It took him 4 years to launch—then he hit $1M ARR in 30 days. | Siqi Chen, Founder of Runway

Siqi Chen (00:00:00) :
Post launch, it exploded very, very quickly. I think we had about seven million impressions on launch day, and leads were coming in so fast that I had to build a GPT automation. That would qualify leads automatically because we literally, we couldn't qualify them fast enough. We were on a $20 million a year run rate at that time. So our revenue was basically forecasted to be effectively rounding down to zero for two years. What are we going to do? Well, we basically can't do anything. We can't spend any money at all. So the conclusion is we have to lay off ninety five percent of the team and just hibernate.

Pablo Srugo (00:00:36) :
How quickly did you sign another million in ARR after the GA launch?

Siqi Chen (00:00:40) :
It was within the next thirty days or so.

Pablo Srugo (00:00:44) :
Wow.

Siqi Chen (00:00:45) :
Yeah.

Pablo Srugo (00:00:45) :
That's a good launch, man. That's a good campaign.

Previous Guests (00:00:50) :
That's product market fit. Product market fit. Product market fit. I called it the product market fit question. Product market fit. Product market fit. Product market fit. Product market fit. I mean, the name of the show is Product Market Fit.

Pablo Srugo (00:01:02) :
Do you think the Product Market Fit show, has product market fit? Because if you do, then there's something you just have to do. You have to take out your phone. You have to leave the show five stars. It lets us reach more founders, and it lets us get better guests, thank you. Siqi, welcome to the show, man.

Siqi Chen (00:01:18) :
Hey, thanks for having me. It's great to be here, Pablo.

Pablo Srugo (00:01:21) :
I'm excited to chat about this with you. I mean, your company Runway is doing this kind of finance simulation. Which I've now, but only in the last year or two, seen a few players start to pop up to try and kind of simulate business decisions, and how they might look on the finance side. You were pretty early to this, I would say since '22. Let's start there. Take us back to when you started the business. Maybe right before, what you were doing that led to this idea in the first place.

Siqi Chen (00:01:48) :
Yeah, well, the start I think business simulation is like forty years old, maybe fifty. When we invented spreadsheets, it's in some ways a simulation of business. They just weren't very useful or good simulations of them and that is what triggered me to want to start this company about five years ago. Is trying to figure out how we build a platform and a tool that can create better simulations.

Pablo Srugo (00:02:11) :
Is it FP&A in general? Is that what you would call the space?

Siqi Chen (00:02:15) :
Yeah, FP&A in general is in some ways a simulation of the business and I think that's sort of not well understood. Because most people think about finance as sort of bean counting and that's the accounting bit of finance but the FP&A bit is strategic. It's about the forward looking, the future of the business and you have to have simulations to do that.

Pablo Srugo (00:02:35) :
But typically, and I don't know where Runway comes in. But just typically, at least my understanding of FP&A is it's highly specialized finance roles that are really using this stuff and obviously, they're then preparing things and presenting them to the CEO, COO, CMO, whatever. To see what projections would look like. But it's not like you have, you know, a founder type person playing around with things, tweaking things, and seeing how that plays out. At least in most FP&A tools that I would know.

Siqi Chen (00:02:56) :
That's exactly right. That is the state of the world, right? And my experience running the company right before Runway and my first company exposed that that is how it works to me. I experienced that, and I think that is the thing that's kind of broken, right? The job of finance, the job of FP&A, is to deliver a forecast and to help the business ultimately make better business decisions. And in the ideal world, you want that to be the same person. But the reality is that business is complex, finance is complex and so you need to have specialized tools, and specialized people to try to do that. And the thing that didn't make sense to me is how poor that experience is as a baseline. That goal of delivering that insight is just not achieved today in the real world. But we spend an enormous amount of time and capital trying to make that happen. And one of my beliefs is that a lot of the time it's just because the tools aren't good enough. Our tools inform our behavior as much as our behavior molds the tools that we create for us.

Pablo Srugo (00:04:06) :
What was, by the way. The business that you were running before Runway?

Siqi Chen (00:04:09) :
So I was running a company called Sandbox VR in early 2020. I joined the company as chief product officer and then around when we were raising our Series B, I was appointed CEO. The founder and the board thought that was the right move. Because I raised the round for our Series A, and I was going to do the Series B, and I was doing most of the job of the CEO anyway. So that happened, and then basically six months into the job, COVID hit. We were four hundred and some employees, and overnight we went down to fifteen because we had to hibernate the company. But the process of getting there required us to make these models. Because nobody knew how long COVID was going to last, right? Elon said in March of 2020 that it was going to be over by June, right? And meanwhile, in January, before the virus really came here into the States. Andreessen started posting these signs on their door saying, we don't shake hands anymore and they were excoriated, dragged in the press over fear mongering about COVID. And so that was the background, right? When it did hit and lockdown started happening or was about to happen, this was like March or so. Everybody was sort of triaging their portfolio companies as a VC, and so we had to create these scenarios for how long COVID was going to last, and Andreessen.

Pablo Srugo (00:05:41) :
Were they an investor?

Siqi Chen (00:05:43) :
Yeah, so they led our Series A at Sandbox.

Pablo Srugo (00:05:46) :
Okay.

Siqi Chen (00:05:46) :
So we were Andreessen's very first overseas investment. They told us that no, you need to plan for two years. Nobody was planning for two years in March of 2020.

Pablo Srugo (00:05:56) :
Also, because it's a startup. You oftentimes don't really, even if you were like, it's going to be two years. You're like, it better be one. Because what am I going to do for two years?

Siqi Chen (00:06:04) :
Yeah, we were on a $20 million a year run rate at that time. So our revenue was basically forecasted to be effectively rounding down to zero for two years.

Pablo Srugo (00:06:15) :
Wow.

Siqi Chen (00:06:16) :
What are we going to do? Well, we basically can't do anything. We can't spend any money at all. The conclusion is we had to lay off ninety-five percent of the team and just hibernate.

