it's something we're all celebrating and it doesn't feel like we had left something on the table.
Turner Novak:
Yeah. And I feel like that's kind of a little bit of a superpower. I'm assuming you all work together at Optimizely in some capacity. So it's like, "All right, it's round two. We already kind of know how to do this, how to work together"?
Grant Lee:
Yeah, there's definitely a bit of that muscle memory working together, knowing each other's styles, things that you almost don't need to say because it's just so obvious of how we work, and that obviously just improves how frictionless things can be.
So, it's been extremely valuable. And in all the first seven I mentioned all still at Gamma., Actually all of our employees, we haven't had an employee leave. And so, it's this continuity piece that's been really helpful and something that we want to continue to sustain.
Turner Novak:
That's an insane stat, 30 people, no one's left yet over five years. I feel like that's got to be a record. I don't think I've ever met anyone that scale over that period of time that hasn't had anyone leave.
Grant Lee:
Yeah, I think it keeps the core team super, obviously the collaboration really tight, and it goes back to being able to move as fast as possible.
Turner Novak:
Yeah. Well, and speaking of the team, so you were in London in the beginning. Were you guys all all over the place, or was it just you that was remote? And I know now you're all in San Francisco. Right? How did that play out and what did you learn about that evolution of how a team functions remotely versus in person?
Grant Lee:
Yeah, so I was in London actually for my wife's work. We were there for a couple of years. And first started Gamma, fund raised, I was out there. The rest of the team was all in San Francisco. So again, this is peak pandemic. So we did have the question of, "Should we build a fully distributed team or should we try to work together in person, or some sort of hybrid environment?"
And when we actually asked our first, I mentioned the first seven, and eventually several more than that, "Where should we be based?" Or, "What is your optimal setup?" Every one of them said, "We want to work in person in San Francisco, no desire to leave." And I was moving back eventually anyways, and so the decision was sort of made for us.
And then basically we went into, "Okay, we know we're going to work together, but in San Francisco we're still sheltered in place, so we're going to find a place that allows us to do this," which is we found a two bedroom apartment, converted two bedroom apartment. We stopped ourselves there and worked in different corners of the office with huge air filters. And that was kind of the first year, I was just doing that in person because it felt right. It was the right way for us to build, and we were just energized by being together.
Turner Novak:
That was pretty bold, because I remember back that first year, I mean especially the first month or two when COVID kind of hit, it was literally, you look at someone with COVID, you're going to get it. We didn't fully understand. So, was it pretty easy to convince people? And were you guys kind of all trying to bubble and protect from the outside world, or?
Grant Lee:
We definitely tried to do it in a safe way. So again, different quarters with air filters and whatnot, but I think part of this is because we had worked together for so long, it was easy just to be candid, like, "Hey, is this something you're comfortable doing? And if not, totally fine. Work from home." But it was something that, at that stage, way easier to have those conversations. And because everybody was fully aligned, it just made sense. It was the right call.
Turner Novak:
Yeah. So I know the first, you mentioned you did a private beta, you did a public beta, you did a launch. What were kind of all those different iterations of the product over time?
Grant Lee:
Yeah, so we started off early on doing basically a private beta. I was just putting up a landing page and then creating this wait list, and it allowed us just to see who's even interested in what we're building. And we were collecting, I think early on is 4 or 5,000 signups. And we could see kind of the personas and what they were looking for. We had a few different open-ended questions just to gauge, "Okay, what is the pain point we're solving for this particular person?"
Turner Novak:
So you were still building it or was there-
Grant Lee:
Still building it. Yeah. It was very, very rudimentary. It had the very basic ingredients of you could use it, you could present it. Was it super reliable? Was it super versatile?
Turner Novak:
Questionable, yeah.
Grant Lee:
Yeah, maybe.
Turner Novak:
But you kind of use the wait list signups as almost like getting some data on who was interested in potentially trying something?
Grant Lee:
Totally, yeah. We didn't drop everybody from the wait list directly into the product. We basically had batched them up. And so, every batch we let in a few hundred and we see what happens. And for the people that were more engaged, we would follow up and start talking to them and ask them, "What is it about the product that they were finding most valuable? What were the biggest things that were missing? How were they trying to use the product?"
