Howie Liu: Building Airtable to Feel Magical

0:15That's Howie Liu, best known as the founder of Airtable, the platform that made a lot of Airtable. poured themselves into creating a product so intuitive and distinct that users didn't ask what's different about this. They just felt the magic instantly and wanted to share it with others. This is Mike Maples Jr. of Floodgate and it's go time with Howie Liu. This is Mike Maples Jr. and welcome to the Pattern Breakers podcast where we explore why some founders radically change the future and how they stand apart. Together we'll learn about the counterintuitive mindsets and actions behind their remarkable success.

1:01Brace yourself for a world where chaos is welcome, naysayers are often a positive signal, and movements galvanize misfits who transform the impossible to the inevitable.

1:24Beneath the surface of the simple and highly intuitive Airtable user experience was a deeper insight. Spreadsheets were originally designed for finance people, yet millions of people now were tweaking them in thousands of different ways to do just about everything else, from project management to content calendars to recipes, even travel itineraries. So Howie Liu and his co-founders spent two years designing a product that made the transition from a spreadsheet to a hybrid database feel obviously valuable in less than 30 seconds. And it worked. In the crowded world, especially now in this AI era, where every new product fights for attention, showing your insight, Your real difference with real product and quickly has become vital for achieving breakthrough success.

2:14If you want to stand alone, it helps if your difference is viscerally felt instantly. Let's talk to him.

2:30Hallelu, welcome to the podcast. Excited to be here. You know, Airtable wasn't your first rodeo. Why don't we start there? Let's start before the beginning. You know, how did you get into startups? And, you know, what did you learn from your early startup experiences leading up to Airtable? I mean, I think I just always really loved computers, the internet, like the idea of like, wow, like there's just so much out there that and anyone could go and participate, right? Like, you know, the fan sites at the time I was into like Pokemon and so on. And like anyone could go and make stuff and like put it out there and even like, you know, monetize it.

3:06Like, I built like a first website, I think in junior high or so, learned HTML. And like, I mean, it was a silly like fan site, probably for Pokemon. I just remember those early experiences feeling so empowering that like, you know, here I was, I didn't have like a ton of resources, didn't have capital. And yeah, I could create something of value. And like it just all you had, all you needed was an idea and time. And like, I had both of those. and then fast forward into like you know high school college taught myself how to do real programming you know 05 through like 2009 that's when I was in college and I would call that kind of the maybe like either right before web 2.0 or right in the like early innings of web 2.0 and it just felt very exciting because like there you know was like a new form of product value that you could create with like an emerging technology and like at that point I was like okay, I think I'm going to learn the skills to go and build value like this.

4:05I have the time still and sweat, like still no capital. But then the bottleneck became like finding a good idea, right? And it turns out when you're in college, like, you know, most of the problems that are proximal to you are college student problems, right? Like, how do I find a sublet? How do I find roommates? And like, in hindsight, like, these were like fairly small problems. And so I think early on in this journey, I kind of realized like a big part of, you know, entrepreneurship has to be like finding a good problem to work on that's like worthy of like your time and frankly, like the employees or VCs that you'll eventually attract.

4:40It wasn't until after I graduated that I landed on the first problem that I kind of felt was actually worthy of going full time on. And And that was the idea for eTacts, contacts with an E. And it was basically like a CRM-like product that would pull in your emails. And we even built a scraper for AT &T and Verizon and T-Mobile to grab your phone calls and text history and automatically show you this kind of evergreen dashboard of everyone you had talked to. So if you're a VC, for instance, you could say like, oh, when's that last time I caught up with so-and-so entrepreneur? Or what are my notes from the last time we talked?

5:17You know, that felt like a big enough problem and something that like I had experienced firsthand moving out to the Bay Area and like interning briefly at a, you know, a different tech startup, like seeing firsthand, like maybe there's a need for professionals to have a tool like that. Okay, you got e-tax off the ground. How'd it go? Yeah, well, so I teamed up with a college friend of mine, Evan. And I mean, I went to Duke, I graduated in 2009. I teamed up with like one of the literal handful, like I could count on one hand, I think the number of nerdy types that I encountered in school who like also read TechCrunch and cared about like, you know, this like whole web startup thing.