Pablo Srugo (00:06:26) :
And not to get too deep into that story, but what was the product that COVID was going to have such a big impact on the top line?

Siqi Chen (00:06:32) :
So, where do you live?

Pablo Srugo (00:06:33) :
In Ottawa, Canada.

Siqi Chen (00:06:34) :
OK, so we don't have a location in Canada yet. But what Sandbox VR is, is if you watch Star Trek or The Matrix, it's basically like the holodeck or The Matrix in real life is the best way I can describe it. So we have these rooms, we call them holodecks, around the world. So now there's seventy locations. We're in Dubai, we're in Macau, we're in Las Vegas. We have two locations in Las Vegas, a bunch of locations in California, and America. I think about twenty different countries now and basically, you can think of it as like the movies where you buy tickets. So similar business models to movies, but instead of watching a movie or a character in the movie. So, you can be a crew member in Star Trek, or you can explore the world of Stranger Things or play Squid Games. All these are titles that we have at Runway that we've officially licensed with all the different IP holders and the difference between at home VR, and what we do is the biggest difference is that you're in this real-time motion capture stage. So you have a full body, and it's multiplayer in the same location. In VR, you see all of your friends. They look like different characters from the game, but you can touch them, you can high five them. It's a full body experience that will transport you.

Pablo Srugo (00:07:48) :
So naturally, it was going to be closed during COVID.

Siqi Chen (00:07:51) :
Highly social, contact, outdoor experience. So it was all going to go to zero, right? So that's what happened. After COVID, it's doing $100 million a year, we have one thousand employees, and profitable. At the time, it was a hot deal, hot company until COVID hit. So when that happened, I was working with our CFO. We hired a CFO from Netflix, fantastic, fantastic finance person and we were just creating these spreadsheets for how long COVID was expected to last. And we ended up with twenty different versions of the spreadsheet. And then if we changed one, the others didn't change. And there were errors in them, because there were so many of these versions and it was such a complex business. And basically, on the tail end of it, when we laid off ninety four percent of the company. I laid myself off, the entire leadership team. I was the founder of the company, I was sort of the higher CEO. I talked to Andreessen's team basically right after, and they were like, what do you want to do next? And I said, has anybody thought about making good finance software? Because we have Figma for design, we have Notion for knowledge. There just doesn't seem to be anything good that's equivalent in quality for finance.

Pablo Srugo (00:09:02) :
That's interesting. It wasn't something you've seen for a long time. It was really just a very recent, deep pain point that you had just experienced.

Siqi Chen (00:09:09) :
So that was a trigger, right? But this is the fourth time I've been a CEO, the third time I've been a founder and so my first company. I was a year out of school, and I found myself in the weird position of running a venture backed company, which is a story in itself, right? I had just literally graduated, I moved to the city, and I didn't know anyone, and I somehow became a founder. This was before being a founder was even cool.

Pablo Srugo (00:09:33) 
What year was this?

Siqi Chen (00:09:34) 
This was 2007.

Pablo Srugo (00:09:36) :
Oh wow, okay, yes.

Siqi Chen (00:09:38) :
Yeah, I'm older than I look. Asian, don't raisin it. I'm forty-two, so people think I'm younger because of the hair.

Pablo Srugo (00:09:47) :
But it's true, from '07, to now. I mean, the startup world has changed dramatically. It's become institutionalized. It was very much at the edges at that point.

Siqi Chen (00:09:55) :
Yeah, I was Y Combinator's very first dropout, and Y Combinator was not cool. It's actually weird to see people say, like, YC whatever. As if it was a badge from Harvard. I mean, it's an amazing program, and it's highly prestigious. I'm just saying it wasn't back then, in '07.

Pablo Srugo (00:10:10) :
It's funny because it probably was the most impactful and best back then. Not to say that it's bad now, but that's just my assumption, actually. I wasn't there early, you could tell me but I would assume it was probably incredible back then.

Siqi Chen (00:10:20) :
Yeah, it was just twelve companies and Paul Graham, right? That was it and Jessica. Paul and Jessica, now it's a whole institution. And it's like more prestigious than Harvard and MIT. But man, it wasn't back in the day, and Paul really made it what it is today. But anyway, in '077, at my first company, when I was a year out of school. I majored in math, but I never thought of myself as any kind of business person. So, when this company started working, we were making gigs for Facebook, and we were about forty employees, and profitable. But I would get these spreadsheets from our fractional CFO because I was told we needed one. Which we did and I just couldn't make heads or tails of it. I was like, what even is contribution margin? And I guess, I'm supposed to look at this and understand some kind of forecast, but she had no idea how the business works. How do I know she doesn't know how the business works? Because she never talked to me. Somebody asked, like, how does this work? It's basically like, here is this standard three statement financials, and we're going to forecast trends from here. And I just remember for a really long time feeling really insecure about understanding finance as a business person. And I think that's really common with founders, right? Unless you get an NDA or something. You feel like you should have an NDA to understand this stuff.

Pablo Srugo (00:11:39) :
I still have this. I mean, so many first time founders that I work with. It's the same thing every time, right? It's like, OK, you got to do a cruel, you know, you got to account for things like this and like that. They're like, I just want to know where the money went and I'm like, but there's a whole system. But it's, you know, just the load to understand gap financials and cash finance, and all this stuff is high. And so a lot of people just start with a spreadsheet, and just put dollars in here. Then they went there and it can become a mess later. But anyways, I faced the exact same issue in my first company. So it's just the thing that keeps happening over time.