And we kept doing that through the initial set of wait lists until we started seeing a pocket of users that was just organically coming back on their own. And they weren't our friends, they weren't people in our family that were just being nice, but there were people that just heard about the product and found some value in it. And that built up enough conviction to say, "Okay, we're ready to do our public beta launch, which is we're going to remove the wait list. Anybody that comes in can sign up, gets dropped directly into the product."
Turner Novak:
So how did you qualify that and know that it was time? Was there, "Okay, we have 5,000 retained daily or weekly users," how did you know it was the right moment?
Grant Lee:
Yeah, we didn't set super aggressive usage metrics. We had basically enough people coming back, where I'd say it was below thousands where people were just using the product. And the feedback, there was enough pockets of users were like, "Oh, this is great." And then they're trying to pull sort of features out of you. And so it was more of that sort of feelings like, "Okay, there's something here."
And then we gave ourselves four months to essentially prepare a public beta, which is add a little bit more polish, extend the functionality. We launched on public product time, which gave us also a chance to exercise the muscle of launching and getting word out there, and just see what it takes to even announce a product, and flexing that muscle a little bit.
And so yeah, the product launch ended up doing great. Those are, again, our public beta launch. We ended up winning product of the day, week, and month, and so gave us that little bit of extra validation like, "Hey, people are digging the early version of the product."
Turner Novak:
So when was this? Beginning of 2022, middle of 2022?
Grant Lee:
End of 2022. Fall of 2022. Yeah.
Turner Novak:
Okay. Fall of '22. So this was around the time ChatGPT came out, maybe a little bit before?
Grant Lee:
A little bit before. So the launch was pre-AI. The product was still manually creating things. And then, what we started realizing was obviously AI was moving incredibly fast, both on the image side as well as the tech side. And most presentations are a combination of texts and visuals. And so we knew we needed to basically revamp our entire creation flow so that we could integrate AI into it.
Turner Novak:
'Cause yeah, I'm sure that a lot of people listening to this have gone through this or thinking... There's probably a lot of people thinking about this, like, "We need to re-architect our product for AI."
I do want to spend a couple of minutes just talking about how do you do that? So, was the moment that you realized you needed to do it was when it was right within the months after ChatGPT launched and you realized the different texts and image capabilities?
Grant Lee:
Yeah. Well, we started by just playing around a bunch more ourselves internally. And very beginning of 2023, we gave ourselves, again, this four month sprint idea of, "We're going to do our first AI launch. We, in the meantime, are going to completely rebuild the onboarding experience and the first user experience."
Turner Novak:
Why was that important?
Grant Lee:
We knew that we needed to get people to understand the value of the product immediately. We didn't have time to get them to slowly understand or adopt the product. They needed to see that in the moment, and that would kind of benefited us in two ways. One, if they see the value and they activate, then they're more likely to obviously then retain themselves. And two, if that experience is so magical, they're going to tell other people about it. And there'll be an ability to generate this sort of word of mouth immediately, this organic word of mouth, which is obviously a powerful lever.
Turner Novak:
So it's just get that right in the beginning, the first minute or however fast you got it to?
Grant Lee:
Exactly, yeah. 'Cause most of us are the Scott Belsky, selfish, vain and lazy. Most of us are that in the moment of trying a new product. And so, unless it resonates and that first mile is something that could be magical, most people were never going to come back. We knew our current, at the time, our current onboarding was a little bit just to required the end user to do so much work to just even grok what we were doing, understand any amount of value. And the learning curve was so steep that we were going to lose too many people right out the gate. And so how could we flatten that learning curve dramatically and make that experience feel a little bit more magical? That was the whole goal of that four months.
Turner Novak:
So, it sounds like you simplified the onboarding, made it like there's a quicker, I don't know, time to value or time to magic moment, whatever people want use to describe that. What advice would you give if I'm like, Hey Grant, our onboarding sucks, it's too long"? What should I do to get a better onboarding or faster, quicker onboarding experience?
Grant Lee:
Yeah, it's going to vary a ton for different products. Sometimes you actually want to infuse maybe slightly more friction because it allows you to better qualify the type of users that you can serve. And then maybe through either personalization or a unique onboarding, based on how they're maybe responding to different questions, you set them up for success. So you're taking some of those inputs.
