🎧🍌 How To Go Viral with Siqi Chen (Runway)
Creating the fastest growing product ever, getting shut down by Tim Cook, making B2B more consumer-ey, secrets to storytelling, when to forget best practices, and tactics for fundraising
The episode is brought to you by Attio
Attio is the next generation of CRM.
It’s powerful, flexible, and easily configurable.
Join OpenAI, Replicate, ElevenLabs, and more and try it today
To inquire about sponsoring future episodes, click here.
Siqi Chen is the Co-Founder and CEO of Runway, the modern and intuitive way to model, plan, and align your business for everyone on your team. Or in Siqi’s words, “revolutionizing the $80 trillion business industry”.
A few months ago, Runway’s new website went viral across the internet. Siqi takes us inside how that happened. He’s no stranger to going viral, having previously built multiple companies that went from zero to millions of users within weeks, including the fastest growing product ever before ChatGPT took the crown in late 2022.
Timestamps to Topics Discussed:
(02:59) How Siqi goes viral
(06:13) Why conversion rate doesn’t always matter
(11:23) How to make B2B software more fun
(14:31) Working on the Curiosity and Spirit rovers at NASA
(16:50) Re-designing at the entire codebase and product at his first startup job
(21:13) Earning the nickname “FB Millz” making a million dollars building Facebook games
(25:11) Selling to Zynga and building the fastest growing product before ChatGPT
(28:36) Building a game with 90% Day 1 retention
(30:09) Being played by Kim Kardashian, Jack Dorsey, and getting shut down by Tim Cook
(32:10) Almost getting fired building a growth team at Postmates
(35:04) Building Sandbox VR: “escape rooms in VR”
(40:25) Meeting Kanye
(44:39) Getting the idea for Runway when COVID hit
(47:32) Why spreadsheets run every business
(54:40) Disrupting the $80 trillion business industry
(57:55) Making formulas 50x easier than Excel
(01:02:32) Why Runway’s building a painkiller
(01:06:27) How to fundraise
(01:08:32) Why the first question from an investor is the reason they won’t invest
(01:11:29) How to tell a company’s story
(01:15:24) The three layers of a story
(01:17:28) The importance of positioning in storytelling
(01:18:49) Runway’s flexible remote work strategy
(01:21:34) Why their hiring strategy changed over time
(01:22:39) Siqi’s single interview question & the three traits he looks for when hiring
(01:26:10) Unlearning consumer to learn B2B
(01:30:26) Navigating the first three years of no customers
(01:31:45) What surprised Dylan Field the most about building Figma
Find Siqi on Twitter and LinkedIn
Referenced:
Runway’s design agency, Reform Collective Design Agency
Transcript
Find transcripts of all prior episodes here.
Turner Novak:
Siqi, how's it going? Welcome to the show.
Siqi Chen:
It's going great. It's good to see you, Turner.
Turner Novak:
So I wanted to really quick hit on first, I think it was a couple of months ago, you redesigned the website for Runway, and it went viral. I remember ... I think we are in at least one group chat together, other group chats I'm in ... I think I was really busy that day, and I just kept seeing it popping up on my phone, and I was like, "What is going on?" I was seeing it and while I was in between Zooms, "What is this?" And I looked and the website took over the internet basically. What happened that day?
Siqi Chen:
Yeah, it was nuts internally, but we released this new website, and we were working on it between that and branding work for about six months. And I just wanted to get it out there, and get some feedback on it, and it was I think a Saturday, and we had it up, and we just pointed the DNS to it, and we said, "Hey, we got some new jobs." Oh, we wanted to hire some people, had some new roles, and a new website. "Check it out." And it just exploded. That's what happened.
Turner Novak:
So why do you think it happened?
Siqi Chen:
Yeah, it actually exceeded our expectations wildly. It was something that we were hoping to happen. During the development design process, the thing we kept on thinking about is, how do we do something that tells the story of what Runway is, but pushes what people think good design is, in the space of B2B forward, and is built for not just conversion ... actually not at all for conversion ... but for interest, like surprise and entertainment.
And so all of that crystallized to something that was fresh and new. And so this is actually tied to what we think our mission is at Runway.
So when we look at what we're building, we are building this finance platform, but finance platform people don't like using ... not even finance people. And the real end goal of this is to make this accessible to everyone else in the company too, not just finance people. We think that has very deep impact and meaning around there. And in order to do so, we want the website to reflect that.
And so what does it take to achieve that mission? It has to be a product that you can see, and something that is approachable, and the vibe is like, this is not professional software. It's a relatable, we talk to you like a human. And so those were some of the thought that went into design.
But I think the other thing is, what is the opposite of a B2B website? It's like stock photos. You don't show a product. Let's just do the opposite and see what happens? And people have been hungry for something different, and I think that's probably the biggest lesson, is that this idea of best practices. We follow no best practices. I got an email last night from a woman who wrote a loon saying, "This is everything you're doing wrong in terms of B2B best practices on your website, and here's how you should change it." And I said, "Thank you very much, but we did this very intentionally, because best practice, it being the same as everyone else. And on a meta level, if you do the same thing as everyone else, you're going to get the same results."
Turner Novak:
One really interesting point that you hit on, was it was not optimized for conversions, and I don't even think it's live yet, right? It just says get access. You can't even really convert and start using it yet, so it wasn't really like you're trying to ... you were just like you said ... put it out there ... guy or some people. How did it go?
Siqi Chen:
We were getting a lead every few seconds.
Turner Novak:
Okay. Wow.
Siqi Chen:
Just the slack was just going through. So we have some internal AI tools, and one of the things we built is automatic lead qualification. So anytime someone will come in, enter an email, we would clear bid it. So we have some more information on what it does, who the person is, and then we have very specific criteria on whether these qualified, that the AI tool would explain in plain English language like, "This company, because they're that size and this role, this is why they're qualified."
And that thing was just pinging every few seconds, it went off, which is pretty wild. But I think one of the things, if you're thinking about it from just a pure growth perspective, we didn't optimize for conversion, but we did optimize it for growth. And I think that's the mistake that a lot of growth projects make, which is they're not looking at the entire funnel. The conversion is probably lower, but the amount of impressions and the amount of attention got is like a thousand x higher than it otherwise would've been, and the overall results were way better.
And so sub-optimizing for a portion of the funnel is a classic growth mistake, and we tried to not do that.
Turner Novak:
Yeah, it reminds me of, when you think about if part of the point was just hire some people, let's say, before this experiment for the website, no one knows what Runway is. You're trying to hire people. It's it own set of challenges, but let's just say 10 million people saw the website. I don't know what the number was, but basically your entire pool of candidates and people in our world, all saw the website and they're like, "Holy shit, this website is amazing. This looks like a good product."
Then when someone's interviewing or getting recruited at Runway, they can tell their friends and their friends are like, "Oh, that's cool. I know that website. That seems like a good product."
Technically yeah, you had this leaky bucket on conversion, but that wasn't even the point really yet, it was just awareness, knowing it exists.
Siqi Chen:
That's really important, because this connects to some of the experience I've had a Zynga, too, and some of the most important things in business and product is either difficult to measure, or it takes a long time to measure. And so if you're, for example, optimizing for direct response conversion, you're looking at how many clicks did you get, and how many people signed up? But what we're trying to get to do is, our buyer may want to think about this once or twice a year, and our goal needs to be that when they do that, they think Runway is in the mix, because we have a lot of confidence that if they just talk to us and look at us, they will end up buying us. So we just need them to be more aware of us, and that requires something different, too.
So I think, thinking holistically, and having the freedom to think about things more in terms of the long-term results, it's underrated in growth, in general.
Turner Novak:
Yeah, I remember you had a really interesting comment in one of the groups that we're in, I think you remember saying, "I specifically, all these B2B websites look the same," and I've seen that comment from people. I think probably Linear is the example. Everyone copied their website design, and you think it's essentially a function of "best practices". So something becomes a best practice and everyone just copies it, and then it's no differentiation. It's not really a best practice any more.
Siqi Chen:
It's hard and it's expensive in time, and it takes some courage to do something new. I think Linear, there's a reason why people copied it, because it was new and it was amazing when it came out, and that's why it made such an impact. And you're not going to get the same results if you do that. That's not how it works.
We got relatively similar results. People go to different firms. I've heard multiple stories, and the agency would say, "What do you want?" They're like, "We want Runway."
Turner Novak:
No way. That's awesome.
Siqi Chen:
It's happened ... I've heard of it multiple times, and I have dozens of people asking me, "Which firm do you use? I want that same thing." We didn't get there because we used one firm. They're a great firm, by the way, it's called Reform Collective. We worked with them at Sandbox, but we worked with them really closely. It's like, "Here's the vibe, and here's why it needs to be different, and here's how we're different."
And so if you want the same results, you can't just say, "We want that." You're not going to get the same thing.
Turner Novak:
And I think it comes back to a question from a mutual friend, Pradeep. He said, "Should enterprise be more consumer-ey? And then how do you do that? How do you be consumer-ey and enterprise?"
Siqi Chen:
I think it's happening in pockets here, and in terms of the productivity tools that people are used to using, whether that's Airtable or Figma or Notion, those are tools that clearly are consumer-grade products that are subsuming many enterprise use cases.