Siqi Chen (00:12:11) :
That's probably actually the right way to do it, right? When you're starting out. You want to be cash based, that's more useful. You want to create a model so you understand how the business works. Those are good things and that's what creates this disconnect, right? So then you introduce finance, and you have accrual accounting, you have the three statements, GAAP, compliance, all these different things. And that's designed to solve a very specific, different problem, right? That's for compliance and controls, and standardization. And the needs of a founder are around operational understanding. Understanding the cash view, the reality of where it is today, rather than virtual money in the form of accruals and amortization. And there is no bridge between the two, right? And this is why you end up with not just a CEO model that's disconnected from the finance model, but you also have a sales model that's disconnected from the finance model, and the product model, the growth model. Every department ends up having their own model as you scale and so the belief here is that a lot of this is informed by complexity, but we invent tools in order for humans to manage that complexity, to make sense of more complexity, to offload complexity to a tool. So that we can impart intent and understanding to it. If you think about the spreadsheet as a concept. What a spreadsheet is very good at doing, the purpose of it, why it's powerful, is that it offloads the calculations that we cannot do. That are not fast enough for us to do, we can't fit it all in our heads, into a platform in the form of a computer that is very good at exactly that. But pre-AI, it wasn't very good at the other stuff, right? The creativity, the strategy, the context understanding and so we think about the opportunity the same way. Which is, how do we make this business less complex and the gap between finance as a concept and people who are operating, how do we bridge that gap? That has been our North Star and it's been a very long, and fulfilling and often painful journey to get there.

Pablo Srugo (00:14:22) :
And so when you go in, you know, 2020 post, not shutting the other business down but reducing it drastically, and you talk to A16Z about this idea of kind of Figma for finance. The high level problem is clear. What's in your mind as what might be a way to solve it?

Siqi Chen (00:14:40) :
I just went inside and thought about my experience operating. And I think that was probably the right place to start. We also talked to an enormous amount of finance people, over a hundred, just to see what the pains are.

Pablo Srugo (00:14:52) :
Before you started Runway or after?

Siqi Chen (00:14:55) :
After, this is a problem I wanted to work on. So we raised money, and that's one of the things about being a second or third time founder. No matter what happens, it's just easier the next time. People want to bet on people who keep on swinging the bat. So what we realized is two things. One is their finances experiencing a lot of pain. But those pains are different from the pains that I experienced and so one of the dilemmas you often have as a founder first getting started is there is your personal experience. There is the utopian world, the ideal world that you think ought to be and then there's the live reality, of the customers and users that you're going to sell to. Untangling that can be quite difficult. Because one of the trade-offs is it is very useful to talk to customers, solve problems immediately, and build momentum. That is a very good strategy. However, you can end up with the faster horse problem, right? People experience these pains only because of deeply embedded consequences of the systems they're used to. So it is a creative act to figure out, well, how much do you solve for the immediate pains? And how much do you solve for the root cause of those pains in the first place, and build something more ambitious? And that is very difficult to figure out. There is often a trade off between how quickly you can get started today and make progress and whether you're setting yourself up for long term success, right? And when we thought about the problem, the way we thought about this is if we're going to build for the long haul, we want to get the primitives right. The foundations of what we're building right, and so what we've started with is a view of your finances that was incredibly flexible. We built this ultra pivot table, basically and the idea is most founders would look at a financial statement or their QuickBooks and find it completely useless, right? It's very difficult to get insight from it. What you as a founder or CEO would think about is, where do my money go? In what categories? And you want those categories to map this way in your head. And in those categories, it's not GNA and SNM. It's like, I have this department, I have this project, I have engineering costs and hosting costs. And I'm trying to trade off these different sources of spend. I want to find where I can save money, and you want it to all add up, and you want to call fear. So we built that.

Pablo Srugo (00:17:24) :
We'll dive into that. Let me just cover a few specifics. First, you said you raised some money. How much was your pre-seed?

Siqi Chen (00:17:29) :
$5 million.

Pablo Srugo (00:17:30) :
OK and then why did you end up on, because this is one of the challenges with this space. I mean, with many spaces, but this one in particular. Where it's like the ICP really matters, because the problem of, you know, a fifty thousand person CEO and the finance department they have, and a fifty person founder type, and the finance person they have, is like night and day difference. Where did you decide to focus, and why did you pick that ICP to start?

Siqi Chen (00:17:56) :
Yeah, so I kind of have regrets, and I am ambivalent on my regret on this. So we picked the founder startup architect, I think myself. The reason is because I understood the problem I've been through before. I can get distribution into it because I'm plugged in. I understand the persona, and I was a customer. And I think the last one is particularly important. There was this viral thread about how Twitter or how Cursor doesn't have PMs, and I think Linear also said something similar. And people are like, yeah, we should have PMs. And that actually makes sense, but you have to really think about for whom it makes sense. It makes sense when the builders are the users, right? But not every company, product, or market is going to be that.

Pablo Srugo (00:18:50) :
But most would probably not be that.

Siqi Chen (00:18:52) :
Most would probably not. In those markets, you probably want FP&A. Ideally, you want as much overlap as possible, but it's not possible in most companies. So we picked that market because it was a market where I could be a user and understand what is good.

Pablo Srugo (00:19:11) :
When you did those hundred conversations, was that all already with this predefined ICP, or did you go broader for that initial set of seeing what the problem looks like?

Siqi Chen (00:19:20) :
Yeah, we went broader. We talked to finance leaders specifically, and we got a deeper understanding of what the pain points are. And basically, there was a decision point. And the decision point is, do we focus on the accounting bit, the source of truth, compete effectively with QuickBooks, or do we focus on the financial model way of forecasting a bit? Initially, what we started with is a front end on top of QuickBooks. Because we didn't want to compete with QuickBooks directly, because we thought that would be difficult. It might have been a good idea to compete with QuickBooks early, because that would have actually been easier to build and the market would have been larger.

Pablo Srugo (00:20:06) :
Then to integrate into QuickBooks?

Siqi Chen (00:20:08) :
Yeah, then to build on top. Because people were actually highly open to this idea of a better NetSuite, a better source of truth. But at the time, what we optimized for is, what is the fastest path to learning. Which was, in retrospect, not a bad decision. But we ended up building a relatively easy to pick up product for small businesses. So it took us less than six months to really launch this.