For others, you just want to get them right out the gate, understand enough about the product so that they come back. And I think we fell more into the latter, which is we removed much more friction, we gave them a starting point, which is, "Just enter some general idea of what you need to create and we'll show you what a first draft could look like."
Get them to understand, "Hey, Gamma is actually a different way of creating content, but the building blocks don't need to feel intimidating. We can get you there and you can start playing around with it within the first 30 seconds of onboarding." So that was more of our goal.
And I think most founders need to go in and understand balancing that out. In some cases, maybe adding a little bit more friction or getting people to integrate, or getting people to really get plugged in so that their first experiences feels like they're set up for success versus a little bit too haphazard, like, "Oh, this is a toy, not a real product" sort of situation.
Turner Novak:
Did you personalize the onboarding a little bit? Based on an answer, you'd tweak what path they went down?
Grant Lee:
We haven't yet done much of that, but this is something that we're starting to explore more, is around personalization, segmentation, RV one of this. We didn't try to do that. The way you kind of went and chose your own adventure is just by nature of the prompt you would put in. So if you're creating something for a sales template, we'd give you a first draft that felt very relevant and it didn't need to be overly engineered based on, "Oh, are you a certain type of user in a certain type of geography?" We just let it be more prompt-driven.
We're now at the stage where I think there's other layers of adding that friction where it may make sense. And with that, comes the ability to personalize and get people even more unique starting point that puts them in a place where they can actually see themselves using the tool after the first run.
Turner Novak:
And I know there's this whole saying of if you're building an AI-native product, with every new model release, you want it to make your product better. Did you benefit from that over time?
Grant Lee:
Oh yeah, of course. Yeah, we are still benefiting from that. The advancements, it's cool to work on a product where the surface area obviously covers a ton. We have text, images, we'll have video, we'll have audio, we'll have many different formats that we'll be able to support.
And because there's been advancements on underlying technology across the board, anytime there's advances we can say, "Hey, what are we going to pull in? How does this get passed to our end user in terms of value delivered?"
And it's just a constant sort of iteration, learning, putting that back in. And yeah, just the basic equation of every business is pass as much value to your user as possible, or your customer as possible, at the lowest price, and you can build something pretty sustainable.
Turner Novak:
Yeah, 'cause I could imagine, even if I'm thinking about using Gamma for this podcast, and I think about maybe your roadmaps and things you just indicated, I could send you a link to something that's kind of like, I don't know, onboarding you to the peel or giving you the TLDR. You'll read this for two minutes and we'll get you going.
It could literally be a custom AI generated avatar me saying, 'Hey Grant, it's Turner. Super excited about Gamma. I'm excited to jump into the podcast. Here's things." You could probably even incorporate a prompt where I'm like, "I want to talk to Grant about building the product and advice for people on building AI-native software." And you can tailor that into what comes in through the actual link or presentation that you're getting from me. It probably changed dynamically too, I'm assuming?
Grant Lee:
Yeah, all of that. We're just barely scratching the surface. I think at the end of the day, tools like ours, they're communication tools. And we obviously infuse a bunch of creativity into that as well, but allows you to express yourselves in ways that just weren't possible in the past. The sort of one-to-one versus one-to-many. The sort of automated, the personalized, all that becomes easier over time. And yeah, we're just getting started.
Turner Novak:
It's kind of interesting that the core insight was that Slides are sort of like a writing exercise, not a design exercise, that that got even more pronounced as you built it to be more AI-native, 'cause you're not really designing, you're just saying. You're texting it and saying, "This is what I want you to make." And there's probably a luck element to that, I'm assuming? You didn't know ChatGPT was coming, you didn't know the LLMs were going to get this good?
Grant Lee:
No. Yeah, we did not. I think it was great that we had a writing first approach, because that has naturally extended the fluidity that you can work in Gamma because you are already in that flow state of writing, describing. And then in some cases, you're leaning on AI to be more generative, and can help you extend your ideas and shape it in different ways. So yeah, definitely.
I think where sort of luck meets the preparation piece of it is that our vision has always been around really simplifying and making it super easy and fast for the end user to shape their ideas. And just so happens, our approach, married with AI, has been one where we're actually able to really fulfill on that promise much earlier on.
Turner Novak:
Was there anything that you wish you did differently when you redesigned it? When you went back from scratch, "We're going to build this completely AI-native, re-architect everything"? I don't know how much the design changed, but anything specifically that if you go back and tell yourself two and a half, three years ago and just say, "Just don't do this one part or skip this or something"?