In general, should enterprise be more consumer-ey? Yeah, I think that's our entire game plan and strategy here.
So the way I describe part of the opportunity for Runway is, you think about traditional software. You have two choices. It's relatively easy, and everyone wants to make something really accessible and easy to use, and friendly and consumer-ey, but the problem is, traditionally you have to build pretty simple products. It's pretty easy to make something like a text editor more consumer-ey. You just type, but it doesn't do a whole lot else. Or you have to build something like a enterprise resource planning platform, and it has to do everything. And because it has to do everything, it's really difficult to use. It's hard to make it good.
And I think that's the tradeoff here, and our observation is that in the last half a decade or so with Figma, Notion, Airtable Coda, this generation of software, what these products have done, is proven that you can actually push forward the efficient frontier of this tradeoff. Things can both become more accessible and a lot more powerful than you previously thought possible, and it was made possible because they designed the right abstractions and primitives, to give the power and flexibility to end user in a really accessible form.
And this is kind of a new thing. People call it no code or low code, but that's the way I think about it is like, "Yeah, you're exposing to the end user the flexibility of software itself in visual form," and I think that's an enabler of this. So it's happening.
The value of happening, I think, is pretty profound. The value is people like using your software, and your software can be fun, and for us, those are two really important qualities, because when software is fun, especially in knowledge work, it makes you smarter, it makes you creative, you can explore and be curious about how the business works, what's going on in your business, and that curiosity is what leads to insights.
I think Amplitude Analytics, that's probably the most well-known analytics platform. It had that quality. You could dig in and play with your data and explore, and that's what I loved about the company. I used to be an investor in it.
Turner Novak:
They're public now, yeah.
Siqi Chen:
Yeah, they're public now, but that's what great software can do if it's approachable and consumery, and enterprise software traditionally isn't that. Everyone hates using it. It's just pure paperwork.
Turner Novak:
I don't really know what the tipping point is, but it's like ... I don't know, you just think of when you say the word enterprise software, I just think of this rigid, hard to use, S-A-P, E-R-P thing, not good design.
I think one of my internships during college, I literally had to upload PDFs into SAP. It's like, "Why am I uploading PDFs into SAP? What?"
Siqi Chen:
Yeah, it's wild.
Turner Novak:
So you have kind of an interesting journey how you've gotten all this consumer learning. I think your very first job during college, I think, was at NASA in the Jet Propulsion Lab. Won a couple awards. My question is how do you get a job at NASA? That's a big thing.
Siqi Chen:
My dad had a friend. I had terrible grades. I had no business being at NASA, and my dad's friend owed him a favor, and offered me a job.
Turner Novak:
That's amazing. But then you ended up being pretty good, if you won multiple awards?
Siqi Chen:
Yeah, that was really transformative for me. I never was a great student. I went to an okay school, and during that job, it felt serious. We were working on the Rovers, of Spirit and Curiosity, and I'd never really built anything in production before, and I've never done machine vision, and I had a really interesting project. I got to teach the Rovers, and we're working on machine vision for it, and I realized that I can do this, I can pick it up, I can build something useful, and that just gave me so much confidence.
And it's one of those things where I first somewhat experienced imposter syndrome, in a sense that, because I was able to build this really useful thing, that it must be really easy, anyone can do it. And I was explaining to my friends, "Yeah, it's not that difficult. It's just, we're detecting the dots to calibrate a camera, and it's automatically," and they're like, "That seems really hard." I'm like, "Okay, but that's just because you don't understand it," right?
But yeah, the first job really changed my outlook on I'm not a loser.
Turner Novak:
Being fair on all of this, my very first job, it was one of my really good friend's mom’s. She worked high up in this manufacturing company. I got a job in the factory, my very first job, and it was basically just, we had a good relationship, and he was basically, "Hire my friend," and I think probably I was pretty... maybe I was good, but anyways ... you usually need that. You need somebody to go and do something for you, that very first opportunity.
Siqi Chen:
I don't think I would be in the same place without that. That was definitely a stroke of luck.
Turner Novak:
You started to get more into tech and startups. How did that happen? How'd you do that?
Siqi Chen:
During the school year, I was looking for a part-time job. I applied for a job in a cafeteria, and I lasted literally a day.
Turner Novak:
Wait, so you got the job, did you quit or was it just like, "I can't do this?"
Siqi Chen:
Yeah, I was like, "I don't like it." So I looked in San Diego for any startups that existed, and one of the companies that was hiring was a company called Webmetrics. They did low testing for websites and web website monitoring for response times.
And my manager, my engineering manager, was a person named Lenny Rachitsky, who is now the world's most famous Product Manager. Number one, most well known ... which is still wild to me. He was my engineering manager. He was really great. So nice, and it was a real job, and I enjoyed that too.
It was just fun to code, but also spent a lot of time on user experience and, yeah, I enjoyed it. It was only one of the very few startups in San Diego, is the answer.
Turner Novak:
Lenny told me there's a story of, I think you told them, "I'll only take the job if you can beat me in Horse," or something and the CEO beat you in HORSE and so you end up taking the job, or something like that.
Siqi Chen:
I don't remember that, I wouldn't have made that bet, because I'm not great at basketball. I think they made that bet, that I have to take their job, if I don't beat them in HORSE, and I did it. I think it was the reverse.
Turner Novak:
Okay. Something like that, yeah. Well he said, "What really stood out though, was that when everyone was having lunch, you'd be reading programming books, getting better, learning, et cetera, and he said you just rewrote the entire code base when you joined, because it because it all sucked." Those are his exact words.
Siqi Chen:
Yeah, I think it did suck, but it's not because I had a distaste for it, I just really enjoyed it. I wanted to be organized. It was like, in software engineering terms, it was a big ball of mud. The entire product was one file at a time, if I believe correctly, more or less ... one giant pull script. And I wanted to make it object-oriented and have it be organized, and I just enjoyed the shit out of it. Wrote a bunch of unit tests and rewrote it. But the thing that he probably forgot, that I'm actually more proud of, is I redesigned the entire product.
Turner Novak:
Okay, maybe that's what he meant, too.
Siqi Chen:
So I did both, and that was rewarding for me, because they actually were interviewing design agencies, I think, and I redesigned it already, and they were like, "Let's just go with that one. The one that Siqi did." That made me super proud.
Turner Novak:
Yeah. And then you did a couple other startup jobs over the next couple of years. I think there was VO, if I'm saying that right, Powerset ... anything interesting that you learned in those roles?
Siqi Chen:
The thing about VO was interesting. I took down their website in production for a few hours and, not knowingly, and no one could figure out what it was, and it was me, and afterwards I was made a product manager instead of an engineer, which was fun.
So that's how I got into product. I got promoted because I sucked to be an engineer. Powerset was interesting, in that I moved here for that job.
Turner Novak:
San Diego up to San Francisco?
Siqi Chen:
Yeah, and I'm surprised I got a job, because everyone else there was incredible. The co-founders of GitHub were there, Chris and Tom, and that's how we met actually.
So yeah, feel really lucky to have gotten there.
Turner Novak:
How did you think about getting those first couple of jobs when you're just right out of school. You're building a track record per se, you're learning? What did you prioritize?
Siqi Chen:
I would prioritize whoever wanted to hire me. I just wanted a job, and I wanted to work at some kind of startup, and Webmetrics and VO, those were the two startups in San Diego there were willing to hire me. There weren't that many.
And Powerset just seemed interesting. They had a great brand, a great team, a mission of building the next great search engine. They were trying to build basically ChatGPT or Perplexity before those things existed. And this is like 16 years ago. The technology was far from there. I went there and the people were fantastic.
And so that's probably the luckiest break I got, aside from NASA, just being able to work at Powerset, even though I was only there for six month.
Turner Novak:
And then you started a company. Did you leave because you were starting a company?
Siqi Chen:
Yeah, so what happened was, so before Powerset, I applied for YC, got turned down, I think it was spring of '07, and I moved to San Francisco for the Powerset job. And I somehow got in with the YC crowd, and we were playing a game of Diplomacy, the board game, at a party. I had so much fun, and right around that time ... this is May of '07 ... the Facebook platform came out, and I thought, "Well, that was a game that is fun, that is social. Maybe someone should build a version of that for Facebook." Except that I was one person. I had no money, and I thought Diplomacy was going to be too tough to build for one person.
So I basically built a version of Werewolf/Mafia for Facebook, and while I was working in Powerset, that game blew up, and I was making a couple thousand dollars a day in ad revenue. And by way, the funny thing about that is, the reason I was making a couple thousand a day in ad revenue is because I wasn't smart enough to make an actual chat interface. So it was a common interface, so while you're playing Werewolf and you want to see the chat, you had to refresh the page continuously, manually, to see new messages.
Turner Novak:
So new ads would pop up?
Siqi Chen:
Yeah, you would get paid by the impression. It was a wild time. At Powerset, word got around, and I was known as, my nickname was FB Mills, which is like FB Millionaire.
Turner Novak:
Okay.
Siqi Chen:
And my co-founder, Alex, was also on a product engineering team at Powerset and said, "Let's build something together," and that's what led to the first company.
Turner Novak:
And then did you keep the Mafia? I think you started a new game, too, once you left.
