Founder Stories: David Hsu, Product market fit and How to find it

Hi everyone, thank you for hanging around. This is the last session, right? And then the night sessions begin, I guess. I'm Brian Schreier from Sequoia Capital. This is David Su from Retool. We're really excited to take a little bit of a break from the last session, which is It's great to talk about the future and navigating this, this winter that's coming ahead. But they also mentioned retooling for that. And we'll get into some of that in this session. But it's really going to be about product market fit and dreaming and creating. So excited to talk about that with, with all of you. I actually started my technology career in Europe at Google, where I worked for 6 years as we were opening our first European headquarters back in 2004. And then more recently, I've been at Sequoia for 14 years. We've been pretty active in Europe since I started at Sequoia, partnering with companies like Unity, UiPath, Klarna pretty early on. And we have a team that's live on the ground in London. So we're really, really excited about the ecosystem over here. David actually also did his schooling in Europe. He went to Oxford, and this combo major, or I'm not— what do you call it at Oxford? You wouldn't call it a major, would you? Or would you?

Uh, we call it a subject.

A subject. The Oxford subject, uh, of computer science and philosophy, which makes for really interesting conversations. Uh, he founded Retool. We met in 2019, uh, and the company has scaled 50x, 5-0-x, since we first met and partnered in the Series A. So we've seen David do an amazing job taking Retool from just getting to product-market fit and really scaling it up into a large business that serves very large customers as well. So with that intro, I'm gonna turn it over to David to tell us all what Retool is and how you started.

Yeah, so Retool is kind of like Legos for code. It's kind of this fast way of building internal tools. And the way we started it actually was we ourselves built way too many internal tools. And as kind of lazy engineers, we were like, there's got to be a faster way of building all this stuff instead of coding from scratch. So that's where Retool first started. And when Brian and I first met, I think we were 4 people at that point in time. We were around $1 million. Yeah, we were 3 or 4 people doing around $1 million in ARR. I think at that point we did not think we had product-market fit, but when talking to Brian, we had to somehow convince Brian that we were a great company and had product-market fit. So that was a fun experience.

Cool. Well, you, you did convince me that you had product-market fit, so this is gonna be an interesting conversation. So tell us, how do you define what product-market fit is? And if you didn't have it back then, you know, do you have it now?

Yeah, so before starting a startup, I had sort of been sort of fairly interested in startups, and most people had referred to product-market fit as something like, oh, you, you will know it when you have it. It's almost like hitting a geyser or something. We went through Y Combinator, and YC, one of the talks was from the founder of Segment, Peter Reinhardt, who's now an investor in Retool as well. And Peter at that point said that it's so obvious when you have product-market fit, It's going to feel like sort of the rolling a rock downhill instead of rolling a rock uphill. And I remember when we were probably at like $3 or $4 or $5 billion ARR, I was still wondering to myself, like, still kind of feels like I'm pushing it uphill. Like, when's it ever going to go downhill? And so for us, the feeling of product-market fit was sort of never zero or one, never black or white in that way. It was very much sort of a process. I think at one point We looked around and I think we're around $10 million ARR, maybe around 20 people in the company, maybe 30, something like that. And we're like, maybe we have product-market fit now. And it seems like customers are buying this thing, so I suspect we do have product-market fit. But it was actually very hard. And even today, I would say it's still to some extent pushing a rock uphill. Hmm.

It's, it's refreshing to hear a company that almost went from $1 to $10 million ARR by, by surprise or by accident. So I like how easy you make that sound. Um, it's not that way for, for all the companies that I've met out there. Um, you guys did a lot of hard work. Uh, one of the things that has been said is that, David, you have a formula for finding product-market fit, and I would like to write that down.