5:56we applied to YC fairly early on because, you know, I discovered YC from a few years prior, like got the callback, got the interview. And before literally going through that process, like, you know, YC was a website to me. It was like something that I had seen from afar, like on the internet, but like this made it feel real and made the kind of path of like doing a startup of our own feel real for the first time. So then you kept playing through on eTax. What was the resolution of eTax before you kind of decided to start Airtable? Yeah, I mean, it felt like a lifetime. The whole company existed only for a year and a day.

6:38You know, we launched the product. We got some early traction. You know, I think we were in a really interesting space. Pretty early on, got some like, you know, calls from different companies that were, you know, interested in either acquiring us or acquiring us. And so got some different, like, you know, knocks on the door. And at some point, we just kind of realized, like, okay, like, we could go this alone. And I do think there's, you know, an independent path to this company, still have funding, we had hired a very small team. But also, my co founder, and I basically had the humility to recognize, like, there's so much we don't know.

7:10And we could try to learn it, like through trial and error on our own, or we can go and, you know, learn at a different company. And then the second time around, like we'll each do our own next companies at some point, maybe together, maybe separately, ended up being separately, but like, we'll have so much more, you know, kind of experience and acumen to do it differently the next time around, right? And maybe pick a different space, right? And so, you know, eventually we decided to go down that path, entertain a few different possible suitors, ended up picking Salesforce partly because I really respected Salesforce as a company that like clearly had done something right in enterprise and SaaS, which like at that time, like Salesforce was kind of the only notable SaaS company, right?

7:55And did you, I'm curious, like one thing I've seen with some folks is their company will get acquired and it's like they're in purgatory until they can do their next thing. But I've also seen some people say, hey, being a big company is not for me, but there's a few things I really learned. I'm curious about Salesforce in that regard. I really felt, I mean, so the reason why we picked Salesforce and I was personally excited to go there as opposed to, let's say, a consumer tech company is I always felt like consumer tech, like if you look at the great consumer tech companies, most of them were founded by somebody who did not work in the last consumer tech company, right?

8:33Like Larry and Sergey didn't work at Yahoo. Neither did Jerry Yang work at whatever precursor Yahoo, right? You know, and every consumer tech company, it's like you almost are, you know, at least the history would imply that like, you're better off like inventing from a fresh canvas than like working within a bigger company and kind of learning their ways. I literally like thought like, not only are they paying us to come in, I would actually pay them to get to go in and learn about this world, right? That was previously very foreign to me. And I think something about SaaS appealed to me because it felt like a more deterministic area to build some value in.

9:13Like consumer tech, it kind of feels like a little bit of like a flip of the coin, right? Like, is your thing, is Vine going to be the next, you know, snap or not, right? TikTok, you know. And I like the fact that in like SaaS, it felt like, okay, you pick like a really good problem space and you solve it. There's not a guarantee, but there's a much more predictable path to creating like a big business that way. So that concept really appealed to me. Salesforce, you know, was the really like SaaS titan, SaaS pioneer, especially at that time, I went out of my way to like do a good enough job and like kind of do interesting enough things there to like kind of get noticed and get to spend time with lots of people throughout the company and learn from them.

9:55So then the idea for Airtable, then I'm curious about kind of the moment of conception of the idea, you know, did it, did it just occur to you in a flash or did it, how were you visited by the muse with Airtable? It kind of came to me as a convergence of one, like the power I saw of the Salesforce platform. And so recognizing that like, you know, Salesforce had become a giant, even at that time, like, you know, It was at$2 billion in revenue. At that time, I was like, wow, this company has reached incredible scale. And the reason why I ascribed to it is like, it's not just that they built like a web-based CRM.

10:33It's actually that that CRM is really an app platform and a database underneath it masquerading as a, you know, kind of just like a CRM product. But the fact that it's completely configurable means that like, you're not selling one hard coded, brittle piece of software. you're actually solving, you know, many different problems. Like any customer can configure their CRM instance or even like use at that time, like they were really pushing hard on the platform itself, like force.com for non-CRM use cases. But many, many, if not most B2B apps can be distilled down into like these, these different app contracts.