Siqi Chen:
Well, we were working on it in the evenings after Powerset, and that was a new product, and eventually became Friends for Sale, and the idea behind it was ... it was kind of theoretical, and I was really into Facebook, the idea of the company, and at the time, one of the things they talked about is mapping the social graph of the world, because it was before making the world open to connected. I think it was even pre newsfeed.
I thought about that problem and I observed that the thing about Facebook and MySpace is that they all have the social graph, but the graph doesn't have any weights to the edges. In graph theories, you have nodes and edges, and you can assign numbers to the edges, and the weights ... in social graph terms can represent, for example, how important is this friendship, or how popular is this person?
I imagine it's probably very in demand to be friends with LeBron James. How is that represented in a social graph? And so I wanted to build something that would be able to assign weights to the edges of the graph.
The idea was basically, okay, how do you assign the weights? What if you had a market economy, and how do a market economy work? You have the limited number of friends a person would have, and then people bid on people to become friends, and that's what led to the Friends for Sale game. Eventually it became ... I didn't think of it as a game. I thought it was a social network, and it was designed very much to look like a social network, but that eventually became the second largest on platform.
I think we early '08, we pretty much reached everyone-
Siqi Chen:
... second-largest on platform. I think early 08, we pretty much reached everyone on Facebook. It was just me and Alex part-time, and we were running a rails cluster that was larger than Twitter in 08.
Turner Novak:
Yeah, that's what Lenny told me. It was the largest Ruby on Rails project ever or in the world at that time.
Siqi Chen:
Yeah, and it was my second rails project ever. Never administered a server, never had to horizontally shard a thing. And what was interesting is a lot of the team that was at Serious Business actually went to Twitter.
And for example, some of the tools we built at Serious Business, and very few people know the story, there was a big project at Twitter where they had to migrate their master tweet database from a 32 bit ID to a 64 bit ID, because they were running out of IDs for tweets. They used the tools that we built at Serious Business, which was a company that Twitter acquired, to migrate with no downtime. We were the first because we had to solve these problems first.
Turner Novak:
Because you just had so much unique IDs or whatever.
Siqi Chen:
And two of our engineers ended up being two out of the foremost senior engineers at Twitter.
Turner Novak:
And then I think you were acquired or sold to Zynga at some point?
Siqi Chen:
That was January of 2010, I believe. So that was about two and a half years after we were founded. So, we were about 30 people. We were profitable, but Zynga was just huge and Mark Pincus knew to what he was doing a lot more than I did. It was a good exit. I was able to have some financial security and that was also a fun ride too.
Turner Novak:
Did you still run it while you were at Zynga, or did you start doing other products?
Siqi Chen:
Zynga acquired it and ran it for a short while. Then they shut it down, Friends for Sale, this is a couple years later and traffic was tailing off and their gains were just so much more profitable than anything that we made. So we immediately started working on some new products.
The two products I was involved in is one is this Treasure Island digging game. At a time, it was actually the fastest growing product of all time. So we got to 7 million daily active users in a week or something. And the next product after that, Alex led and I contributed to it as well. Alex was my co-founder, but that was also the fastest growing product of all time. I think we got to tens of million, 30 million active users in 30 days, which was was the record until ChatGPT came around.
Turner Novak:
Wow, that's crazy. So what was that second one called?
Siqi Chen:
CityVille.
Turner Novak:
CityVille, Farmville but you build a city?
Siqi Chen:
It's a Sim City plus Farmville, I'll say you.
Turner Novak:
That concept just crushed it for Zynga.
Siqi Chen:
Oh my God, they were so good at what they did.
Turner Novak:
And then at one point you left, you started a new company called Hay.
Siqi Chen:
Yeah. So two years after acquisition, I was halfway vested, and I thought, all right, we're going to do the next thing. And we didn't quite know what we wanted to do. I co-founded with Alex again and we were experimenting with what is possible in mobile.
One of the things that surprised us is, wow, you can have API access to all your photos, and there's background location. What we have to gather is, and this is 2012, and so the photo app wasn't quite what it was today. It's just a grid of photos.
What we did is we could take your grid of photos and we can organize them and cluster them into events. Like you're at a place, you're at a wedding, you took a bunch of photos with the same location data, so it's really close in time, really close in space as well. And we can say, "Hey, this happened at City Hall," and you're browsing through your timeline of photos and it's just a remarkably different experience.
It feels like, oh wow, this is almost like a diary and a life line. So we combine that with a background location and the idea was you just keep your phone in your pocket, you take photos, you're with different people and we just tell you everything that you did in a day. And so we worked on that for about a year and a half and then we launched it and Apple loved it.
We were a global editor's choice. We're on a front page full banner in the App Store for a couple of weeks, and we got a few million downloads off of that and that was our first iOS project. So we're pretty happy with it.
Then we obviously raised some money beforehand, and we just couldn't find a way to make it a venture scale business. It's effectively a private product and so it would've been an interesting lifestyle business, but the scale of mobile at the time and our ability to monetize and people will miss pay and our ability to distribution, all didn't work.
And so eventually we pivoted the company to a remake of Friends for Sale. But this time for Twitter and on your phone, that was the bullet we had in our back pocket.
Turner Novak:
This is a separate app, like a new mobile app?
Siqi Chen:
It was completed from brand and it was called Stolen. And that was a bullet we had in our back pocket for when, if this main thing that we like doing isn't going to work, we know this thing is definitely going to work.
Turner Novak:
You got the playbook down already too. And from what I saw, 90% day one retention, which for people who don't know what that means, I mean that's absolutely insane. I don't think I've ever seen a product have 90% day one retention. That basically means you download it, you leave, you come back again the next day.
Siqi Chen:
It was pretty wild. People were spending hours on it every day.
Turner Novak:
Really. Whoa. What was it? What did you do again?
Siqi Chen:
So basically it's the same concept as Friends for Sale we built for Facebook. You get to be the one and only owner of @JustinBieber, for example.
Turner Novak:
The tag or you own the account?
Siqi Chen:
A card, basically there's from a trading card, they're like, you're stealing people's cards. There's only one Justin Bieber card that's attached to the Twitter handle in existence on the app. And people can at any point steal it away from you if you're willing to pay a premium over what you paid.
So it was highly dynamic, and people would just be viciously trading Justin Bieber back and forth until he's worth millions of dollars.
Turner Novak:
And people actually are paying money in some capacity?
Siqi Chen:
We sold a virtual currency and people spent... Jack Dorsey and Adam Bain spent thousands of dollars personally on it.
Turner Novak:
And I think Kim Kardashian was playing it too from what I saw.
Siqi Chen:
So it was invite only, that was the other thing. It was invite only, and people came in, like a few invites and within two weeks, despite it being in invite only, you needed an invite code to get in, it was trending worldwide on Twitter.
And what happened is, and I still have the message on Facebook, Kim Kardashian's cousin asked Drew Houseton to ask me on Facebook for an invite code. And I said, "Here you go." But this happened.
Turner Novak:
I mean, it was similarly just super viral. I think I saw a million daily active users or something was within a couple weeks.
Siqi Chen:
Well, it was only live for two, but yes.
Turner Novak:
And then what happened?
Siqi Chen:
And then Zoe Quinn, who was the center of Gamergate at the time, was offended that she got stolen, and wrote something to Katherine Clark, a congresswoman from Massachusetts. She tweeted at her and Katherine Clark wrote an open letter to Tim Cook and Jack Dorsey saying that we open up a potential new avenue of harassment on Twitter, because you can give people a nickname.
And nothing bad ever happened on Stolen, to be clear, the argument was that it could have happened. And so Jack Dorsey really liked the product. He played it a lot. He was like, nah. And Tim Cook never played it, didn't care for it, and pulls it from the app store, and that was that. And we went from every VC reaching out to no VC responding to an email, and we ran out of money, and we were acquihired six months later.
Turner Novak:
Wow, man, that sucks.
Siqi Chen:
It's character building. You're grinding for three plus years, like this is it. We made it, and it lasted for two weeks and then we did everything we could to try to fix it and nothing worked.
Turner Novak:
So you acquired by Postmates and then you were at Postmates for a couple of years, if I'm remembering right, doing growth stuff?
Siqi Chen:
Yeah, I was there for about a year and a half, and we started a proper growth team. I think when I joined, we had a growth marketing team that sent marketing campaigns and there was no AB testing, there was no experimentation engine.
So I remember the best story I have about Postmates is when I joined AB testing was considered a bad word. It was a very brand forward sort of company.
Turner Novak:
So no optimization, data, not important?
Siqi Chen:
Which is great for me. I was like, that means there's so many coins under the cushions here for me to just pick up. But to get there, we had to instrument and have good data for the conversion rate and the flow, and we need to have experimentation capabilities and we didn't have any of it.
So when I joined in, I was a director of product and I said, "We should start a growth team." They said, "Cool, here's two engineers." First three months, we basically put it instrumentation and server framework. And for three months I would have to go to CEO and the CEO would ask me, "Where's the growth?" I said, "Nothing's happening. We're working on things."
And I was pretty sure I was going to get fired. But once we had it in place, the three months after we're seeing double digit conversion rate improvements every week. And it got to a point where the growth team went from two people to one point comes to 70 people by the time I left. And so that was really fun. But that was an exercise in working under very different cultural circumstances and tried to make something work.