Pablo Srugo (00:20:36) :
How did V1 work? What exactly did it do?

Siqi Chen (00:20:38) :
Yeah, so to use your experiences, you have QuickBooks, and you sign up for Runway, and you just want a really good view into your expenses. And see how much cash you have, how much runway you have, and then be able to see month by month. All of the expenses and what categories it went to according to your own organization, right? So it was so fast for you to just organize things into whatever buckets that you want. You could just drag and drop. So, you know, Mint, right? Mint.com.

Pablo Srugo (00:21:09) :
I was going to say Mint, right? And Mint had the problem where it miscategorized pieces. So I'm curious how you, as a founder, quickly would say, these are, you know, marketing, and this is project A, and whatever.

Siqi Chen (00:21:19) :
Yeah, so that was the coolest thing that we built. Later on, we were sort of one way, we kind of undid it but we're going to bring it back. Because it's so awesome. So in Mint, when something's miscategorized, how do you fix it, right? You have to open up the details and then recategorize it, or write a regular expression rule. And then say apply it to all future transactions matching this, whatever, right? And that's a lot of work. I feel like you're constantly fiddling with it. So nowadays you can probably write an AI rule, which is nice. But when we did Runway, V1 was very cool. We had a drag and drop. So basically, if for some reason your Cloudflare expenses went to the sales team department. You could literally just drag it in the same tree into the engineering department, and it'll just remember. And that works on any level. So you can drag an entire vendor, an entire account, or just one transaction, and it will just learn. It was just so fast and then you can create new categories, new dimensions very quickly. So if one day you want to organize all of your expenses by project, you could just create a tree called Projects, and you can create names of your projects, and all of your existing expenses will be under no project. So then you can just roll into the same tree, and they say, okay, everything in Sales is actually going to be this project, and you just drop it there, right? And it's just so fast to organize your expenses and your income.

Pablo Srugo (00:22:49) :
This was like the Figma inspiration in terms of the user interface. At least with screens that tied other screens.

Siqi Chen (00:22:54) :
Figma was and still is a fairly big inspiration. It looked very different from Figma, though, obviously, as a finance tool.

Pablo Srugo (00:23:01) :
Okay, yeah.

Siqi Chen (00:23:02) :
But the inspiration that we took from a product like Figma or Notion is the idea that you want to have these building blocks. That are flexible and expressive and powerful. That map to how people think about whatever's in their head. In our case, of business, in the case of Figma, it's how they think about a product or design. In the case of Notion, it's how they think about the knowledge that fits in their head. So in our case, yeah, the inspiration was, what's the equivalent of a component in Figma or a block in Notion for us? And we call them blocks in Runway V1. And the idea is, one block is, I have as a concept this project, I have as a concept this department, I have a concept of this company. And you can think of a company as the sum of these blocks, right? In terms of expenses, at least. You have your sales expenses, your engineering expenses, your hiring expenses, your revenue and the sum of that is sort of that income. And so can we just make all of those the same thing? Can we make it sort of recursive? That was the idea.

Pablo Srugo (00:24:07) :
And the end outcome, like the value for this founder-operator. Would be a clear view that he or she really understands in a way that's speaking their language, not accounting language, but also not incorrect language that they might set up on a spreadsheet that then is impossible to reconcile and starts to get too messy.

Siqi Chen (00:24:25) :
Exactly.

Pablo Srugo (00:24:26) :
How did the launch go?

Siqi Chen (00:24:27) :
Well, we actually technically didn't really launch until a year ago. So it wasn't much of a launch. I tweeted about, you know, how we're thinking about the business and what we're doing, about where we want to take it.

Pablo Srugo (00:24:39) :
And this is 2021 at this point?

Siqi Chen (00:24:41) :
This is late 2020.

Pablo Srugo (00:24:42) :
Okay.

Siqi Chen (00:24:43) :
And we just had some people sign up, basically. It was PLG, so anyone could sign up.

Pablo Srugo (00:24:49) :
Free, or was it a freemium? What was the model?

Siqi Chen (00:24:51) :
It was free at the time we weren't charging and we got very strong reactions. People really, really liked the product. But the problem is, they weren't willing to pay a whole lot of money for it and that's like one of the learnings. Is that there's an enormous amount of difference between something that is valuable and something that people really like a lot as an experience. You want those to be the same thing as much as possible, but oftentimes they're not necessarily the same thing, right? Value is about solving really deep, impactful pains that cost people a lot of time or money. I think Sequoia describes it as a difference between a vitamin and a painkiller. And vitamins can be sugar coated, right? They can taste good and so what we learned is that as a founder, you think you have the problem of wanting to understand your finances better. So you're like, okay, I need to get a fractional CFO, I need to get these payments, I need to be good at this, I need to get Runway V1, and it'll tell me exactly where my money goes so I can do a good job of being frugal or whatever. And the thing is, Runway does that. It was very good at doing that. It did the exact thing that you wanted and what founders found is that in every small company. All of your money went to people, maybe a little marketing. That's it, that's actually it and you had your Figma subscriptions or whatever, but that's basically a rounding error. Next is salaries and if you're in the market. Maybe marketing, as a small company, buying large. That's what we see and that's cool to see. But the next month, that doesn't really change. The month after that doesn't really change and you're not spending that much money anyway, because you're small. So how much money are you going to spend on this finance platform that tells you these things, but it's not really that interesting. Because it was actually really good at telling you the truth, and the truth is, it's not that interesting. You should be spending your time on getting to product market fit, then looking at your finances.

Pablo Srugo (00:26:48) :
I'm going to ask you for a small favor, a tiny little favor. In fact, it's not even, now that I think about it. It's not even really a favor for me. I'm actually trying to help you do a favor for you. Just hit the follow button. You won't miss out on the next episode. You'll see everything that we release. If you don't want to listen to an episode, you just skip it, but at least you don't miss out. It's funny, you're surfacing the classic eighty twenty, maybe it might even be ninety ten in this case and by doing so, you're kind of getting rid of a problem. But because it's a one time thing, once they see that it's not a thing, that everything else is noise. The only outcome could be, if you're rational, okay, like you said, I'll go spend my time elsewhere. Thank you very much, but I'm certainly not going to pay you. I mean, maybe I'll pay recurring, but if it is going to be like 10 bucks a month. Because why would I pay anything more than that?