11:10Like you have a relational database, you have like a logic layer and you have an interface layer and a product, whether it's Salesforce or before that, Microsoft Access, Lotus Notes, Final Maker Pro, et cetera. If you can deliver on that like kind of app platform form factor, you can solve for a really wide range of B2B app use cases, right? And there's a really large camp to be won. I had also just read Innovator's Dilemma at that time. And like, I was like, oh, this is like classic like opportunity for, you know, me to just come in and build what I know best, which is like, you know, I'm never going to beat, you know, an enterprise company at the game of like going top down and doing sales.

11:47But I think I could build like a more modern, more easy to use product that kind of initially almost looks like a toy, relatively speaking, and comes in below the current competitors, right? Like I'm not going to compete against Salesforce, I'm going to come in with a product that looks like almost like a consumer type of product like Dropbox did. And like, I strongly believe that the best companies are built, not only when there's a great problem space and like kind of a great product market fit opportunity, but like there also has to be passionate skillset fit. And so for me, like I had a passion for like product UX and like I thought, you know, I think I had cultivated a skillset of being able to build modern web-based collaborative products.

12:27And so I also felt like it was a good way to exercise like the thing I was passionate at, like versus I don't think I could have gotten passionate at like building, let's say like a top-down sold like EMR system for, you know, health tech. Yeah, lobbying the government. Exactly, yeah. Business profits, yeah. Totally. So you got started on the Airtable idea, and it sounds like you kind of had a relationship to this topic, right? You'd been at Salesforce and seen the power being a system of record. You had all these apps that were kind of at the edge that kind of were doing simple things with databases.

13:01And I remember this because I remember trying to use them. But you just, I guess, a bunch of things came together to cause you to think that you could do a cloud-native version of this? Yeah, I mean, I think, you know, there's like kind of like moments in time where an opportunity opens up, right? And in this case, like, I remember doing some like early research on the space and finding actually that like Microsoft Access was still a pretty large product, even though it basically had fallen by the wayside and like Microsoft didn't really care about it. I mean, it was like still generating a fair amount of revenue as kind of this like weird second tier product in their suite, as was FileMaker Pro.

13:40Like FileMaker actually was a product of Claris, which actually was a subsidiary of Apple. But like Apple literally didn't care about it. Like it went like I went and talked to a lot of people who were like working on these products and kind of came to the conclusion that like there's real market pool still for this type of product. but for like more accidental than like actual market or strategic reasons, they'd all fallen by the wayside and left an opening for somebody, i.e. me to come in and build like a more modern and web-based, like making the bet on like, okay, we're going to go all in and making this web-based instead of trying to build like a desktop client version of it, right?

14:17Because we could have built what was sketch to Figma, but like we chose to go all in on the web that here, which I think was the right thing. The product made sense. there could be a version of it that would be even more accessible than the predecessors, because it was kind of a pain to install Microsoft Access even, right? You had to literally install it onto your desktop. And if you wanted to share a database with anyone else, you had to install Microsoft Access on their computers as well, and then create a networked share drive with a shared file representing the database to collaborate on it, right?

14:50It's crazy. Whereas obviously with the web, like you could just have a much more instantly collaborative experience and like set up a database in like minutes, not like hours. So I just felt like the moment was right. And like there was a there there that gave us some confidence to say, like, we're not going to we're not building something that the world might not need. The world clearly already needs this. And the current forms of it suck. What was your thought process there? Why did you decide to take I think it was two years, right, to ship? your first product. Yeah. So yeah. What's up with that?

15:25What were you thinking? I think there's like, I mean, look, there's multiple ways to, to get to product market fit. And I mean, I think the lean startup way is a valid one, right. And, and great companies have been built that way, like Twitter. But yeah, I mean, I think like you can either just like shoot a bunch of darts at the board and see what sticks and then kind of iterate from there. I'm not going to like try to premeditate like the perfect thing. I'm just going to go and try a bunch of stuff and then lean into what I discover empirically from that. I think the second approach, though, is like, I mean, there are good precedents for this as well.