Turner Novak:
Interesting. I want to give credit mutual friend Kevin Lee said, we need to talk about that. Just implementing growth and just doubling conversion rates.
Siqi Chen:
It was a good time. It was fun.
Turner Novak:
It's almost like a parallel with runway. No metrics, not optimizing conversion. You got a little bit of an opportunity to maybe actually optimize a little bit at some point.
Siqi Chen:
The funny thing is experimentation went from a bad word to basically everything that they did by the time I left was wrapped in experimentation. And the way we sneakily made that happen is I would be in a meeting, and we would talk about what we did. And I said, "Hey, we did this and this had a provable 25% lifts in symmetric."
And they're like, "How do you know that?" "When we roll something out, we try to measure what it did, and so we have this thing called a rollout framework and you're welcome to use it." And they're like, "A rollout framework that sounds great. We're going to use it when we roll something out."
Turner Novak:
So, it's usually like a foreign concept.
Siqi Chen:
And it was basically everyone decided AB testing might be a good idea.
Turner Novak:
And then now it's you can't not AB test. You're crazy if you're not measuring everything.
Siqi Chen:
Well, everyone wanted to improve the impact of what they did.
Turner Novak:
And then after Postmates, super cool company Sandbox VR, I think you were not a founder, but you were the CEO, I think, if I'm remembering right.
Siqi Chen:
So I wasn't a founder, I was a very small investor in this company, and I thought it was a terrible idea.
Turner Novak:
What was the idea?
Siqi Chen:
The idea as was pitched to me was, "Have you heard about escape rooms?" I said, "Yeah, what if you had escape rooms but in VR?" I said, "I've never done an escape room. I don't know, but that sounds like a really dumb idea." And then the founder said, "Escape rooms are like a $10 billion business."
It's like what? And I looked it up and it was like somewhere around there. I'm like, what? The thing about escape rooms is you can't replay it. Imagine you could just swap out the digital assets and play escape rooms. I said, okay. So what happened is this company was founded in Hong Kong, based in Hong Kong, and we were fundraising from Postmates, do an Asia tour with Bastion and Kristen, who was the CFO and CEO of Postmates.
So, I was with them and we stopped by Hong Kong, and I made a stop by their first store, which apparently went viral after they release this video. I tried it and my mind was just... My mind was exploded. It was the coolest thing I've done since I used the iPhone for the first time, is this Holodeck from Star Trek, but in version 0.1 form.
So if you haven't done a sandbox, basically it's a VR experience, but you're on a realtime motion capture stage with multiple people. So, you have this full body and you can high five people, you can shake their hand, you can jump up and down, and it's like you're there because you're not just a floating head.
That was incredible and they have these special effects and that moment I was like, I got to work here. This is amazing. I think I can help a lot. So I joined as chief product officer, but what happened is I had a network here, obviously, and we rebranded a company and I did most of the new deck and help with the fundraising.
When a series B came around, the CEO and founder of the board was like, "Hey, you're doing basically a lot of the job as CEO anyway. Maybe we should just swap the titles," and said okay. So that happened about six months before COVID, and my primary drill said get a series B done. And obviously that didn't happen.
Turner Novak:
I think you launched the week of COVID or we're going to or something. It was just the worst possible timing.
Siqi Chen:
We launched in a dozen stores, it's like the week of COVID and with intention of having a strong raise or series B, our stores were printing money. And so we went from 400 some employees down back to around under 20 over course of a couple of months. I laid myself off and most of the executive team and the founder hibernated it.
Now, it's incredible. It's on a pretty clear IPO track. It's got about a thousand employees now and it's roaring back. By the time, man, it was the entire revenue stream went from couple tens of millions to zero. It's character building.
Turner Novak:
And I guess just so people understand the context, you guys were like, I don't know, building, renting, retrofitting pretty big spaces. They were basically big green rooms.
Siqi Chen:
Exactly. Imagine you go to a mall and it's basically you're putting four motion captor stages or two to four inside a mall space and you walk in. That was part of the appeal for me joining Sandbox. I joined Postmates because I thought it was interesting to be able to move atoms, having only done digital products.
One of the reasons I joined Sandbox is I figured with my background, very few people would have the experience of operating, constructing retail businesses. I just thought that was a fun, eclectic thing to do. I don't know what's going to do with it. But it was very much a retail business.
Turner Novak:
And it's probably interesting timing with, if I'm remembering that was the peak of the death of malls. So I'm sure the real estate was relatively affordable-ish or at least there's no demand for these emptying dying malls and you're like, "Hey, we're going to build a destination experience in the mall and it's in person."
I mean, I'm sure it made the economics kind of interesting just in the sense of the real estate. It was probably pretty affordable, relatively.
Siqi Chen:
Yeah, actually, I learned a lot. It was really eyeopening about real estate. So there's two things that are really surprising. So one thing is the real estate is affordable if you understand the incentives of the owner, which is that we're actually an attraction.
We attract real estate. People pre-book, we don't rely on foot traffic. And so that ends up being one of the reasons why real estate is a lot more affordable. The second thing is we basically for most of our locations, we pay next to no rent and landlords pay to construct the stores. So, it's a surprisingly capital efficient business once you have a little bit of traction.
Turner Novak:
Just because you bring so much traffic to the mall.
Siqi Chen:
Correct.
Turner Novak:
Wow, that's wild. That's unexpected. I mean I guess that's why you had so many people interested in the series B too.
Siqi Chen:
Once you look in our hood, it's a surprising lucrative business.
Turner Novak:
Interesting. So I had one actual last question on Sandbox from Nev Drawer at Shrug, I think he was a Sandbox investor, also a Runway investor if I'm remembering. So he said, "I got to ask you, there was a time you played Sandbox VR with Kanye."
Siqi Chen:
This is after Andreessen invests in SandboxVR, and Ben Horowitz is like best friends with Kanye. And Kanye was in town and we had this one story at Hillsdale and Ben emailed and said, "Hey, you want to play with Kanye? Kanye wants try it." I said, "Yeah, let's do it."
Turner Novak:
So this was five years ago? This was older Kanye?
Siqi Chen:
It was 2019 around there.
Turner Novak:
Hillsdale is LA, right?
Siqi Chen:
Yeah, that's right. So if you're familiar with the Hillsdale Mall, this is one restaurant, I think Paul Martinis, I think it's called. And they have this patio and we're supposed to meet at 10. And I think I got there a couple of minutes late, and Ben and Kanye were in the patio hanging out, having a drink in the morning on Saturday.
I walk up to the reception of the restaurant and I said the coolest words that ever came out of my mouth up to that point. And to this day I said, "I'm with Kanye." And the reception woman was like "Right away, sir." They had roped off this area in the patio and Kanye was just on his phone being Kanye.
Ben was eating some salad and I was just looking at him waiting for finish his call. He wasn't especially talkative. And we met again, and I have a different Kanye story, but this was the first time. And so a few minutes later we leave this restaurant and we need to walk towards a store. So we need to go outside, go through the mall and enter our small store.
The entourage was a dozen people between his people and his bodyguards. And I just remember walking through the mall next to him and these kids would walk by and just do like, this is a random mall, it's not even popular. And there's Kanye West. And they're like, "Oh." And so they're getting out of their phones.
By the time we walked up 150-ish, 200 meter distance to the mall, there was already a crowd of kids following him. And a few minutes later, all their kids are in Snapchat telling their friends and kids were driving in and the people were peering into the window front of the store and Kanye played a thing.
The other thing I remember about Kanye is that he's one of the two people who has played Deadwood Mansion, which is our first title. It's a very scary zombie experience. I think you've done it if you've done Sandbox. He was completely unreactive to it, just undisturbed. Everyone plays it, they're freaking out, they're running, they're like, "Ah." And he's just looking, unreactive to it. And I was like, wow, this guy is wired different.
Turner Novak:
Maybe he's just played a lot of Call of Duty.
Siqi Chen:
I was like, does he even like it? And he took it off. He's like, "This is great." And we had a whole conversation about investing and they came back down in LA and we had another meeting where he tried some new stuff.
Turner Novak:
Interesting. He was probably just dissecting the product, the art. He was probably just so focused on observing.
Siqi Chen:
So, the only other person had that had a reaction was Michael Ovitz. And Michael Ovitz is pretty old and we thought he would get a heart attack. We were like, scared for his health, and he was in there and he was also completely not reactive.
He was looking at a thing, and when he was done, he took off the headset and he went through, "Here's the story arc, here's where it could be improved, here's the setup of the scene." And I was just blown away. These two people are wired so different.
Turner Novak:
They were probably just completely dialed in to just the experience that they were so immersed in it that they were studying it and didn't react. Because they were just so focused on observing it and just living it, soaking it in. So then, I mean, it's a really interesting transition then into Runway.
You talked about, you had to figure out what happens when COVID hit. That was like when you came up with the idea for Runway. Can you just talk through that a little bit?
Siqi Chen:
When COVID hit, one of the things we immediately had to do is Andreessen was triaging all of their portfolio companies for COVID impact. So, they were an investor in Lime Bike, us, a bunch of other companies too. And everyone was going back to their insider investors trying to get into Fusion because everyone was distressed and no one was about to invest in some other company.
