Darren [00:00:01]:

Hi, everyone. Welcome to another episode of One of One. I’m your host, Darren Gold, CEO of the Trium Group. My guest today is Wade Foster, co founder and CEO of Zapier. In nearly 15 years, Wade has grown Zapier from a side project at a Missouri hackathon into one of the most capital efficient software companies ever built. Essentially bootstrapped, Zapier now helps millions of customers automate their work by connecting the thousands of apps and tools that businesses use every day. But what I wanted to explore in this conversation is not the business story itself. It’s the leader underneath it. A man who grew up in Jefferson City, Missouri with no entrepreneurial blueprint, who built something genuinely countercultural by trusting what he knew to be true before the evidence arrived. In our conversation, we talk about the Code Red Wade declared in March of 2023, the prescient decision to enact a company wide pause that has led to 100% AI adoption across the organization, what it takes to lead people through that kind of change, and the critical distinction between personal and institutional AI adoption. Please enjoy my very timely, relevant and important conversation with Wade Foster. 

 

Wade, hey. Good to see you. Thanks so much for being in this conversation. I’m really eager and excited to get started.

 

Wade Foster [00:01:27]:

Yeah, thanks for having me.

 

Darren [00:01:28]:

Darren, this conversation comes at an ideal time given what’s happening with the world and particularly AI. And there are very few conversations I’m in with a CEO today where the topic isn’t raised. And there’s really a couple questions, and I’d love to start with those. One is the question I keep getting asked, which is how do I get my internal organization to adopt AI at scale? And I know that’s a topic that you’re very passionate about and that you’ve written very generously about. And the second is, what does AI mean strategically? Like, what are the big strategic implications for my business? And I know that’s also something that you’ve been focused on a lot. So I thought we’d start there. And I’m wondering whether the March 2023 timeframe is the right place to start. You’ve talked about that moment publicly, and if it is, I’d love you to bring us back to that moment and we can sort of trace the last three years together.

 

Wade Foster [00:02:17]:

Sure. So you know what Darren’s referring to. In March 2023, we called a code red at the launch of GPT4. And you know, that has that kind of kicked off our AI transformation. Now, you know, 100% of folks in Zapier use AI daily and a lot of stuff happened in between. But I think if you go back to a key moment, that that date was important. And the reason we called that Code Red then was, you know, the GPT4 launch was notable for a few reasons. One, it came just six months after the launch of 3.5. So it was kind of the first signal that the pace of improvement here was going to be pretty quick. The quality improvement of 4.0 over 3.5 was very noticeable. You know, it felt like you could feel like if this momentum continued, these models were just going to become more and more and more capable. And three, there was major price reductions across these models. And so you felt like the models were coming faster, they were getting cheaper and they were getting better. And if that became a trend line, it was going to have huge ramifications for our business, kid to the old businesses. But, you know, we were pretty fixated on our own at the time.

 

Darren [00:03:34]:

Yeah.

 

Wade Foster [00:03:35]:

And you know, we called the Code Red. It was the first time we’d ever sort of, you even use the word code Red internally. So there was a lot of folks being like, okay, great, Wade, you got our attention. What the heck does a Code Red mean? And truth be told, we were trying to figure out, you know, we were like, hey, this is important and we just need to give it some serious attention. And we did a lot of things around that Code Red. You know, we revamped our entire roadmap to make sure that we were introducing these AI capabilities into our products. Building products that, you know, assumed that AI was a core feature instead of just sort of like a bolted add on piece of infrastructure. But probably the most impactful thing that we did was we paused the company for a week and we ran a hackathon. And you know, part of that hackathon was procuring tools for folks. It was getting our, you know, data retention policies dialed in kind of all the sort of, you know, I dotting and t crossing to use AI correctly. But from there it was just a week of hey, go build with this stuff, you know. And engineers were using the OpenAI APIs. Non technical folks were playing around with ChatGPT and a handful of other tools, but it was mostly chatgpt. This is very early innings of AI. So there wasn’t as much out there as there is today. But the end result after a week was we went from probably around just just sub 10% of folks using AI weekly to north of 50%. So in one week we saw a huge increase in the number of folks using the technology regularly. And when I talk to other companies who have been successful in their, you know, AI transformation, you hear a lot of different techniques and things that have tried. But there, there does seem to be one consistent theme amongst those who are being most successful and it is that they do hackathons or workshops or co building, like whatever word you want to use for it seems to be a common ingredient for almost everybody that has managed to get their, their internal folks really using this technology. A lot of, you know, if there’s a second theme, it’s that the leaders and the CEO of the company is also using it a lot. It’s hard to expect your company to change if you and the leaders aren’t changing too. And so those two ingredients I think have, have really been critical for us and many others to be able to do this. And it’s the number one thing I recommend for people who are like, hey, how do I actually go get people to do this? And it’s like, well, you gotta actually put your hands on the keyboard and do it. You know, it kind of sounds silly, but that’s really what it boils down to.