Grant Lee:
It's so hard to go back and say, because I think for us, and maybe this is because we came from Optimizely, so much of how we approach things are iteration, experimentation, testing things out. And so, we get a lot of things wrong, but that's built into the system and how we build things.
And so yeah, for sure we probably could have navigated differently, but our approach probably would've been the same. And so, we would've still tested, get things into the hands of users, test our assumptions, our hypotheses, and then go back and refine that and just be relentless about doing that over and over. And so, yeah, I don't think it would've changed much. I think it was just baked into how we operate.
Turner Novak:
Yeah, okay. And I know you have about 50 million people that use the product today. You talked about there's a wait list, you did the public launch, maybe a couple of thousand, maybe tens of thousand people. That's a pretty big gap. So, how did you go from a couple of thousand people using it, to a pretty significantly large product today? Anything specifically that happened over the past couple of years on that side, on the growth side?
Grant Lee:
Yeah. I'd say the main inflection point through it all was that four month sprint leading to our first AI launch. And so this was 2023, spring 2023 getting ready, we completely revamped the onboarding flow. And at that time, we had about 60,000 users and we launched. And then all of a sudden, we're adding 5,000 users a day, 10,000 users a day, 20,000 users.
Turner Novak:
Oh, so it was increasing?
Grant Lee:
Yeah, yeah.
Turner Novak:
That's cool.
Grant Lee:
And it kept on increasing, and so that's where the flywheel started spinning. And that was really the initial inflection point. And we've been lucky that since then, growth has remained steady and we want to just build on that momentum.
Turner Novak:
Is it like somebody shares it, like you send me a presentation or a link and there's, "Made with Gamma, try it out?" type thing. Do a lot of people come from that? Or is it they go to Google and they just search Gamma, "What is this?" Make a slide, AI-generated slides"? Is there certain channels that seem to work better as you really scaled up?
Grant Lee:
Yeah. There is a good chunk of when you share a free Gamma, you'll see the made with Gamma badge, and people come through that. The vast majority of it is more through word of mouth. So someone might say, "I'm using Gamma for this." Or if you're presenting for instance, you'll see that it's being used in Gamma, being served by Gamma. And so, people will then just inherently ask, "Oh, what are you using?"
And then people that really love the product are telling all their friends, and so then their friends will go and search for it or come direct. So that still ends up becoming the most, for us, the most meaningful channel, is just strong word of mouth.
Turner Novak:
And then how do you think you stand out today as an AI company? What's the most important thing? 'Cause there's a lot of people doing a lot of different things, they're all growing pretty fast, or some of them are growing pretty fast. What becomes most important when everything's kind of AI-native?
Grant Lee:
What becomes most important is still the question of how much is the human in the loop? And are you creating something that completely replaces the human or can the human have the ability to actually still be part of the experience and edit and shape? We're actually approaching that from, I think, a few different ways.
For our core prosumer product, where the end user, it needs to be a delightful experience, you go in and you can get a great first draft and then the human can continue to nudge and push things in the right direction. We want that to feel, again, effortless. It should feel like something you go in, it feels nice. It's not the PowerPoint moment I described, which is like, "I'm really dreading going into this, and I worry about touching anything 'cause it's so fragile and it's just going to mess up the entire layout." We want that to not be the case.
And so our, I think unique advantage, is human and AI working well together. And as the AI gets better and more capable, it can just really guide that experience too, but the human is the driver, still the person navigating.
I think where we're going to be exploring much more automation is via our API, where being able to allow others to build on top of Gamma, and where maybe things can actually be almost fully or almost entirely automated, and where a human doesn't necessarily need to go in as much, and it's more about how can you create content at scale reliably, predictably, and then what does that really give to you?
And so I think that might have a different set of constraints overall and maybe an overarching different approach, but that can also be equally powerful and we want to be able to serve both audiences well.
Turner Novak:
So what's the API? Sounds interesting. I maybe have an idea of it, but that sounds kind of interesting. What is that going to be?