So we had to scenario plan for what we're going to do, how we're going to survive through COVID if it lasts for three months, six months, 12 months, two years. Remember in June or it was it in April, Elon tweeted that this is all going to be over by June.
Turner Novak:
I remember that, yes, he got a lot of heat.
Siqi Chen:
And Elon was like, "I'm really smart." Turns out not so much on that topic.
Turner Novak:
Classic Elon, he is confident in what he says.
Siqi Chen:
And Andreessen on the other hand, I remember, if you remember in January they put up these signs on a door. It got made fun of for saying we don't shake hands anymore.
Turner Novak:
I remember that, yeah. That was early too. That was February or something.
Siqi Chen:
Late January, I was there. They were really early, and they got made fun of for that. I remember when we were doing the scenario planning and working with the firm, they said, "We believe this is going to be two years." And nobody believed that at the time, and they were spot on.
So it's one of those things I think about when I think about what makes different firms so great, things like that. Naval has a saying, "Truth is that which has predicted power," and they predicted it correctly. And one of the few people in the world to have done that.
We had to create these scenarios for how long COVID is going to last and what the financial plan is. I just remember doing that on Google Sheets. I remember working with our CFO emailing these Google Sheets back and forth and there were errors. It was slow, it was confusing, it was really frustrating.
Long story short, we had to lay off pretty much everyone, hibernate it and do what we can to preserve capital and get some capital to survive, and we basically had to get it down to no one. So that's what we did. I laid myself off along with the rest of the executive team.
About a half an hour later, I was having a conversation with our new CFO who came from Netflix, and I asked him, "Hey, this scenario planning for COVID, we could have used something different from that? Notion, Figma are all products that exist. There was something that is already there I just haven't heard of. Someone must have solved this problem."
And he said, "No, this is the best we've got is pretty much Sheets and it's not great." Or, "There's really expensive software like Anaplan, but everyone has to using it too." And I thought this seems like something interesting to build. I mean that's how Runway got started.
Turner Novak:
Interesting. So why didn't it exist before 2020 when you started it? The only options were Sheets and maybe Excel?
Siqi Chen:
Sheets, Excel, and that's it, that’s it really. And Excel and Sheets are fantastic pieces of software. They can do anything. It's like the gold standard of this. I think the more we get into the development runway, the more we realize that there is deep structural reasons for why this doesn't exist. The existence of Notion and Figma and Airtable are fairly recent phenomena, right? So, that's one.
And two is, I do have this idea that there is this vast financial conspiracy where these tools are so awful, and part of the conspiracy is that these products are so terrible that you pay somebody who is willing to use it. And I am getting paid because I am the person who knows and is willing to use this piece of shit. And so, those are really powerful incentives. So, I think that's one of the reasons why that is the case, along with the advancement of both what is possible in technology and design.
This other factor too, though, is the changing strength of transparency in organizations. So, we've seen this story before. So, if you think about what the world was like, for example, before Amplitude, to get data for retention and growth, you had to go through the data team, right? And not everyone had access to it. And so, it was considered somewhat sensitive in a lot of organizations, like what user counts or retention was, right?
And the existence of Amplitude and tools like that democratized access to people who aren't data analysts, who aren't part of the data team. You're a product person, business person, engineer, you can get to the answers yourself. And now, we kind of take it for granted that you can work at a company, you kind of just understand the metrics of the product. That wasn't always the case, because the tool wasn't there.
And Figma is actually a similar story. People think of Figma and they think it's a really great design tool in a browser. But we forget that before Figma, it was considered impolite to even look over the shoulder of your designer when you're designing, right? When you have a designer and a person walks over and look over your shoulder, it's kind of a faux pas. But now, everyone is in Figma together, and they're collaborating, it's made design more strategic in the same way it's made data more strategic, informed by Amplitude. There is no such equivalent for the business itself.
And the thing about finance that I think is really misunderstood and the opportunity, is that finance is not about bean counting or the spreadsheet that no one really understands. It is actually about the entire thing. It is about the entire business. What this spreadsheet is a software simulation of a business. It's coded in Excel language, it simulates the future, and you use it so that you can understand the impact of the decisions that you make today in the future. Which means that for it to be accurate, it has to take into account accurately how every department works, how the product works, how marketing works, how sales works, and connect to all the data sources, and people need to have a stake in it. And none of that is true today, in these models.
And so, because the expectations for transparency and the competitive advantage that companies have when people collaborate and the insights from people on the ground on what should be done, and the advantage of having everyone understand the context of where you're going, you need something that's more accessible, and that has never existed. But that's a relatively recent shift that I think has crossed a critical point over the past few years. And now, everyone understands and everyone markets like, "Oh, we need to have something accessible and we need to collaborate." But to do that, you need to have a very different product.
Turner Novak:
Yeah, it's a really interesting point where you say, design, super important, impacts the product, marketing, growth, impacts how big the company is, etc. Engineering, you're building the product. But all these things tie into the number. When you say, "We need to grow faster," it's because you need the spreadsheet to go up. When someone says, "We need to design a new product, we need new product lines," it's because that flows into the spreadsheet that then values the business at the end of the day. There's a very qualitative piece of we're going to design something new, but then there's a quantitative of, it has some kind of value at the end of the day. So, it's an interesting point.
And then, when you were talking, it kind of made me think of, okay, when was the spreadsheet invented? When was Excel invented? It was invented in, I actually don't know, the 70s, 80s?
Siqi Chen:
'78? Multiplan, I think. There was a killer app for a first personal computer, so right around then.
Turner Novak:
Really? And this was pre-internet. So, the spreadsheet was designed before you could collaborate. Computers were not connected to each other.
Siqi Chen:
That's right. It was like the very first killer app for a personal computer.
Turner Novak:
Yeah. And then, when I think of what was Sheets, that was the first kind of internet native, I mean there's probably other ones, but it's the biggest internet native spreadsheet. But it had to be compatible with Excel and with offline spreadsheets, really, at the end of the day. So, it's the same thing. It really hasn't changed. And what are we at, 46 years? Maybe, I can't remember the number you said. But anyways, been almost 50 years of the same thing basically.
Siqi Chen:
Yeah. And I think that's a positive, in many ways, right? It's like an accessible interface. People can quickly pick it up and understand the basics of it, and they even have to change it. They've added a lot on top of it. And so, I think that's great. The issue is that, because it's so powerful, because this product can do anything, it can do any specific thing super well, and that's the opportunity. So, one really simple example is, you think about what is required for people to collaborate on a model for the company, right, and why it doesn't happen. And the reasons are actually really simple. So, let's say I have a company model and I want to be able to share the context of this model with my head of sales or a head of product. I actually can't do it. And the reason why I can't just share the model is because this model has everyone's salary in it.
And on a spreadsheet, you can't just say, "Show everything but hide this one column." That's not a feature that exists in any Excel or Sheet product. And so, the observation here is that, it can do anything, but it can do any specific thing well, and there are many use cases that are very specific things. Business planning turns out to be a pretty specific thing. Airtable unbundled Excel for a pretty specific use case, which is like the database use case. And you can do a lot more that you can't do in Excel as easily. And so, that's sort of our observation is that, if you think about what the use case is, the abstractions required are different to make it really accessible and useful. And that informs our product strategy at Runway in a pretty fundamental way.
Turner Novak:
Yeah. You have a pretty fun LinkedIn description of what you're doing at Runway. It's disrupting the $80 trillion business industry. I mean, that's just global GDP basically. You're just changing the way businesses run, essentially.
Siqi Chen:
That is the ambition here. The mission of Runway is to make business accessible and understandable to everyone. And this is a really personal and important mission, I think. It's personal to me because I've always felt insecure about being any kind of business person. I have no business background, right? I don't know what this finance stuff was, even when I was CEO, and I wanted to understand it better.
But I think, on a broader level, this is not a product design for any one industry. Really think about it in terms of people and actually adding meaning to their work. So, the two things that make the mission important is, from an enterprise perspective, we believe that if you have more people in your company who understand how the business works and can contribute to the running of the business and contribute to insights, you will make better decisions.
And if you have a tool that people understand how the business works and understand the plans, it's easier to align everyone to go in the same direction, right? So, one of the things I talk about when we talk to candidates for example is, have you ever experienced someone high up making what seems like a really dumb decision, right? The pointy hair boss doing something dumb.
And everyone has some version of this, and my observation here is that half the time, they are really stupid. There's a lot of stupid leaders. But half the time, it's probably because they know something that you don't. And had you known it, would've made complete sense, right?
Turner Novak:
You're saying, as a lower level judging the dumbness of a higher ups decision?
Siqi Chen:
Right. They have some context that you don't have, which is why the context over control value at Netflix is so powerful. If people know the same things and they have good judgment, chances are people will at least understand and find reasonable of where you're going, and maybe move in the same direction. And so, understanding how the business works and what the trade-offs are helps you do that too. It helps you align on the right decisions and make better decisions. But then, from a personal just working experience standpoint, whether you're an engineer or PM, we know what makes work fulfilling and fun is autonomy, mastery and purpose.
And we kind of pretend that you can have purpose without needing to understand how the business works, or the context around it. And the image I have of this is basically, you are chickens in a farmhouse and you're getting fed every day in the form of massages and high salaries. And one day, farmer John takes a bunch of chicken out and cuts their heads off because you did not understand that you live on farm. And that's what it is when you don't understand what the business is and how it's doing. You get laid off one day and you're like, "What happened? I thought everyone's getting massages and things are great." But that shouldn't be how it works.