15:59Like, think about Tesla, you think about SpaceX, like, these are not companies that you could like, lean startup your way into, right? Like, you have to have like, a plan. And you're solving like, a really hard problem. But like, you know, that if you can solve it, there's a big payoff at the end, there is a surely a market for like, you know, scalable, cost efficient, space shuttles, right or launches right um as there is for an electric car that's actually performant and efficient the real question is more like how do you get there and can can you like find a technological path to get there without dying right and so i think like i think that latter path just always appealed to my style a little bit more like we you know just even like what we were talking about before and why b2b to me was always a little bit more interesting than consumer like i wanted to build something that had like a real craft and quality to it so i I wanted like consumer grade UX, but applied to like a known problem that if I could just figure out how to solve from a technological standpoint, there's a clear market for, right?

16:59And so, you know, like the two and a half years we spent, first of all, that wasn't like two and a half years in a vacuum. We first started with what we thought was the most important part to de-risk. So like there's no point in building a complicated real-time, like, you know, kind of database engine and like the whole backend infrastructure for this. if we can't actually make the product and the UX intuitive enough to get people to just start adopting it without reading a manual, right? Like we have to make the actual product good and we have to make it good on the web. And at that time, like there really was a question of like, can you create a rich enough product experience purely like in the browser, right?

17:36So I'm super curious about this, right? Because I remember when you guys launched on Hacker News, I remember reading about you guys in the early days. There were two things that I remember that really jumped out at me. One was this idea that says, okay, we have to make it easy to use for non-programmers. How did you think about iterating? Like, how do you iterate when you don't have users explicitly? And how do you learn from iterations without being in the market? Yeah, so we did try to get real user feedback even before we officially launched the product. So meaning like the first version of Airtable was actually like a front end only prototype that basically completely faked the backend or the persistence layer.

18:19And the goal was not to actually get people to like put real data, real workflows into the product. But it was good enough that we could shop around. I mean, honestly, mostly to friends and family. But we tried to pick like friends and family or, you know, kind of people in our network who had intentionally diverse workflows. So here's a person who works in BD at Walmart. here's a person who works in VC, right? Like, and see, how would you manage your pipeline, you know, in this prototype, right? And like, you know, basically ask them as a favor to sit with us for like an hour or two hours and like, show us how they could use this product, even though they couldn't actually use it steady state.

19:00And then, you know, getting some validation from that, like we built like a really hacky persistence layer. And then eventually we built like a proper back-end persistence layer. But every step of the way, we tried to cut all the corners we could. We didn't actually ship the real-time collaborative version of Airtable until the very end. We thought that would be a very important key capability and that betting on having that real-time collaborative architecture would be a long-term winning strategy. But it was a lot of technical effort to put, you know, to work on upfront, unless we had already vetted, like, okay, every single other part of this product is working and makes sense.

19:40Okay, now we're going to solve the heavy lift of making it all real time, right? So it was kind of like a steady progression of like, getting more usage validation at different stages, even if it was only like a small handful of people. There was actually one, one friend, he worked at a friend's company. And, and I remember, like we got him to use air table for his um i think it was like a partnerships workflow he was a bd person at a startup and like at one point the product just wiped all his data like lost all his data because it wasn't a proper persistence layer and he was like pretty frustrated because like he had put all this work into it and it just all disappeared and we felt really bad sent him like i think a very nice bottle of whiskey but like you know it's like those were still shortcuts that I think were worth taking because otherwise we wouldn't have gotten any feedback until like the very final reveal moment, right?

20:34Yeah. And it's interesting. One of the things I've talked to some people about before is this idea of implementation prototypes. And so like a lot of the great products I've seen, they demonstrate in the software itself, the fundamental insight behind the product, right? So like, you know, I remember the first time I ever saw a Figma workspace and a bunch of precursors come into it all at the same time, or the first time you ever saw like eight lines of code for Stripe, there was a way that you could show what was different about it in a 30 second window of time and you couldn't unsee it. Right.