Great. Well, I think for B2B SaaS companies, finding product-market fit is actually quite formulaic. For consumer, I think totally different. But for B2B SaaS, our thinking always was sort of sales was the best way to find product-market fit. And when we first started with Retool, we sort of had the core idea of Retool, but to be honest, we had no idea how to pitch Retool. It was sort of this, you know, interesting piece of technology, if you will, that we had built. The first time we launched Retool was on Hacker News, and I think the Hacker News title was called something like Excel but with higher-order primitives. People were very confused about what this was. I think no one really understood. We tried sending outbound emails actually with that title, and no one responded, sort of unsurprisingly. Then we pivoted a bit. We were like, okay, well, you know, maybe that sort of messaging doesn't really resonate. We kept the product exactly the same, and said, hey, what if we were to focus on, let's say, internal tools specifically? Because actually sort of companies build a lot of internal tools. And so we sort of pivoted the messaging, and that sort of turned out to be much more effective. And then we entirely just did sales for probably maybe 18 months or something. It was entirely outbound sales. Uh, and I think B2B SaaS product market fit is sort of quite formulaic when you do, uh, sort of outbound sales in that way. Because, uh, for example, at that point we had a bunch of investors that invested in us, and we actually had, uh, tried getting a few investor intros, for example, to be like, hey, could you introduce us to other companies that you invested in or know? They might be building internal tools, which, you know, is all of them, kinda. And a lot of people would take these intros, but actually it wasn't very effective. It was kind of like a very low signal intro. People would always take the intro, they would always act kind of interested, but actually wouldn't actually end up buying the product. And so we're like, we gotta be more hardcore about this. What if we just did outbound sales? And so we did that. And through outbound sales, we iterated a lot of the messaging, we iterated a lot on sort of the profile that we pitched to, even the market, actually. So one funny thing about product-market fit is that it's both the product, but it's also the market as well.

And so— David, can I ask, just to jump in with one point? One of the things that I often say about working with you is most founders come in one of two flavors, which is either hyper product-focused or hyper go-to-market or sales-focused. And I feel like you've done both really well, and you really get both. And what I'm hearing from your answer to the formula for product-market fit is there's a lot of sales in there, which is maybe something new, maybe something that's kind of surprising to folks who are coming at founding or building product more from a technical angle. So can you talk a little bit more about how you think about sales and how you weave that into product-market fit and building a business?

Yeah, I think the rise—

as a technical person.

Yeah. Yeah, I think the rise of PLG has actually been really bad, actually, for a lot of early-stage founders.

Do folks know what PLG means? There's some nos out there. So PLG is product-led growth. It's, it's the idea of, uh, it's the idea that you can build a software product or a consumer app and it kind of sells itself without having to be driven by more traditional sales and marketing. PLG is very common in the developer space. Which is the market that Retool actually sells to. And the reason product-led growth is a strategy that the developer space employs is because developers do not want to talk to sales and marketing people. I had a big sales function when I was at Google, so, you know, I don't feel bad about myself. It's just the way it is. You think about a company like GitHub, which was a wonderful outcome. They didn't have salespeople calling engineers. You know, they had open source, and they had a very large network of folks who were using GitHub for code who just adopted the service themselves. So that's PLG, and it's interesting to hear David say, "Hmm, maybe it's not such a great thing to chase." Yeah, I think PLG causes people to optimize for very weird things, and probably kills the company, actually.