Turner Novak:
Yeah. So, you're really just opening up more visibility into the business, helping people understand the company that they're a part of, that they're contributing to at the end of the day. So, you had two things that really stood out to me in terms of how you build it. There's automation that's built in, and then you say the formulas are 50 times easier than Excel. What do those two things mean? Because that sounds pretty big if you can pull them off.
Siqi Chen:
Yeah. Well, think about how you write a formula in Excel, right? Let's say you have a Runway in some naive way, you divide bank account by burn rate. But, that's fine, you can do that, right? But what is the formula? The formula is A1 divided by B1, or A2 divided by B2.
Turner Novak:
Yeah, those are hard to audit if you're going in.
Siqi Chen:
What does that mean, right? And this is where the abstractions make a difference and the primitives make a difference. So ,the core primitive in Excel is Excel. That's why it's called Excel, right? And so, you put a formula in one cell, you paste it across, and it does some spark transformations, and the formula reference cells, right?
So, think about what happens though. The way Runway works, for example is, our core extraction is, one of the core extractions is a row. And with a row, you can give it a name. So, you have a row called cache, and you can just name it. You can create another row called burn, and you can just name it and you can create formula there. And the formula for Runway then is cash to buy back. You have this one formula, and that one formula goes across the entire row, right? And so, you could have 100 rows, but it's still one formula. And because it's easy to read, because it's semantic, it's easier to debug, it's easy to understand.
Turner Novak:
So, it's basically just giving the people the ability to name the rows and the columns, makes it so much simpler to use and understand.
Siqi Chen:
Yeah, and that's one of the things, right? I think we have other ideas here. So, we think about the concept of abstractions a lot. So, abstractions, the right abstractions make things easier to understand and enables new workflows.
So, for example, probably our most powerful abstraction is this idea of plans. So, imagine you're doing a growth model, right, and you have a row called conversion rate. And maybe you know that in your product roadmap, your conversion rate from registration, install registration goes from 50% to 55%, 60%, right? On the spreadsheet, what does it look like? You have a real code version and you have one cell that's 50, another cell that's 55, another cell 60, right? But how do you as a product person actually think about what you're doing? You think about it in terms of product roadmap, and you think about the features that you're going to build to make that happen.
They have a certain time, they have a name, they have an owner, that context is completely disconnected from those numbers, right? So, we have a concept called plans, and basically it lets you say, "Hey, the reason why this change from 55 to 60% is because we're going to work on this future at this time by this person, and here's what we expect to happen." And now that you have these plans, you can just drag the plan on a timeline. You could say, "Okay, this is what happens when it happens in Q3 instead of Q2," and the numbers would just work and stay in sync with the spreadsheet.
Or you can just drag the impact up in there. It's like, "Well, what happens if the conversion rate of this future goes from 55 to 70%?" And when you make those changes, we have another extraction called a pull request, which is for by GitHub, right? Right now, if you make a change, you can make an entire copy of it. But with pull requests you could say, "Here's a change we're proposing to the plan." You can have a conversation about it. And so, all of these things are things that you can't do if you just have a cell and you don't have new abstractions that make it easier to understand.
Turner Novak:
So then, I know you have a bunch of integration. So, are you integrating with, I'm just trying to come up with an example. You're integrating with Asana for some of the product roadmap stuff, is that kind of how that works? So, if in Asana I'm changing my plan or whatever I'm using for my road mapping, I change when this thing gets released, it flows into Runway and it impacts what I'm seeing from the "spreadsheet finance perspective"?
Siqi Chen:
Yeah. You could do that. People actually just try to connect our plans directly. Yeah. People integrate everything, right? They integrate Jira, they integrate the CRM. And so, everything that you use, really, to run a business eventually falls down to the bottom load somewhere, and is a potential area for a Runway to integrate. And that's not what happens today in software. Because finance is siloed, you basically integrate with your general ledger, your accounting system, and that's about it.
Turner Novak:
I majored in accounting, I fricking hate accounting,
Siqi Chen:
Yeah. But it should integrate with your analytics system, right? A change of retention and growth, it should integrate with your accounting system. And you should see the changes in real time, and that's not how it works.
Turner Novak:
Let's say, you're trying to sell Runway. You've got to sell it to a CFO, a finance team. They're used to Excel, maybe they use Sheets. What's the pitch usually? How does that go, convincing them to use this whole new type of paradigm of finance software?
Siqi Chen:
What's funny is we actually try to not have to do any convincing, and this is somewhat of a surprise to us, is that it's not a vitamin. It is very much a painkiller. And when you are on Excel at a certain scale of company, you experience very serious pain. And the pain comes in roughly three forms. The first one is, the raw amount of time it takes to update your model as new data comes in. Because it has to take into account all these different things, what you have to do every month is pull in a bunch of data from your data warehouse, from your general ledger, from your CRM, and that just sucks to do.
Turner Novak:
You probably have a couple tabs that are like data dump and data cleaning, cleansing tabs with formulas and macros probably.
Siqi Chen:
Exactly. And that only gets worse over time. You never have less data, right? You only have more data sources over time. So, that's the first thing that happens. The second thing that happens is a company gets to a size where you need to collaborate between different departments, understand what people's plans and goals are. And today, the way you finance people actually do it is they create these empty spreadsheet templates. And as a department in the built space, I got this, right? I get one of these and it's like, "Can you tell me how much you're going to spend and how much we're going to make?" And it's just pure paperwork, because I don't know what it's for, I don't know where it's going, I don't know why you use it.
And so, collaboration becomes a pain. And so, something that helps people understand, okay, what are you going to do and how does it affect the business, and employee collaborating in a reasonable way becomes increasingly painful and needed. And the third thing is, the business itself just becomes way too complex. You have multiple entities, you cross multiple countries, you have multiple customer segments, you have millions of SKUs in some cases. At some point, the sheer complexity of it makes it so that Excel can't handle it anymore. And so, all of these forces are pushing for someone to go to something different.
The problem is, your choices in the market today, it's basically keep on dealing with the pain, or it goes to this other thing that actually doesn't solve your problem, that everyone uses, that creates a whole 'nother kind of pain, which is the established Anaplan's enterprise products of the world. And the funny story is, I talked to a Fortune 100 CEO one time and they used Anaplan, and we talked to a lot of users of Anaplan and everyone said it's the worst thing because it's really difficult to model with. And they said it was great, like, "It did exactly what I was hoping to do, and I'll introduce a person who set it up for us."
Maybe they got it right. We talked to that person, it was like, no, it was awful. You have no idea. That dude had no idea how much time and pain it was to set this up and have other responsibilities. What is he talking about? I was like, okay, that squares.
Turner Novak:
And is it usually, there's just someone on the team that just eats the shit sandwich, just they got to own this thing and that's their job?
Siqi Chen:
Yeah. I think that it's part of the underrated, going back to your question about why it's bad. When you're a certain size, you can just hire people to use the software, and that's the incentive and it makes sense. But, because what that hides is the opportunity cost of having something that people truly want to use and can use. The insights and the efficiency gains you gain from that, you can't really measure but their present.
Turner Novak:
Everything we've just talked about, I know that it took a while to kind of get through this. It wasn't like this just clicked right away. So, I guess, going back, Runway, fundraising, you raised that first round kind of after Covid. How do you fundraise? Take us inside. You've obviously done it a bunch of times, I know you're pretty experienced at it. How do you approach fundraising and what are kind of best practices, and what do you do that's maybe a little bit different?
Siqi Chen:
There's two parts of fundraising. One is process, the other part is the pitch, and the pitch is part of the long story.
In terms of process, you treat it as any other sales process, right? You want to have a pipeline, you want to have as many meetings as possible in a concentrated period of time. So, the thing that's different about fundraising and sales is that there's a business transaction that has a market clearing price. And so, if you think about, in those terms, then you want to minimize the supply and maximize your demand to maximize the price, the clearing price. And so, the thing about minimizing supply is, imagine you're linking two companies. One company says, "We're raising 5 on 10, we're halfway there." The other company says, "We're raising $2.5 million and we're already subscribed on 10." Exact same situation, another one just sounds better. So, limiting your supply is one way of increasing effective demand.
The other one on the demand side is, imagine you have access to 12 investors in world, if that's your entire network. There's a huge difference in having those meetings over the course of three days versus a year, right? Concentrating your finite demand in time is a very useful technique in maximizing your demand, and you need to create heat in order to close a round. One of the truths about fundraising is that rounds will either be oversubscribed or they were never close. And that is totally true. And so, whatever you can do to minimize your supply and maximize demand is very helpful. So, that's on the process side.
I have this process where I try to tranche my meetings, in three tranches. So, there is, the first tranche is people who are going to invest no matter what. And these are fed leads, I get feedback on it, I get some momentum into the round.
Turner Novak:
Those are the first?
Siqi Chen:
Yeah, that's the first tranche, right? And by the time I get to the second tranche, which is people I don't know, but also people I don't particularly care about, I have a bit more feedback into the pitch. And there's some momentum, it's a round already.
Turner Novak:
So, you might get some conversions in that wave?