Siqi Chen (00:27:33) :
Yeah, exactly. So then, at the same time, we had a wide swath of companies try this, and some of the companies were larger. And for the larger companies, it was cool, but they had very different problems. And most of their problems were future looking. They needed to understand the impact of decisions worth tens of millions of dollars across capital expenditures, hiring, strategy, and they wanted a better tool to do that. And their tool today is Google Sheets, right? And in order to solve that, you had to build effectively Excel. And we knew that would be really, really difficult. So we didn't want to start with that. But that is where the value clearly is when your problems are at that scale, right? You're willing to pay a small percentage of the cost of that problem in order to make a slightly better decision. But building that product was going to be very time consuming. So that's part of the reason why we didn't start with that, right? But we learned fairly quickly that actually, in order to make a dent in this understanding and offer your business space, you kind of have to build that.

Pablo Srugo (00:28:45) :
What was status quo though in that space? Because that is, at least classically in my understanding, the FP&A space, right? Financial planning space and so what were people. Spreadsheets have always been part of that tool set, right? But what else were people using, and what were the blind spots or the gaps that were left?

Siqi Chen (00:29:03) :
Yeah, so people were using things like Anaplan, Adaptive Insights, Planful, Cube, DataRails started a few years before that, Vena and so these are tools that either would bring data into your spreadsheet or will replace your spreadsheet entirely. And the value is twofold. One is, so backing up for a second, what are the highest pains of a finance team historically? The highest pain has been actualizing your model and how much time and effort it takes, which is clock time every month. Because what happens every month is that time passes and you are maintaining this model. This model needs to forecast to the future, right? And for any simulation to work, you need to have a starting state. The starting state needs to be where you are today and so your general ledger, your QuickBooks, your HR systems, your CRM, your data warehouse. All of your existing sources of data are effectively that starting state. So what that means is every month you have to pull the data from all of these different sources, and there are more and more of them with more and more data. And you have to somehow twist them into a configuration that you can use to update your model. So that's just enormously painful. The second thing is, in order to plan and collaborate, you need to talk to a lot of people and get information from them. So if you've ever run a department at a mid-sized company, every quarter you're likely to get some kind of spreadsheet in an email. It could be a Google Sheet, it could be an Excel, and it's just this template from the finance department. And you just fill in your budget and travel, and your revenue and your salary, whatever. And then you email it back. And you can bet that's probably the last thing on your to-do list, right? Because you have a hundred other things you'd rather do and the finance team is just nagging the different department leads on Slack to please do this. So that's annoying for the finance team to have to do and the third thing that happens is just business is enormously complex and you can't trust your model anymore. So at some point, all of these pains become so painful that you go on Google and you're looking for some kind of software to solve for these pains. And that's when you buy something like Anaplan or Adaptive. And the problem with the experience is that if you look at the NPS scores of all these platforms, they're in the gutter. Everyone has really negative NPS scores and so the reason is they actually will solve most of these things I'm talking about. So that's the really interesting thing. So all these platforms connect to data integrations like we do. They have some budgeting workflows where you can assign tasks to different people and project manage your quarterly budgeting review cycle, your annual planning cycle. And they're dimensionally modeling tools, so it reduces the number of formulas in your model by a significant amount. What happens, though, and this is a really subtle thing about a space. After you spend six months, sometimes two years implementing one of these things, and you're paying implementation specialists to do it. What happens at the other end is that inevitably you realize one day that you need to make a change to your model somewhere. Maybe there's a new business unit. Maybe you decide you change your go to market motion. Now you have a PLG motion and now you have to change your model, and it's basically impossible. You have to hire an implementation specialist to rebuild everything. It's going to take another three months for something that you could have done in an hour in a spreadsheet. Because spreadsheets are really good at that. They're really flexible and so that's the nuance of this problem. And the reason why it's happening is because when you are a user or a potential customer, you're experiencing these pains. You're not thinking about that six months from now because you don't have that pain right now, you're on Excel. Excel is great for that but you buy this thing and six months later you find out that you can't do it.

Pablo Srugo (00:33:07) :
And aren't some of these, like Vena? I thought specifically, built on top of Excel, no? Has Excel as a backbone?

Pablo Srugo (00:33:12) :
Yes, so that's true. There's two different problems with that. So on Vena and Cube and DataRails. Their churn problems are fairly acute, and we get people switching from them all the time. And the issue with that is, one, the structure of bringing in your data into your spreadsheet also makes your spreadsheet rigid. You can't just willy nilly change it. It assumes your spreadsheet is going to be a certain structure, so it does very dramatically make it rigid. So that's one problem. You can maybe try to work around that. There are different solutions that do better or worse jobs, but that is one part of this reality. The other part of this reality is, well, yes, even if it does solve that problem. It doesn't solve the other problems that the Anaplan and Adaptive platforms do solve. Which is around collaboration and version control, and all these different things. And so at some scale, you will need to solve those problems too. So it's either one or the other, and they're both broken. And it just hasn't been, and it's very, very difficult in retrospect. Which is what we've learned, to build something that has the power of a finance platform but also the flexibility of a spreadsheet. And that's been, I think, unsolved because it's just really, really, really hard.

Pablo Srugo (00:34:31) :
So when you decide you want to go there, do you then do this again? Do you have another one hundred conversations with people? How do you decide? Because you said it's a very complex problem. How do you decide what you're going to build as your new V1?