Today, when I talk to founders, for example, a lot of founders will tell me, "Oh, you know, we don't have much revenue yet, but it's PLG, and, you know, the reason we don't have much revenue," It's because the revenue's going to come. We just have to build a great product, and it's all going to work eventually. That very rarely works. I think you were an early investor, I think, in the seeds for Dropbox. Dropbox is like a once-in-a-generation company. That doesn't happen very often. Chances are, if you think you're Dropbox, you're probably not Dropbox, actually. And so for us, we were always very sales-driven to say, like, hey, currently we have this website. We have this product. People are signing up. But people are not paying us. Why are people not paying us? Well, let's go talk to them, actually figure out why they're not paying us. So then we talked to them, we were like, oh, actually it's totally the wrong market, actually. So one funny anecdote was when we first started Retool, one sort of idea we had was, oh, a pretty good market for Retool might be FileMaker developers. Retool's kind of like FileMaker, it's like FileMaker 2.0 basically, to some extent in the cloud, if you will. FileMaker is a sort of Apple product from 20, 30 years ago. And so what happened is I ended up joining a sort of FileMaker user group on LinkedIn, and I was like, this is gonna be huge. You know, we're gonna sell all these FileMaker people on Retool and we're gonna make, you know, we're gonna sort of find so many customers here. And what ended up happening was I think we got a list of like 200, 300 emails from the LinkedIn user group, and I emailed all of them and I think we got 3 replies out of 300 or something. And they were all like, no, no, no. But one guy was like, hey, I'm actually happy to hop on a call with you and tell you why this is such a bad idea. And I was like, well, that's a start, so let's take the call and see what happens. We took the call and he was like, well, it's a terrible idea because FileMaker is good for ABC reasons and you're bad for XYZ reasons. And therefore FileMaker is great and Retool sucks. And we were like, well, thank you for the feedback, see ya. And so that's an example of something that if we had done the PLG way, sort of in a marketing-driven way, We would have been like, "It's pretty weird, a lot of signups, but no one's using the product, no one's paying for it. Why is that?" You don't really know, actually. But then when we did that in this sort of outbound sales way, people maybe sort of feel more okay to be mean when you're doing outbound sales. And so they sort of tell you the truth more, basically. And that helped us iterate a lot. And so in the early days, the sort of core messaging that we had was almost entirely driven by the outbound sales that we did, and we were sort of test messaging We test pricing, for example. How Ritual's priced today is entirely because actually we would test like, hey, we charge $5 a seat, we charge $10, what about $100? When we say $100, what's their facial expression? Are they happy, are they sad? If they're happy, we're like, wow, now we learned we can charge $100 per seat, it still works. And so that is the kind of learning you can only have via sales. And so for that reason, I think PLG is actually sort of a very dangerous path to go down or a trap to fall into when you believe that, oh, you know, if I build it, the people will come, or the customers will come. In reality, I think it's very rarely the case. Ryan, I think you and I were talking backstage. You were sharing sort of one sort of counterintuitive thing or surprising thing about PMF Sequoia sees.

Oh yeah, well, I think, I think one of the things that we see is that many companies actually just never find product-market fit. Which is— it's just really tough to admit. And I think that what people tend to miss is that it's not really a science, you know. It's definitely more of an art. It's more of a holistic effort. And as David was talking through his approach and Retool's approach to finding product-market fit, I think one of the things I would love for the audience to take away is how customer and sales oriented that effort was for them. And I think I know from, from starting to work with David very early on that when he talks about— when most companies talk about sales, if it's a bigger company, they're probably talking about the salespeople. You know, when you're talking about Retool, when it was 2 or 3 folks, the salespeople, you know, the engineers were the founders, the product people were the founders, and the salespeople were the founders. And so whether you're starting a small company or you're building a new product line at a much larger company, when you're putting a new product or feature into the market, you're probably going to be more on the product and engineering side, but you also need to think like a salesperson because that's how you're going to get the feedback from the market on the value proposition, how it's working, and what your customers really want. Versus what you want to build for them. And there's always a little bit of a give and take there.

Brian, I'm gonna ask you an unscripted question, if that's okay. What percent of Sequoia seed stage companies do you think find product-market fit?

Oh, that's a good question. Um, not one I'm prepared to answer in a scientific manner, um, for my last comment. But I would say that at Sequoia, you know, we start with companies at the pre- pre-seed stage all over the world. Actually, we just launched our pre-seed stage and seed stage Catalyst called Arc in London this summer. It was a lot of fun. And I would say that of those pre-seed stage companies, none of them have product-market fit yet. And for a typical seed company out there, I think, you know, a quarter of them would be lucky to find product-market fit. It's, it's really, really tough. Now, I think that half of those companies probably think they found product-market fit, uh, and because sometimes you can cover up the cracks with a different type of sales and marketing, which is just spending a lot of money to get people to start buying or using your products. Uh, and that's another common kind of failure mode in this space that you want to be careful about. What do you think companies should do if they can't find product-market fit?