Grant Lee:
Yeah, they'll be a few different flavors of it, but essentially it's being able to allow people to combine their first-party data and feed that into something like Gamma where we are now the content engine. So no matter what you might need to do in terms of combining that information and shaping it for different outputs, so it's a report, a PDF, a slide deck, whatever it may be as your output, we'll help you automate all of that and just run in the background. We're the sort of content infrastructure for you and you are just piping us basically what you need to be shaped and distributed.
Turner Novak:
So what all that could that connect to? Salesforce, your data warehouse?
Grant Lee:
All sorts of things. Yeah, you can imagine obviously for a sales use case, any of the customer, wherever customer data might live or customer information might live. If you're a consultant and you need to tap into a body of research that you've done and pull that in. Anytime you want to personalize or be able to do specific types of market research or pull from different geos, zip codes, all of that can be piped in. And so it really depends on what data you have access to and what you want Gamma to be able to take advantage of.
Turner Novak:
Interesting. So if I'm a teacher, maybe the report card for a kid could be hooked up to this, and I'm sending you your kid's report card and it's just hooked up into whatever my grading software is. I don't know what teachers use, but it sounds like that might be a use case?
Grant Lee:
There could be definitely something there, and I think even simpler than that is just maybe personalizing some of the content that goes out. So for certain types of students, you include certain types of information. Or maybe for certain types of learners, you might give them more visual content versus stuff that is more long-form information dense.
If all those things are knobs that can be controlled, and so rather than having a one-size-fits-all sort of content production, you might actually fine-tune a lot of in the background and just allow that engine to run and spit out what needs to be spit out.
Turner Novak:
And then this is a little bit of a different topic, but I'm curious, you're sitting at the forefront of AI. How are you guys using it internally? Are there any interesting things you've done to leverage... You're smiling, so there's got to be some stuff you've done. Maybe the secrets that you're willing to share. What are some of the cooler ways you guys use AI internally?
Grant Lee:
Yeah, I don't think it's anything crazy. It is basically the way we use AI is that every corner of the company tries to deploy AI or leverage AI, and it's a constant sort of learning, sharing and willingness to try certain things out. Some things are still in uncanny valley territory, and that's fine, but we're able to test it.
And so yeah, I'm sure other AI startups are doing similar things, where engineers are exploring the Cursors of world, designers are exploring Lovable's Bolts to help with prototyping. Marketing team is constantly leveraging different deep research tools, or tools like Notebook LM to synthesize vast amounts of customer data to create personas, create archetypes of different customers we're going after.
And then customer support, leaning on infrastructure built by Fin and Intercom. And then using that to create better support content, and really creating that flywheel of, "Where are there gaps in our own help center content?" And streamlining, automating that.
We use a ton of AI to localize our product, and so get machine-translated content out in different forms of our web app, as well as our marketing site. And then going in and also having humans review that and prove upon it.
So honestly, it's like every single function is deploying some version of AI and AI tooling, and the question is just how much can we use that as great leverage?
As we continue to build, every single person should be thinking about actually building in a way that incorporates AI as part of the plan. Not just a nice-to-have, it's a must-have. And it might look different at different stages, and it might look different for different functions, but everybody's thinking about it.
Turner Novak:
I'm assuming you use Gamma too internally?
Grant Lee:
Oh yeah, of course.
Turner Novak:
Okay. You didn't mention it, so I was like, "Yeah, I hope you're using it."
Grant Lee:
Every team meeting, presentation, and we even have fun use cases around Fridays. We have this thing called Gamma Ramma which a team member can present on, just a fun topic, something silly. It could be something related to just pop culture and just educating the rest of the team, but using Gamma to create that presentation. So, we're trying to have fun with it but also use it as the actual tool we communicate information with.
Turner Novak:
Do you have a favorite, while we're talking about AI tools, do you have a favorite kind of under the radar or new one you've discovered or you still just like ChatGPT is kind of the king?
Grant Lee:
It really still feels like ChatGPT is the king. There's been a few out there that I want to spend more time playing with, like granolas of the world seem interesting, obviously friend Optimizely, Team Optimizely founder, Dan Kurding, Limitless, which seems like a really cool application. So there's so many out there, and more and more of them. I'm trying to figure out how to fold them in and give them a try.
Turner Novak:
One thing I want to talk about, anything that you guys specifically do differently from other companies in San Francisco? We kind of hit on it a little bit, but maybe some of those constraints has kind of molded how you guys think or operate that's maybe a little different than your down the fairway, venture-backed hyper-growth startup?