21:08And, and what I, what I learned is that for many products, you don't have to have the actual product to deliver that. Right. So it's, it's a little bit like in car shows, sometimes they have a car, but it doesn't really have an engine in it. If you turn on the rate, it's a concept car, right? And so one of the things that I started to notice with a lot of the startups that I worked with that had success is they would do these kind of concept car demo implementation prototypes. And what they were really trying to do was show enough tangible reality to tease out desperation. They'd say, hey, I'm going to show this to you.

21:43It's not completely done, but what jumps out at you when I see it, when you see it? and you knew you were on the right track if their pupils dilate and they say, when are you going to build this and how soon can I have it? Whereas if it wasn't just an obvious ring the bell situation, it's like, okay, we still haven't found it yet. How did you use the fact that you were kind of UI focused at first to tease out surprises where they were more desperate than you thought or things that you thought they'd like, they didn't like? I mean, in a sense, the prototypes we had were the car without the engine, right?

22:16I mean, literally, it didn't have the back end, but it had like all the semblance of working. And so I think we were testing for a few things. One is like, is the UX intuitive enough for people to just kind of pick up and start using? Because I think if there was too much friction to figure out how to use a product, we just believed it would fail because we knew intuitively this would have to be a PLG product, right? And then second of all, I think like to your point, even just being easy enough to use is not enough. Like it has to be compelling enough that like, you know, you can lower the friction to adoption to nil, but like people still have to have a reason to try out this new thing.

22:50It's not like a consumer social app where like they're going to try it just for novelty sake. And so what is the utility that we can provide them on, you know, above and beyond the status quo, which for most people was basically Google Sheets, right, or Excel? Like, you know, most people are not using Access or FileMaker today. They're actually using a spreadsheet for non-financial workflows. So, you know, they're creating their CRM list in a spreadsheet. They're creating like an inventory list in a spreadsheet or the recruiting list, et cetera. And so could we show enough compelling value in an actual interactive prototype that people would get excited about doing that workflow in Airtable as opposed to the status quo, right?

23:29And so to your point, like we really tried to front load as many of these like sizzly features as we could versus like more what I would call like subtle ones. And some of them were like, one, just the ability to attach files into the product or have like colorful, rich field types. whereas especially at the time Google Sheets was like very bare bones it was like text and numbers everything was black and white and now here's Airtable you have like the option of colorful check boxes and drop downs and like an attachment you could just tell it was different right you knew it was immediate it's like yeah yeah and we actually even played that up with the colors and like kind of the visual aesthetic like we made Airtable into an extra colorful product to really differentiate and it wasn't just superfluous colors it was like you know the colors were embedded into the functionality, like the rich field types, like the colorful drop downs, like we use that to really visually show like this is different from what you get.

24:25Yeah. Oh, yeah. There's a lot of lessons in that. Some of the startups I've seen, they make a really compelling product. They take all this risk. But when it comes time to describe how they want to show up in the world, they kind of flinch. They use what I call comparison language. We're a better CRM. The other approach I like to call a reframe, where you're saying, I reject the premise of the current way of doing things. And one of the ways that you reframe is not just in how you describe it, but in how the product shows up, how it works, like all the cursors in the Figma, design as multiplayer rather than single player file versions, handoff, all this kind of stuff.

25:07I consider myself still like my primary competency is like being a product designer, product UX thinker. To me, like UX is the product in like in many cases, if not most cases of products, right? Like a form factor and like how people see it and use it and intuit how to use it. That is literally the product, right? Like you could have all the complicated backend and technology you want. But like, if it doesn't show up in a immediate and visceral way for the user, like it's all moot, right? And so, you know, this is actually relevant as well to like building AI products, using visual metaphors to really show the user, show the customer what's going on and like play up magical things or capabilities is so important, right?

25:53Like, I mean, and in the world of AI, it's like, these models are incredibly potent. And yet, like, It's kind of belied by the fact that the default interface for it is a blinking cursor in a chat box. It doesn't really show the magic of what's happening behind the scenes and all the different diverse use cases you could have for it. And so I really believe in giving visual metaphor to what is actually happening from a technology and magical capabilities standpoint to the user. And I mean, look, that's like the entire reason why personal computing took off, right? Like the GUI was what unlocked the power of computers for a mainstream audience.