Wow. Uh, so I think in a B2B SaaS world, actually, the way I think about it is sort of there is always a path to a product-market fit to some extent, in the sense that it's almost like a chess game. Like, I think when you think about no matter where you are in a C-stage company, I think there's like 8 moves you can play where you will find product-market fit somehow. Uh, now Now the question is sort of how do you make sure you're on those 8 moves, basically? And that's a sort of challenging thing. But I think being sales-driven, being customer-driven are probably the 2 main things. Especially maybe sort of being outbound sales-driven, I think especially so. Because again, you know, as mentioned, when you're outbound sales-driven, people don't lie to you. When I remember sort of the happiest days we had in early Retool was when we sent an email to this, actually another Sequoia company based in Latin America called Rappi. Like a delivery company. At that point, I think maybe a few thousand people or something. So I had just cold emailed the CTO saying like, "Hey, we have this fast way of building internal tools." And I think within 30 minutes he replied and said, "We'd love to see a demo." And I was like, "Well, great. How about tomorrow?" He was like, "How about today?" And I was like, boom, like product-market fit. So, you know, whereas like in that case, I was like pretty certain like this guy actually wanted the product because like we had no relationship, right? At that point, we're not a Sequoia company. There was no reason for him to respond to my email. The fact that—

Do you think that's product market fit or value proposition market fit? Because he hadn't actually seen the product yet.

I would agree. Yeah, so I would say—

I think we're hitting a lot on the value prop here and how getting that right can really inform your product development work.

That's a good point. I think probably value prop fit, whatever you want to call that, is probably number one. Because for us, if I think about that sale, you're totally right, because when we sold that when we sold Rappi at that point in time, our product was fairly immature, actually. What we were pitching them really was, we have this solution to this problem they have. And the problem they had, we actually knew pretty clearly, was they sort of have all these internal tools they want to build because they're a very operationally heavy company, but they don't have the software engineers to go build all these tools. And so that's why he was so interested in Retool. But when he first saw the product, like, the product itself actually had a lot of shortcomings. There were a lot of things we had to go fix. We, for example, didn't have a lot of the key features. I think we might have built on-prem actually. I think we built that over a weekend or something after he said he was interested in on-prem and he couldn't use it without on-prem, for example. So I think you're right. I think he was broadly interested in the problem, but the fact that he was a stranger that we did not know who was so excited by the problem that we were solving that he would hop on a call within like an hour or two, like to us that was gold. That's very exciting. Uh, that is something you could only get, I think, via sort of pure outbound sales.

Yeah, and I'm guessing you didn't have much time, you know, to pitch this. Was this email back and forth, or was it a phone call? I can't remember.

It was email. We had emailed him.

Short emails, probably, though.

Yeah.

You just— I think one other thing that people— when you're working on the value prop, people assume you have so much space, and you really have very little. You have one sentence, typically, to really catch someone's attention. In order to get in a meeting in 30 minutes. Like, you really have to have it nailed, which is a, which is a pretty big deal.

Yeah. One thing I'll say on that too is we experimented a lot with the messaging, but the core product deal was always the same. I think one mistake I see early-stage founders, uh, uh, make is that sort of they have no core idea at all. They're kind of just like, I don't know, I don't have an idea. Let me test various Google AdWords and see what idea, you know, I have 7 ideas and let's see which one's the best one. I'm gonna go pitch all 7 ideas via cold outbound email, see what happens. That, I think, very rarely works. I think maybe sort of a startup philosophy, if you will, that I have is I think you have to have some strong differentiated opinions about the world that then you go and test. But you can't just say, I have no opinions, let's see what happens. And so in our case, our opinions were something like, we think that the current way of building software is stupid. We think there could be a much better way of building software, sort of a much higher level way, kind of like Legos for code, basically. And so that was our opinion, and that was, you know, kind of a controversial opinion. And the second controversial opinion we had was that we wanted to sell that to developers. I remember when we were a small startup, plenty of people would come to us and say, what's your idea? And it'd be like, oh, we're selling this drag and drop way of building apps to developers. And people would be like, that sounds like a terrible idea, you know, I'm a developer and I would never use a drag-and-drop product. Like this, you know, you should go shut down, you should go do something else. We in fact had a YC batchmate that told us, oh, this idea, I tried this, you know, 3 years ago, it doesn't work for XYZ reasons, I would go do something else too. And so I think what you have to sort of believe is you have to sort of believe these sort of maybe 2 or 3 core beliefs you have about the world that are different from what other people believe. And then you have to test that in a very rigorous, sort of outbound salesy way. But you can't be too sort of fluid with what you're testing, because if we were testing 7 ideas, honestly, we'd have no idea what, you know, resonated or not. So I think it was, we had a core idea and a core value prop, and we continuously test messaging on that value prop, find the messaging that resonated, which was build internal tools fast, and then work with the customer to build the product. That, I think, is the formula. Finding product-market fit.