Grant Lee:
I would say there's some big differences in just the culture and how we operate. So one is, I think there's this tendency right now to believe that all the hyper-growth AI startups are operating their 996 shops, which I don't know if you're familiar with, but 9:00 AM to 9:00 PM, six days a week in the office, in person, and that's everybody. And we do not do that.
We have very flexible, a more hybrid work culture, where we're primarily based in San Francisco, most people come into the office most days of the week, but it's not mandatory. We don't force that. We have, basically anchor days, Monday, Wednesday, Fridays, where most of the team is in because we have our team meetings, but you want some focus time and working from home is the optimal way to do that, that's totally fine by us.
And we are also not of the culture of being a boiler room where the pressure is just constantly there. We're trying to build for the long haul, which means we're constantly trying to balance how much creative energy you can bring into building a product like ours requires the level of energy where you can't just force that through brute force energy, just working endless hours.
It means to be that you have space to think, space to be energized, and recharge your batteries. And we really much encouraged that, but then for us in person, obviously still has its benefits, so we haven't moved completely remote. We still want to be able to be in the same room, jam on things together very, very frequently. It just doesn't need to be 24/7.
Turner Novak:
Yeah, it sounds like it's working, 'cause your point earlier, you haven't had any turnover yet, so sounds like for... And it's been pretty successful for everything I can see from the outside at least. So, it seems like it's working.
Grant Lee:
Yeah, it's working so far and I guess what everyone will want to see, just what do the next four or five years look like? We're four years into a journey now, and a lot of these hyper growth startups are maybe even earlier into a journey, so they're 1, 2, 3 years into it. And so yeah, I think there's many different ways to win. And as a founding team, you got to find a way that feels authentic to you, and one that you believe you can sustain for a long, long time.
Because regardless of what type of startup you're building, if you want to be successful, it's not a one two year journey for most, it's multi-year if not multi-decade journey. And so, give yourself a chance to be on that journey for long enough. I think that's the sort of mindset we've tried to had.
Turner Novak:
So on that note, how do you balance trying to grow fast and meet the moment, but also not burning too much cash, remaining profitable, not spending it on dumb stuff that you don't need? Or has there been anything you've been able to center around of just the right balance of that? 'Cause I feel like that's a struggle everyone kind deals with.
Grant Lee:
Yeah, the way we balance it is to inherently infuse some natural tension into this. So, we create tension by our internal mantra is, "Hire painfully slow," which means we don't open up a bunch of headcount. We know that the team is going to remain lean. What that means is each team member needs to take on a ton of ownership and has high agency, can do things.
We try to have a very complimentary set of teammates, and we optimize for the type of people we're hiring, like generalists over specialists. And we, even at the manager level, try to invest more in developing player coaches.
So rather than the traditional management layer where you might bring on the VP with the Rolodex and they fill out the entire org chart by hiring. We do that by having these player coaches that are people that have managed in the past but really still love the IC work, and can be great mentors and coaches to those around them, and can play a role in strategy but are still plugged in to the day-to-day.
And so all that means that we can move fast. We feel the scrappiness inherently. We bake in or we try to time box things a lot, so we're setting our own deadlines so that we aren't just kind of letting ourselves drift infinitely into exploration, but there's some sort of semblance of progress, and we can measure ourselves against that.
So all those are things like you have to naturally, one, instill attention and then be comfortable with it, and be open when things maybe aren't working well. And so being in a culture of high feedback, culture of feedback, all those are important ways to engineer your own urgency into the whole setup.
Turner Novak:
Yeah, and you mentioned a couple of times this four months window. Do you guys do product sprints in four months? Is there some significance of that,? Or
Grant Lee:
It's kind of organically how it's happened. And I think what I've found is that obviously hard to sprint indefinitely. If you just don't have a timeline and people are just sprinting forever, it's really hard. Almost anybody can say, "I can crank really hard for a quarter." And that four month is just like you need a quarter in a little bit. You know you need to last-
Turner Novak:
It's an extra month.
Grant Lee:
You need to last not just a quarter. You're going to have to go beyond that. And it's like that mentality, I think for whatever reason, it just has felt like enough time to get our ducks in a row, enough time to build what we need to build, and then feel like it's still an appropriate deadline where we do that and it's not forever. It's like, "Okay, we have a few months here," let's make sure we've got our stuff together.