Siqi Chen:
Right. And the third trache is people you actually really want. At that point, you have all the momentum. Maybe you have term sheets and your pitch is well practiced. The way in which you polish your pitch, during that process, one of the most common founder questions or complaints is that, "I thought we had a great pitch and they didn't tell me why. They just kind of ghosted, and I don't know what the issue was." Right? And the hack is, they always do tell you why they're not investing if you know how to listen for it. And the reason why they're not investing is the first question that they ask. Whatever that question is, that's what they were thinking about while you were talking, and that's the abduction. And so, the algorithm is, how do you update your pitch next time so that that question doesn't even come up in the first place.
So the question is, how do you think about competition? They're probably thinking this entire time, "Incredibly competitive space. I don't know if I want to invest." So, if you add a slide, which I literally have done this before after an investor saying, "Very competitive space, here's why it still makes sense." Question won't come up. And this has happened. So, that's the first part of process.
The second part is the story. And the two frameworks here is, one, how do people actually make decisions? It's not based on facts, it's based on motion. So, the question, for example, it ends up a really good filter. For example, one of the frequent questions you have is, "What should I put into a deck? What should I put in an appendix?" And the thing you put into a deck is whatever gets people to feel something, and the thing that you want them to feel is eventually leads to greed.
Turner Novak:
I feel like I'm going to make money on this.
Siqi Chen:
Yeah. Yeah. The particular emotional journey I like to hate people on, by the way, is go from amusement to curiosity to surprise to awe to greed. And so, I try to create a story that will generate those emotions. So, amusement will tell you that, hey, if you decide to not invest and we're nothing here, at least you won't be bored. And so, you can pay attention. Curiosity then suggests that you might learn something. And surprise is like, oh, that paid off. I did learn something interesting. The odd then goes, this interesting thing could be extremely valuable. In fact, could be one of the most valuable companies in the world.
Turner Novak:
Yeah, how do I get in?
Siqi Chen:
Yeah, how do I get in? Exactly.
The other framework about a story is, if you break down an investment rationally, you can break it down in terms of, one, likelihood of success, and two is, size of outcome being successful.
The thing that dominates the decision, I think is the second term, the size of outcome if successful, and not enough stories, pitches focus on that because that's the nature of venture math, right? You want to be able to believe that this company could be a Stripe or an Airbnb or whatever one day.
And so much of, the issue with the founder psychology is that because founders have to be in the ground day to day, we're so focused on now to execute, that it's hard for us to transition and context shift to, what are you actually selling? You're not selling today. What people are buying is tomorrow, and people don't spend enough time talking about that.
Turner Novak:
Is there a similar thinking then on the recruiting side of the story that you tell? What are the differences between fundraising storytelling, and then maybe recruiting and customer storytelling? How do you think about making that transition?
Siqi Chen:
Yeah. This is actually a relatively new insight, but I think is something that I think is super impactful in how I think about operating and for Runway. So, the role of the story, right, and we have a fundraising story, we have a recruiting story, we have a marketing story, we have an internal story, right?
And the default is to think about these as independent, as things that you do outside of the product, right? You build product, you solve these customer problems, and we need to figure out what our marketing story is. And then, we're recruiting, you need to figure out what the recruiting story is.
I think the thing that I've realized talking to a bunch of advisors and working on our story is that it's so powerful to think about the story as just one thing and everything flows from that story. And that story is like, why do you exist as a company? What is the purpose and mission of this? This comes out to mission and vision. he thing about that mission, vision, culture and values, the thing about that is people also thinking about those things as these independent things that you do, you check off a box.
But all of that is just one thing. The story of Runway is that we believe people should understand the nature of their work. This should be accessible to people. That has profound implications on what it feels like to be working and how it impacts the success of businesses. And that's why we exist. And to do this well, we have to connect the art of product development, but also the nature of being human and having emotions and the squishiness of it, with the hard-nosed numbers and machine that is a business. We have to connect the two.
And so that informs the story that we tell to investors, to candidates, what product we build, how we communicate, what we do to the market. It's all one thing. And so I think Lulu Cheng had a really great article on Pirate Wires about Anduril and how they do comms. And I think about it in similar ways.
There was this picture of the story should be the center. It starts with the founder and then you expand that circle to the leadership team, then to the wider team, then to your customers and to the world. If you can do that and that's how your story is, it's one story and it impacts everything. And it's actually the central access by which your company operates. I didn't think about story in that vein in that way until fairly recently, but it's been so powerful.
Turner Novak:
It's almost like it can mean different things to different people. The story that you're telling means something to you, to an employee, to an investor, to a customer. A customer might be like, "Cool, better software." For you, it's like changing the world. To an employee it's like, "This is a fun place to work. I really enjoy what I'm doing. Be like I'm getting value." To an investor it's like, "We're going to make money." And that story that you're telling, it's the same story, but it means something different to different people depending on where they're at in that circle that's expanding out.
Siqi Chen:
You want it at the deepest level to be the same story because if it's not the same story, it won't be cohesive. The culture of the team, the people in it that you hire, the marketing that you do with the product and your brand, it won't mesh. It won't make sense. But if it's the same story with different facets of expression, there's something deeply cohesive and powerful and weighty that we've seen really powerful results here in Runway.
Turner Novak:
You have this really interesting framework around positioning to yourself, your company, your business. Can you talk through that?
Siqi Chen:
So this was learned from a person named Wolfgang Hammer who was the producer of House of Cards and he ran Cinemax and CBS. We've been working on positioning a story for Runway. There's this framework that he had about good stories that really stuck with me.
This idea that a great story has three layers. The first layer is action, like the events of the story. That's the most surface level of layers.
The next layer deep is emotion, how people feel during the story and people are picking that up. But what makes the story deep and interesting and profound is the deepest layer.
And the deepest layer is unique, and this really blew my mind. And this idea that what makes an interesting story is a conflict of worldviews between a dominant default worldview and a underdog worldview.
And so the global worldview is something that is mainstream, everyone believes and critically there's really good reasons for why people believe that and why it's a dominant one. So in our case as an example, finance is supposed to be complicated and Excel is great and all these things. And there's great reasons for why that is true. And the underdog worldview is a different worldview. You can imagine a better world. There are untruths about the reasons for why the dominant worldview is correct. And it's a personal story.
The thing that really struck me about that is that in stories, the dominant worldview is held by the antagonist of the story. And the protagonist story is always the underdog worldview. And so when I think about positioning, that's been such a hell of a framework thinking about what is the enemy worldview that you're fighting against? Because the counter to that and what makes you different and how we position is going to be the underdog worldview. Weaving that into the story is how you get to a really good position or narrative about how you sit in the world and how to get people to feel something and follow you and all these other great things.
Turner Novak:
Yeah, it almost ties back to this concept of just counter positioning. It's like what's the incumbent startup business, whatever you're competing against? And then what do I believe that's different that's actually maybe better? Or maybe it's just different, you figure out if it's better. Not all startup ideas work, but yeah, that's an interesting framing to think of that. I like that.
Siqi Chen:
And one of the critical things is how reasonable and how seductive the reason supporting the default worldview is. It's so important that you understand deeply those reasons, right? It's almost like in 8 Mile you're doing an accusation audit of the global worldview. Like, here's all the reasons why this is actually great and I get it. But there's a lie here and not everyone can see the lie or is brave enough to call it out, but this is what we believe that's different. And people are drawn to that.
Turner Novak:
And it's probably most helpful if it's a very simple solution yours of, it's so believable. And it's so easy to just... It's like something that people overlooked almost where it's like, "Hey, if we just do this one thing differently, all parties benefit, all stakeholders. It's actually better for everyone."
So a little bit of a related question to story and how that impacts the team. Question that came up from this is from Kevin Lee, mutual friend. I think Arjun Saidi also had the same question. Just how have you cracked org design, like organizing the team? Kevin specifically said you figured out how to make remote work work, was his comment. How do you do it? How do you think about it and what's working for you?
Siqi Chen:
I really don't feel like we've done that. That's nice of him to say, but we're trying to figure it out as much as anyone. I mean, the background is we were founded right when COVID started. We had no choice, so we tried to make the best of it. We tried just about every remote work tool there is. And we even tried a hardware a company that has a remote tool where you can have this iPad on the side and click on a face and just start talking to them as if they were next to you. Nothing really worked, I think. Right now we do just normal things. We use Notion and Zoom and Slack and Height. Mostly the way in which we've made it work better is one, we just have more practice at it. Two is we're closer to the product market fit and the ambiguity is lower than it was before.
But probably the most important thing is we just get together regularly in person. So twice a year we have an offsite, we get together. And in between those offsites we do an onsite, which is we just get together in the same place and just do normal work. So we did one three weeks ago. It's my favorite working experience ever. And I guess the way we made it work is there's somewhat of a balance. We do have hubs in New York and SF and people come in usually twice a week and I do as well.
I think the way in which we made it work is I think we've tried to not be religious about any particular thing. Some work is done better alone at home, some work is done better together. And for people who want to work differently, we give you the option. Instead of focusing on particular form and prescriptive work, because reality is whether it's remote work or any other best practice, the opposite of the best practice also works.
What you have to manage to is results. If people aren't effective remotely or however they choose to work, then we manage them out. And that's the thing that really is the thing that makes it work is that we try to hire people who are effective at how they work. And we try to not be religious and prescriptive about how any one person chooses to work.