That makes a ton of sense. And, um, Retool's been on quite a tear in terms of launching new products recently. So 2 over the summer, I know there's another one coming. How do you think about product-market fit when it comes to new product lines? And maybe this applies to folks who are working on bigger companies with multi-product line types of businesses. And what's next for Retool?

Yeah, it's tough. I think one, um, sort of curse of success, if you will, is that companies almost become less interested in taking big risks the more successful they are. And I have actually long admired sort of Chinese or, you know, almost like, you know, Asian companies in this for this reason is that these Asian companies seem to do everything. If you look at, for example, Tencent, or if you look at WeChat, if you look at Alibaba, you look at Grab, for example, these companies do everything. I mean, on WeChat, for example, You can order food, you can send messages, you order a taxi, you can send money, you can order food, you know.

Apparently this is going to be Twitter soon, right? Once Elon finishes deploying products.

It's hard in the Western world, I think. But I think these sort of— these Asian companies are so competitive, their eyes are so on the ball. They're like, okay, we launched one product, that's great, we're not done yet. We want to launch 5, 10, 15, 20 20 more products. And that, to some extent, has been sort of the thinking at Retool as well, which is, you know, it's great that we have this one product that, you know, now has found product-market fit, but we don't want to be complacent. We actually want to build far more products. So Retool today is sort of a fast way to build frontends. We want to expand onto backend storage, backend logic, mobile frontends, and probably like 5 or 10 other product lines as well. And it's actually quite challenging because figuring out how to sort of invest in new product lines when you have a product that's already working, when you have limited resources, is tough. Because you know that if I apply, let's say, you know, 4 engineers to the core product, the ROI there is actually pretty positive. Like, we can sort of pretty easily quantify what the ROI of that is. Whereas we can spend 4 engineers, you know, for a year on a new product line, it could go nowhere, actually. And so today, the way we think about building new products is we keep the budget very low. So we like to think that sort of new products can be started with basically 2 or 3 people. For us, you know, at Retool, when we were at around $1 million in ARR, when we met Brian, we were only 3, 4 people. And so that's kind of the philosophy that we had for starting new product lines too, which is, could we get this product line to something like $1 million in ARR with again, with just 2 or 3 people. And that means, you know, it's gonna be hard. Like those 2 people are gonna have to do everything basically in order, they're gonna have to do sales, they're gonna have to do engineering, they're gonna have to do marketing, they're gonna have to do product, everything. But that is kind of, so now we're sort of applying the formula again, if you will, to these 2, 3 new product lines that we're launching as well. But it's a lot of fun, and I think part of the startup, part of the fun of doing a startup is being able to launch these new product lines, expand your product offering and find product-market fit again, again, and again, as well as scaling the core product to 9 or eventually 10 figures in ARR. So it's a lot of fun.

Awesome. Well, hopefully many of you out there are working on product-market fit, and this was some helpful information, guidelines to go by. But if you are, you're very lucky because, as David was saying, you literally get to wear every hat in that journey. To create something new and put it in the hands of your customers. So good luck to you all on that. David, thank you so much for the time and sharing your wisdom.

Thanks, Brian.