Siqi Chen (00:34:46) :
So it is a combination of talking to people, and now we have customers to talk to. But a lot of it, you know, is born out of also my authority experience. I built hundreds of models when I was head of product at Zynga, right? I understood modeling. I have some of the muscle memory of using Google Sheets and Excel. So I think having some intuition for what the opportunities are and what is minimally required to get to an outcome is highly useful. And so you got to start somewhere. And the reasonable place to start is work from first principles. What is actually super minimally viable? What is absolutely required? And what is absolutely required is a way of modeling, is a way of writing formulas. So we started with that.

Pablo Srugo (00:35:32) :
Do you rebuild a spreadsheet then?

Siqi Chen (00:35:35) :
Eventually, yes. But to start, you don't have to build all of a spreadsheet, right? So one analogy is when you think of Airtable. Airtable is very spreadsheet like. it has a lot of overlap. It chose a different primitive than Excel, it chose a row but when you can choose a different primitive. When you can basically decompose or unbundle a use case from Excel, you don't have to build all of Excel. You can just start with a minimal set of features as required for your specific workflow. So that's what we did. We had addition, multiplication, we had this idea of a person. So you could sum up employees and salaries. This idea of a plan, and then you can extend on top of it. So it ultimately took another two years for it to really start working.

Pablo Srugo (00:36:25) :
Wow, what do you do? Tell me about those two years, because there is a subset of, lean startup embedded in everybody, right? Even the concept of MVP is generalizable in the sense that even if your MVP takes two years. It is still technically an MVP, just one that takes longer to build. But it is not immediately obvious how you can go about building for two years without having a huge risk that you are just building the wrong thing, right? And then you spend two years, and you launch it, and nobody cares. How do you balance it during those two years when you are building it out and making sure that the thing you end up with. Because one thing, if you take two weeks, you build something, you put it out, it did not work, who cares, you throw it out, right? Take two months, it is big. You take two years, you are like, I better be right that when I have done this thing, people are going to pay me for it. How do you make that happen?

Siqi Chen (00:37:09) :
That is the single toughest thing about the past couple of years, is going through that experience. The truth of it is very painful because you do not know, and you are doubting yourself every day.

Pablo Srugo (00:37:22) :
And do investors doubt you? Stakeholders? Employees? That is the other thing. You have to keep everyone on this. We are going to get there.

Siqi Chen (00:37:29) :
Employees, definitely. Investors, less so if you have the right investors. Yeah, it's very hard and what made it easier are a few things. One is you try to talk to customers as much as you can, and you are constantly reinforced that this is real pain. It's empirically true that there is real pain, because the market already exists. There are multiple examples of billion dollar companies in the space. So you know that this is a valuable thing. You also know that the status quo is bad. One of the things that helps is having other examples. So I had actually an expectation coming in that this would be a three and a half to five year build. From the day before I started the company, that was my basic expectation and that is because when I look at the most valuable companies, and products that I admire, Figma, Notion, Airtable, Coda, all of these were three and a half to five years to launch. It was six years for Figma to get to one million dollars in ARR. Six years, and I remember talking to Dylan, Dylan is one of our investors, when I think I was two years in. I asked him what was the most surprising thing about taking Figma from zero to where it is today and he said what surprised him the most is how few people made it all the way. And the more I thought about the answer, the more I think that is remarkable. Because Figma, now we know, is one of the most successful private outcomes of all time, actually public outcomes now, right? And even for Figma, you could be part of the early team before they got to their first million where nothing was growing. You could look left and look right, and half the people thought they were not going to make it. And so this is why I think about Paul Graham, recently he said, or I think it was Jessica. She said that when they started Y Combinator, their assumption was that startups are a test of intelligence, right? That you must be very smart and they were going to find these very smart people and they were going to be successful. And over the course of the past fifteen years and more, almost twenty years now. What they have learned is that startups are a test of stamina, far more so than intelligence and so I think about that.

Pablo Srugo (00:39:34) :
My question is during that time, because it is true. Figma, Notion, Airtable, they built amazing products and they were under the scene, in build mode for a long time. What is your litmus test as you develop things that this feature is built the right way, customers will love the way that this feature is built? The thing that you normally do in this iterative step, in the normal world, and you kind of get feedback from people. Not so much that they tell you, but at least you can see the usage, you can see the patterns and you get the quality, you get it all. In this mode of building, how do you make sure along the way that you are not going too far in one wrong direction before pulling back and saying no, that is not built right?

Siqi Chen (00:40:11) :
So we had design partners, right? We were never building in a vacuum. We had design partners from day one, and we would turn through them. Because the product just didn't do much, and it had to do a whole lot before anyone could really use it. But the good thing is, when we did churn through them, they weren't upset about it. They were just like, we really believe in what you're doing, so don't even sweat it.

Pablo Srugo (00:40:32) :
And the expectations, I guess, were set the right way from the beginning.

Siqi Chen (00:40:35) :
Right, as a design partnership, we're early and that could be a very good gift. The other thing that helped, I mean, the question was, how do we know what is good? And one of the classes I recommend every single founder and product person take. Is the Product Sense course from Shreyas Doshi. It's the single most valuable course, I think, in the world. I'm not even exaggerating. It is just that good and one of the classes was on taste. His definition of taste is as follows. Taste is the ability to know what's good, without the need for external validation and the ability to explain why. And so the answer to your question is you have to bet on your taste on some level as well. Startups and good products, good products do not get created as a result of a random walk. And what I mean by that is you're going to randomly try things and then see what happens. Good products come from, yes, empirical validation, but also a strong sense of intent and taste. And you have to trust that. It takes an enormous amount of will and stamina, and belief to trust your taste. As well as discipline to get feedback on the things that you build. So you have to do both.

Pablo Srugo (00:41:56) :
When did you finally launch this product of Runway?

Siqi Chen (00:41:58) :
So we went to general availability a year ago.

Pablo Srugo (00:42:01) :
So let's say mid, late '24.

Siqi Chen (00:42:04) :
That was four years in.