Turner Novak:
So how do you make a good hire then? It sounds like that's super important for you guys. What's your process?
Siqi Chen:
We've evolved. We started with writing samples and a work trial early on. I think a lot of companies have kept on doing that. We found that it was difficult to scale in the sense that it really selected for people who are really available or didn't have kids or was able to take a lot of time off. And that's not usually everyone. So I'll observe that Lenny did a series on this recently, but the question was not asked about what about people with kids and family who can't take time off, like had to do with a full long work trial or are in between jobs? And I think that's the real reason why we stopped doing it. I think we made better hires than we ever have over past couple of months.
So, to give you a sampling, our last two engineering hires were both VP of engineering at $500 million companies and are just IC engineers here. Our PMM was a VP marketing at Flexport. And so these are fairly recent hires. So I think our ability to hire people has never been better. Our process is we do a take home exercise. If that goes well we do an onsite, that's virtual. There's architecture, there's coding, there's a culture interview with me. And my interview really is just like, "What's your story?" It's actually literally just one question.
Turner Novak:
Really? And they talk?
Siqi Chen:
They just talk, and I'm just listening. What I'm looking for explicitly is we have some values at Runway, give a shit, build trust, create clarity, raise the bar. And that's primarily the way in which we value, weigh people. Does this person create clarity? Is their explanation clear? Does this person give a shit about their work? And also what we do, and why we do it? We want people to actually... The thing about Runway man, I have been on such a high working here, and one of the things that I think about a lot in just two weeks ago at All Hands, one of my engineers expressed that on Sundays, he can't wait for Monday to arrive. He can't wait for Mondays. And every person at Runway feels this way. And it's so gratifying. It's because you know everyone else here gives a shit, and you know everyone here gives a shit about why we're here. And that's why it feels that way. So we look for that.
And not everyone feels that way about what we do and our mission, that's totally fine. Not everyone feels that way about work and what they do and that's fine, but it's not for us. That's probably one of the more important bits. Give a shit is literally the value. Build trust as well. Does the person understand how to build trust? Is self-aware enough, low you enough to do that? And the final one is raise the bar. Does this person push us to be better? Is it better than our average person? Does this improve or multiply us in some way? Have they demonstrated that in the past? And a lot of that is also affected by this idea of high agency. Do I have a internal locus control? Do I have control over my own outcomes? Because there's other really powerful framework that I have about these feedback loops that are kind of invisible, that are so impactful to life outcomes.
So one of the feedback loops is this idea of agency. If you are a person who, for example, when something bad happens believes that it's because of external circumstances, you're probably right. And the next time something bad happens, it's going to be more confirmatory. As this loop happens, you're going to be more deeply embedded in this idea that everything bad that has happened to me is because of external circumstances. Because everything that happens is additional confirmation. But if you believe the opposite, the opposite loop happens too. If you believe something bad happens to you and you have control over it, you can do something different. The next time something happens, it confirms it. The next time something happens, it confirms it. And you build on top of that and you get momentum that way.
And so the people with the positive feedback cycles and they're very, very different. I think that's one of the best traits that's correlated with what we're looking for too.
Turner Novak:
So it's people who are excited about what you're doing, they care, they give a shit, you can trust them, they're very good at what they do. They improve the team, they make other people better, and they just keep getting better over time because they believe that they have control over that they can improve. Yeah.
Siqi Chen:
That's an excellent summary.
Turner Novak:
One of the things that I think a couple of people mentioned on was it wasn't like you cracked this immediately. It took you a while to go from the sandbox VR experience to then where you're at today. The current product is very polished. What was that journey like of going from like, "Okay, we know we have to fix this problem," to, "It works." Or to a product that people want to use?
Siqi Chen:
It was personally torturous because the thesis behind people investing in Runway is that we're going to take a consumer founder, apply it to an enterprise area and maybe something good will come out of that.
Turner Novak:
Okay. It's a good thesis, honestly.
Siqi Chen:
I think that's ended up being true. But I think what it ignored is how much you also need to unlearn because the nature of building enterprise products is highly different. The surface area is much wider, and understanding your customer and their pain points is different than building a consumer product. And so a lot of the pain in the first year and a half was just unlearning and relearning. And from the rest of the team, frankly, I'm like, how do you actually do this? Only after unlearning can I then apply what I am good at and reintegrate my superpowers.
Turner Novak:
What were the things you had to unlearn?
Siqi Chen:
The thing about consumer a lot of the time is you just know what you build something for yourself and you know what you like and you put it out there to see if other people can iterate from there. And thing about enterprise is that people are really different from you. You can't just start by building something that you like because the segment of people who actually have deeper pain that are valuable may be completely different from you. So that's one.
And two is if you just build what you like, the product outcomes of enterprise are actually path dependent. So what I mean by that is if I build something that I like and I'm an early-stage startup founder, say, then it's going to attract more early-stage startup founders and you're going to get traction that way. And you're going to build more stuff for them.
But if you haven't really stepped back and thought about, talked to more people and thought about where is the real pain? And with that pain, how much value is there? And what is the willingness to pay around that? Then you might be led to build something for early-stage startups for which there's high churn, low willingness to pay, and the pain frankly isn't high. They're willing to use it because it's cool. Meanwhile, you have a larger customer who's willing to pay thousands of dollars.
Turner Novak:
I'd assume hundreds of thousands. Yeah, massive contract size, millions even.
Siqi Chen:
Yeah. So the path dependency of it is something that is not, it doesn't matter quite as much in consumer because by definition you're trying to attract just about everyone, right? People with a phone can download the thing. So that's really different. One of the many things.
Turner Novak:
It was a skill, like a superpower that you could build for a consumer product, but you almost built, you just had no experience on building for an enterprise customer to the point that you just had to learn that skill while not losing that consumery mindset. Like the product experience, the thinking from first principles, redesigning, coming up with magic, I guess if that's a fair way to think about it?
Siqi Chen:
And the other part of this too is that for the space that we chose, there is no obvious quick, easy wedge here because people rely on it to run their businesses, the core of their business. When you think about what inspired to make that work, just to get off the ground, just to get your foot in the door for someone to start using it. You basically have to build a good chunk of Excel and then you have to build integrations that will have 700 different things. And then you have to build a good chunk of Notion for reporting and collaboration. And that's just to get your foot in the door. That's not even why should people choose your new thing over the thing that already exists?
And that's hard because how do you even get traction without having... There's the minimum so people can actually use it for real. And how do you keep the team going when it's hard to get traction over, in our case, like many years? That was probably the hardest thing to do, but we still have problems and challenges like every other company today, but it's not on nearly at the same scale as what it was a year or two years ago. Once you get that first paying customer and you have some kind of foothold. But getting to that first paying customer was, yeah, it was not fun.
Turner Novak:
How do you get through that? What do you tell yourself as the founder and your team too? How do you keep people motivated?
Siqi Chen:
The story and the mission helps, this belief that this is not good enough and you could build something better. And that's just somewhat blind optimism. The other story that I tell myself is when you look at the recent generation of truly great productivity enterprise type of products that are consumer grade, they all took three and a half to five years to launch. Figma, Notion, Airtable, Coda, even Anaplan back in the early '90s, that was a five-year journey to launch. And I think there's not enough appreciation for if you want to build software of a certain kind, it actually takes a really long time. In the 18-months timelines, Figma made no money, didn't get to a million in AR until I think it was five years, six years in? It was something to do with five or an eight, I think it was, probably made the first million in AR, and then it exploded.
So I think people forget or don't appreciate how long this stuff takes. And so looking at those benchmarks was helpful. Having Dylan Field as an investor and having a positive team was also quite helpful. What he said, I've thought about a lot, and I've shared with the team multiple times. I asked him a question at an internal podcast, lunch-and-learn type of thing. And the question was taking Figma from where it was, which was nothing, to where it was today, what surprised you the most about that journey?
And he said, "What surprised me the most is how few people made it all the way with us." I was like, that's interesting. But the more I thought about that, the more crazy that answer is because it's a team of smart people working on something cool. And well, what it ended up being is one of the most successful companies of all time. And half the people on that team looked left and looked right and said, "At some point we're not going to make it. I don't believe it anymore." That's just the norm. That's when it works out. And so that's helpful.
And the other bit that is helpful is basically I thought my job and the story I told myself is, "Trust me guys, we're on a hockey stick red curve." But when you plot out the hockey stick red curve, we're just on a really flat part of it right here. That's where we are, but we're going to get there. You just got to believe. And there was a lot of that for really good reason because think about it, all their friends are working at OpenAI or Coinbase or somewhere and they're killing it. And meanwhile, your stinky startup has no one using it.
Turner Novak:
Wow. So it's really just believing in the mission, buying in, being in it for the long term. And knowing the payoff when you actually built, finished the product and get it in people's hands that the payoffs can be worth it.
Siqi Chen:
Yeah, setting the right expectations. Building momentum is something we think about a lot too. But sometimes the activation energy is just really high for some classes of products.
Turner Novak:
Awesome. Well this was a lot of fun. Thanks for taking the time to chat.
Siqi Chen:
Same, I really liked it.
Stream the full episode on Apple, Spotify, or YouTube.
Find transcripts of all other episodes here.