Turner Novak:
Do you give yourself a wind down period after that, after a four-month spring you're like, "All right, we're taking it easy for a month or two," or something? Do you balance those sign waves almost, or?
Grant Lee:
Yeah, I guess we don't really talk about it in terms of like, "Oh, we're just going to let our foot off the gas, take a break." It's mostly because after that, there's a whole bunch of bugs that need to be fixed, or things that we didn't get and we need to kind of iterate. So I think it just allows us, like we were building, building, building and now it's like, "Okay, let's learn a little bit, see what's going on. Make sure in some cases where there's immediate fires that need to be put out, we can put those out."
And then, we don't go immediately back into four month sprint. There's always some sort of period where we're actually learning from what's been valuable to the end users. And that sort of period ends up being a little bit more organic. It may be one month or may be a few months, and then we'll start thinking about the next four-month milestone.
Turner Novak:
Maybe it's like a calibration period or something.
Grant Lee:
Totally. Yeah. Yeah.
Turner Novak:
Where you're still foot on the gas, but you're just sensing of, "We're still going somewhere, still driving, but what do we need to level set or adjust, course correct."
Grant Lee:
100%. Yeah. Yeah, and there's oftentimes quite a bit, 'cause you can go in assuming you know all the answers, but oftentimes things that will be surprising that you have gotten wrong.
Turner Novak:
Yeah, so I have two more questions. They're both I made more selfish questions, but I have two kids, and we both have two kids of the same age. Any just, I don't know, parenting tricks or... Especially with running a hyper growth startup, anything you found that just works for being a good parent plus operator? What have you learned that's worked the best?
Grant Lee:
For me, what's worked the best is being very diligent about being on a routine. And so, I think part of this is you're going to be your best self when you feel healthy and well rested. And the only way you do that is by being healthy and well rested. And so being really, I think, what was it, Brian Johnson has the saying of you basically have to be a professional sleeper, sleep for your extra job. And when you show up late to your job, you feel bad about yourself, if you show up late to sleep, you shouldn't feel great.
And so getting that in is the foundation. Sleep then trickles into being able to have energy to exercise and fitting in that interior schedule, being super diligent about making sure you carve out enough time. And then from that, comes nutrition. It ends up becoming pretty obvious. You have those three in a good place, I think the rest can follow.
And then you can feel like when you show up to being a parent, you actually have the energy, you can actually have fun and play. And so, easier said than done, but I feel like it all starts with being on that routine and being as rigorous about making sure nothing gets in the way of that. And if you can do that, I feel like everything else, hopefully the dominoes can fall in the right place.
Turner Novak:
Yeah, I feel like my two kids are both an awesome age now where they want to play board games, play card games. We've been playing Minecraft the past month or two and I'm like, "How am I so lucky that just my kids just"... If I had nothing else to do, I would just play video games or play board games, so this is kind of awesome I get to do it. And it's like you don't want to play video games all the time with your kids. You try to limit the screen time, but it's like, "This is fun, this is actually enjoyable."
Grant Lee:
No, I think you're winning. You got it right.
Turner Novak:
Yeah. A last question. Do you have a favorite founder, CEO or business that you learned from just throughout history? This can be the Rockefeller or it can be maybe more modern recent example, but any stories or founders or companies that you've gotten a lot of inspiration from?
Grant Lee:
I am definitely a Nike fanboy and an Apple. I think with some of these brands that have built enduring businesses, so much of is about recognizing early on that there's also this community element. And you as a founder play a specific role, particularly special role in nurturing that early on. So for us, that translates into we have our own Slack workspace for our power users, we call it Gammbassadors, where -
Turner Novak:
Gammbassadors. Thant's an amazing name.
Grant Lee:
And I'm still daily checking that and trying to interface. And whether or not it's helping them use Gamma or just helping them out in any other way, many of them are building their own businesses and are looking for advice or need help in a certain way. And I don't always have the answers, but I do try to seek and give them some guidance, or introduce them to someone that might be able to, and that part of it feels good. It feels like we're building not just a product, we're building the community, and that sort of thing needs to be done brick by brick.