Pablo Srugo (00:42:05) :
Your pre-seed is relatively easy, just in the sense that you're a proven founder, you've got a thesis, okay, let's go. How do you raise your seed and then how do you raise your A. Which, correct me if I'm wrong, says $27.5 million, right? What is the story you're pitching in 2023 to raise $25 or $30 million dollars if you're still mainly in that kind of build mode?

Siqi Chen (00:42:26) :
Yeah, well, there are certain products that just require a lot of capex, right? Building an offering system for a company is certainly one of those products and those products can be the dimensions for a few reasons. One is if the market size is that large, and the addressable market is basically every company, and the opportunity is large, and there's a wedge into the opportunity. It's clearly underserved. Nobody's happy with any of the products that they have. So that's one, another reason why you would want to invest is because you believe that there is a disruptive opportunity, and this team is uniquely qualified to do it. And so the reason why we would raise such a large round is because we could show the product.

Pablo Srugo (00:43:11) :
Well, that was going to be my question, were they close enough to the space to understand and have taste? Because to have taste in a product, you have to know the use case, the ICP, these sort of things.

Siqi Chen (00:43:19) :
Yeah, so where we differentiate is we're building for people inside a company, right? Finance people are people too is one of our mantras. We say finance is not just finance, finance is everything that helps in the company. So being able to demonstrate the value, but also what a product built for people that makes the finance of the company understandable feels like. What it should do is even if you're not a finance person, you can look at this product and say, that makes sense, I get it, right? It's intuitive, and just like how if you truly understand some domain, some discipline. One of the hallmarks of that understanding is you as an expert should be able to explain concepts in very simple terms to a five year old, right? A sort of amateur wouldn't be able to do that. That's one of the hallmarks of mastery. I think one of the hallmarks of a great product should be that someone with no experience whatsoever should be able to look at your product and understand how to use it, have it make sense and be useful to you. Because ultimately humans use it and so that's what we were able to show for the first time in this domain. And then you draw some lines and you say, oh my God, if this works the way it's supposed to work. And I see some evidence of it, this could be an extraordinarily valuable company.

Pablo Srugo (00:44:42) :
What was that kind of demo that you're showcasing in 2023? What were the key things that were enabled by that version?

Siqi Chen (00:44:48) :
Yeah, so we showed a combination of the Runway V1 pivoting, but also this idea of plans. And plans allow people to connect operational intent, your roadmap, a project, the things that you do, to the model for the first time. So inside a spreadsheet model, and this is the thing. People think finance is a separate thing, and that finance is a separate thing. And basically, our entire belief is not, it really isn't. It just seems like it because it's really complicated and opaque. But it's not different, it's not separate, it's actually intimately connected to everything that you do. So an extreme example is a product roadmap. So you spend all this money on designers and product people, engineers, to improve the product. But why? Why are you doing that? The reason why you do that is because at the end of the day. It increases the value that people are getting from the product, so maybe you can command higher pricing, or it increases value so that your attention goes up and your churn goes down, or it improves conversion rates, or it expands to a new market so your market size is larger. It does these very specific financial things and that context is either disconnected from the model, or it shows up as a good model will have a line for conversion rate. And maybe you change a number here. But that number doesn't just change on its own, it changes because you did something to change it. You ship a new feature, there's a new initiative. There's a connection there that just doesn't exist. So we create these constructs that connect the operational intent with your model for the first time, and we're able to show that. And so having a good story, but also having the evidence where they can see with their own eyes the proof of that story. At least, if not completely true, we're on the path to making it true, was the thing that did it.

Pablo Srugo (00:46:37) :
When did you actually publicly launch the product?

Siqi Chen (00:46:40) :
Inter-general availability, July of last year and we launched this sort of beta for design partnerships marketing site a year before that.

Pablo Srugo (00:46:49) :
And how did the public launch go? How did you structure it?

Siqi Chen (00:46:53) :
Yeah, so for the public launch. The timing was, do we have enough confidence in our product market fit that we'll really start scaling out the machine? About two years ago, AngelList started running much of their business on Runway, and we thought, if AngelList can do it, then we think a lot of companies can also be successful here. So that's when we started the plan for the launch and for the website itself. For some context, two years ago we launched a website, and it went highly viral. I won a Webby, a bunch of people retweeted it, and it got a few million impressions. It wasn't even supposed to be a launch, I literally tweeted, new website, some new roles, check it out, that's it and people were like, this is the most insane website they've ever seen. So that was sort of an unintentional launch, but we weren't really live or generally available until a year ago. A year and a half now. For the launch, we just wanted to do a launch campaign. It's our coming out story where we can tell people what we're about, what makes us different, tell a story of Runway, why we exist, and show a demo of the product. Really, what makes Runway successful is people will be able to see, use, and feel the product, and we really wanted to make sure that's what people saw. So we commissioned a launch video, spent a bunch of time thinking about the creatives, and that was fairly well received, knowing that our main distributing channels are going to be LinkedIn and Twitter. In conjunction with that, we also thought about how to amplify the story, right? It's not like it was today where every startup is launching with some highly produced video, right? I think that wasn't quite as common a year or two years ago and the application strategy was, alright, how do we do something really cool, right? Because what worked for the website is we just took a normal marketing website and thought, how do we plus it? in, you know, plus it is a term coined by, I think the Walt Disney company. Which is how to make things that much better. How do we plus it. What people normally do is, in a marking website just take it to an eleven? That ten percent difference isn't a difference of ten percent in outcomes, it could be one hundred times. We applied the same reasoning to what we could do for our launch campaign. Normally, people may get swag, maybe some T-shirts, do outreach, and a lot of people will just do the video and the tweetstorm. And what we decided to do was, OK, lets do swag, but make it really insane, and have that itself be a story tied to the launch. So, a launch is a story, there's an asset associated with that story, it's on different channels, and then there is the time of the launch. You want all amplification to be concentrated as close to the launch time as possible, and you want a wide network of people ready to amplify that message, right? Those are the sort of ingredients we thought about and so our idea for timing and the network of influencers that wanted to amplify it came down to this idea of let's do swag, but do it better, right? Not just a T-shirt, but a bomber jacket and that bomber jacket is highly designed and branded, and it's bespoke, and it's unique. There's like only two hundred of them, individually numbered. And let's send them to our supporters, best customers, and some influencers. Then we thought, alright, but that still not that memorable enough, still just receiving something in the mail. So we thought, how can we make it more interesting? We brainstormed and we eventually landed on this idea of, what if, you know, it was a sense of mystery. I think about mystery sometimes when thinking about creative ideas for product or marketing. The reason why I think about that is because the guy who made Lost, and his name escapes me. But a producer of Lost, J. J. Abrams. Talk about the mystery box he grew up with, nobody knew what was inside, and that was the point. If you open it, it becomes less interesting. When things are mysterious, that itself triggers something deeply innately human in terms of curiosity. So the idea was to put the jacket in a bag. We thought it should be a chest, but what's practical is a helmet bag, and let's lock it. And we also thought about, okay, our first idea was we were going to lock it but not send the key until launch day. Then we brainstormed variations and sourced a timed digital lock, which was very hard to find, a very niche product. 

Pablo Srugo (00:51:39) :
That's wild.

Siqi Chen (00:51:40) :
But yeah, we sourced a timed digital lock and the lock would only unlock at the exact minute and second of our launch day. So we asked influencers for their addresses, and people are always into a mystery or say it's going to be a really cool surprise. They're like, cool, I can't wait and we mailed out the bags, and people started posting about them. Nikita asked, what is this? And I said, maybe it is a bomb. I just got this rumor. We were already tagging our company beforehand, and then when it all launched on launch day, people were posting about their jackets.

Pablo Srugo (00:52:19) :
What was the outcome? I'm curious, ultimately, how many impressions did you get? And then ultimately, how did revenue flow from that?

Siqi Chen (00:52:27) :
Yeah, so our first million revenue was from me tweeting through it.

Pablo Srugo (00:52:32) :
In how many months?

Siqi Chen (00:52:33) :
I mean, it took us three years to get to one.

Pablo Srugo (00:52:36) :
Three years, but in post-launch, yeah, like.

Siqi Chen (00:52:39) :
Yeah, post-launch it exploded very quickly. I think we had about $7 million impressions on launch day, and on our website two years ago it was I think $5 million. And leads were coming in so fast that I had to build a GPT automation that would qualify leads automatically because we literally couldn't qualify them fast enough.

Pablo Srugo (00:53:04) :
How quickly did you sign another million in ARR after the GA launch?

Siqi Chen (00:53:08) :
It was within the next thirty days or so.

Pablo Srugo (00:53:12) :
Wow.

Siqi Chen (00:53:13) :
Yeah.

Pablo Srugo (00:53:13) :
That's a good launch, man. That's a good campaign.

Siqi Chen (00:53:15) :
We're not product-led growth, so we're sales-led growth, and we have to run through sales. Our average time to close is about twenty days, and that's enormously fast.

Pablo Srugo (00:53:25) :
And then how have things gone since, like growth wise?

Siqi Chen (00:53:28) :
Really well. We've quadrupled year over year, which is exciting. But yeah, there's still a lot of work to do. That was a good start for the rest of the year.

Pablo Srugo (00:53:38) :
Perfect. Well, listen, we'll stop it there and let me ask the last three questions we always end on. First one is, when did you personally feel like you'd found true product market fit?

Siqi Chen (00:53:45) :
Yeah, that moment was when AngelList not only signed, but said, hey, we're actually going to run our business on Runway. They were happy with it, right? They were telling their friends about it and we did a case study with them. And they were just shouting over roof tops about how pleased they are with Runway and had never seen anything like it. That was the first moment. Before that, it was just a slog, and it's been a slog since then. But that first major happy customer who's not just satisfied, but over the moon. Because it's so different and new and it made them feel something, that was that moment. That was about two years ago.

Pablo Srugo (00:54:21) :
And was there a time, especially while you were building. Where you thought maybe things wouldn't work out and you would just fail? Especially in your case, spend a lot of time and then fail.

Siqi Chen (00:54:31) :
I mean, I want to say, of course, my first instinct is to say, of course, all the time, right? And no matter how big you are, you still have that fear. You still have that feeling. Product market fit is not just one instance, right? It's a continuum, and you keep on capturing deeper product market fit, product market fit in new segments. But I don't know if that's actually true. What I would say is there were times where I had no idea and the team had no idea how long it was going to take or if it was ever going to happen, right? But at the same time, what I felt was, it does not violate the laws of physics, and people are not happy with their existing products, and there is a better product to be built. It exists, we just haven't built it yet and I don't know when exactly it'll happen. But I always believed that it would, if you just keep on going. Because my expectation has been set by looking at Figma and Notion and Coda and all these other companies and how long it took. If you just keep going, not all the time, but if you don't keep going, for sure, you will not figure it out.

Pablo Srugo (00:55:38) :
And then last question, what will be your number one piece of advice for an early stage founder looking for product market fit?

Siqi Chen (00:55:43) :
I like to think of it, my mental model of this is that there is this lighthouse that you're trying to reach for, that's your vision and the path toward product market fit is this blindfolded walk through terrain. And you have to be able to have this lighthouse in your vision at all times, but you have to adapt very flexibly to the terrain you're on. So you have to be infinitely flexible in the short term while being very stubborn in the long term.

Pablo Srugo (00:56:10) :
Awesome, well, dude, it's been great having you on the show. Thanks for sharing your story.

Siqi Chen (00:56:14) :
Thanks for having me.

Pablo Srugo (00:56:15) :
Wow, what an episode. You're probably in awe. You're in absolute shock. You're like, that helped me so much. So guess what? Now it's your turn to help someone else. Share the episode in the WhatsApp group you have with founders. Share it on that Slack channel. Send it to your founder friends and help them out. Trust me, they will love you for it.