And I think obviously the Nikes of the world have shown that it's not like you should just set it, forget it. It's like, oh, those relationships are humans, and those humans require a lot of input from you, and nurturing that just takes a lot of time and investment. And if you believe it's worth it, you should do it. And so for us, that's been from day one, we really believe that it's something worth doing. And hopefully, something I could just continue to carve out time that it's something that we really believe is important.
Turner Novak:
And so what does that exactly mean, building a community or having a community? Does that mean that the people that use the product talk to each other, they have their own identity? Does it mean you as the founder or as the company or the brand are interacting with them? Maybe it's all those things, but what does that mean exactly?
Grant Lee:
Yeah, in the very beginning of community building, it very much is a kind of you and then finding individuals out there that see something about your product and trying to just build some of that relationship. And in many cases, they need a little bit of help getting the most out of your product. In other cases, they're eager to give you feedback on how to improve the product. And so, just giving them a channel to talk to you, I think is important, and deeply empathizing with what you could be doing better.
And then the long-term goal is you build a community that's then self-sustaining. You have other community members helping each other, and you are no longer part of the conversation. You have left the chat. You're out of it. And you can then just observe them being so excited to help another person get to a good place, because they know what it feels like to have a product that can serve them in some way and unlock some incredible thing for them, whether it's extending their business, or helping them build all the content they've always wanted to build.
And so, that's a magical moment, and we're kind of on that part of the journey. And obviously, that community can start small, and then eventually you want that to be something that's global, where people have this affiliation with your brand and your product, and a deep love for it, and that's the dream state. 'Cause if that happens, then you're just kind of out there trying to just push the boundaries on the product and the rest kind of happens more organically. And we're really trying to build towards that, where we can be one, where the product gets shipped and we have the millions of people out there just championing the product and just excited to share with the next person.
Turner Novak:
Yeah, 'cause in a way, they're almost like an extension of the team, in the sense of they're kind of doing your marketing for you, they're kind of doing your customer support, they're kind of like R&D.
Grant Lee:
100% Yeah.
Turner Novak:
So it's like shift those off balance sheet or off income statement and just makes a stronger business more profitable.
Grant Lee:
Totally. Yeah. Yeah. And I think it's so underrated because it's not quantifiable. Accounting is very quantifiable. I can go say, "Hey, what was my payroll last month?" But yeah, the value that you just described, that's hard to quantify, but if you get it, man, that's a secret sauce and it could be lightning in a bottle that you can just keep and fuel future growth for your company.
Turner Novak:
Is there a moment that you know that have it? Was there a moment when you kind of knew, "We have a community," quote, unquote? Is it the word of mouth thing usually, or?
Grant Lee:
There's definitely an element of that, but I think it's the same thing as when people talk about product market fit. It's like when you think about product market fit, it's sort of this thing that's almost fleeting or the game gets harder. So you might have product market fit in a small TAM, but if you want to go after a massive TAM, the product needs to change and you're going after a bigger market.
And so for us, the same thing is like, yeah, we have a small community of users today that love the product. To become a global community, the product needs to evolve, the way we interact with the community needs to evolve, what they get out of being in that community needs to evolve. And so yeah, we're at the small stages of growing what we think to be enduring and special community, but it's like, yeah, ton more investment needs to be required, and hopefully we can actually fulfill that over the many, many years ahead.
Turner Novak:
Well, Grant, this is a lot of fun. Thanks for doing this.
Grant Lee:
Thanks so much, Turner. Yeah, this was great. Really appreciate it.
Turner Novak:
Yeah. And just real quick, where can people find you? We want to look you up, Twitter, LinkedIn, both of those?
Grant Lee:
I'm trying to be active both on Twitter and LinkedIn, so just Grant Lee. And yeah, I'd love a follow. If there's anything I can do to help, please reach out. Love to connect as well.
Turner Novak:
Yeah, I feel like in Twitter, you tweet, tweet what you're seeing. It's like thoughts, it's interesting stuff. People should check it out.
Grant Lee:
Yeah, I appreciate it. Trying to. I think there's so much fun stuff going on in the world, so trying to make sure we can participate in the conversation.
Turner Novak:
Cool. Well, thanks again for doing this.
Grant Lee:
Thanks. Take care.
Stream the full episode on YouTube, Spotify, or Apple.
Find transcripts of all other episodes here.