 

Darren [00:06:00]:

So I have a bunch of questions. The first one is, and I’m actually like, if you’re watching me right now, I’m actually leaning in because I’m excited to get into this with you, Wade. And in retrospect, you know, three years later, it seems obvious. It was not obvious at all in March of 2023 and in fact, it’s only now that a number, the majority of companies are waking up to the question itself. So I’m curious about like, what was going on in, inside of you, inside of the senior leadership team, inside of the company at that moment. And what is it about you that gave you the conviction to declare a code red when there was, I’m guessing, nothing within the business that was troubling. But you saw something, you projected forward, you paused an entire company for a week. That’s a big act, particularly given the time. It’s not so much right now. And I’m just curious to sort of get inside your head a little bit more weight on that.

 

Wade Foster [00:06:55]:

Yeah, well, we had been working with AI for a while. You know, my co founders, Brian and Mike had both started basically being full time builders around AI, trying to figure out what the implications of this were. Pre chatgpt, Mike was the president of the company, he was running all the product and engineering and said, hey, wait, I need to step down from that job so that I can go build with this stuff because I think it’s that important to figure out what’s, what’s going on here. And so he and Brian kind of went off to the side and kind of a lab style way and just said, hey, we’re just going to go start tinkering around with this and figuring out what this means. And, you know, my job was to run the company. And so I said, all right, we got this, let’s go figure this out. And, you know, the first thing they built was this text bot. So, you know, they had set up a Twilio account, gave you a phone number. You could text back and forth with this phone number and it would reply back to you. And so I remember asking it like, hey, what should I get my wife for her birthday? Or I got this weird toe pain. Like, just stuff like that. And you know, it’d respond back and talk to you and it’s like a pretty cool experience. You know, every morning I’d kind of wake up and ask it a few questions and just be like, oh, this is pretty fun. Of course, about A month later, ChatGPT launches and it’s more or less the same experience, just way more polished and better in many different ways compared to that tech spot. But it was the same concept and we were like, ah, yes, of course, of course. This should be like a standalone product. And you know, even then it wasn’t like, oh, let’s go call it Code Red yet. It was just more of this. Like, we’re still learning, we’re still getting excited about this stuff. And yeah, we’re trying to get the company excited. And so I remember sharing ChatGPT and the general Channel in Slack and saying, hey folks, check this out. This is a really cool product. I think, you know, I like cool products. You like cool products. Like, I think it could be interesting and I think it could have ramifications for a roadmap. So, you know, maybe, maybe think about, like, if there’s a place where we could use this as part of improving the product experience. But there was no, like, mandates or, you know, hey, you must do this. Yet it was still very suggestive. And over the course of the next six months, we had a number of moments like that where you just saw interesting things happen. You know, I remember a customer had built a prototype of GPT inside of Zapier and was building zaps around this. I remember another customer built, he called it Voice to Invoice Prototype, where he had a jotform form that had a record button on it on his mobile app, and his wife was a gardener who was horrible at sending out invoices. And so he made this Jotform button field where you could just talk to it. And you would say, hey, I was at a client’s place, and, you know, they wanted the irises for this month, then the violets for that much and the roses for that much. And, you know, can you generate an invoice for me? And that would, you know, Zapier would take that from Jotform, would send that to a GPT and would turn that all into a spreadsheet. And then once it’s in the spreadsheet, it would generate an invoice inside of QuickBooks. So I remember seeing that and going like, whoa, that’s pretty cool. Like, that’s the kind of thing, you know, five years ago, you might have seen somebody raise $500,000 in like a seed round and something like that just to build that. And he just did it. And so you had a moment like that. We had a customer come to our company summit. So we’re fully distributed team. We get the team together once a year. And we had one that came to the summit, so who talked all about his experiments with AI and what he thought about it. And, you know, he. I remember subtly encouraging him to talk about what he thought the implications were for, for Zapier. And I was like, don’t hold back. And I remember him saying to the team and the company, he’s like, you know, you guys have a really. You’re at a really interesting moment, because I think if you really embrace this stuff, you can be a really important company. But if you ignore this, like, I think people will go use other stuff instead. So we had like, these kind of, you know, handful of experiences, and there was many more where people were just getting more exposure and, you know, starting to realize something important is here. And all that kind of culminated in that GPT4 launch, where we just remember when that came out, getting access to it, playing around with it. And that night, Brian, Mike and I on a zoom call for like an hour, an hour and a half, and just being like, hey, playtime’s over. Like, it’s. It’s go time now. Like, this is. This is it. And so we said, you know, we talked to the exec team at 7am the next morning on a Friday. And by Monday, code red in place. We’re going for the whole company. And you’re right, not many people were doing that in March of 2023. In fact, I don’t know of hardly any that were at that point in time. You know, there was enough people in the company that had exposure to these concepts where some folks were like, I get it. I see. I see why this is. But there was a lot of people that were also like, code Red are like, sound pretty alarmist, Wade. You know, you’re scaring people. Like, that was a bad thing. And I remember thinking, like, that’s actually why I called a Code Red. It’s like, I wanted people to pay attention. I wanted people to pick their heads up and take this seriously. And so that critique sort of felt like, you’re right, but, like, you’re critiquing. It should actually be like, thank you for calling Code Red. You know, of course, in hindsight now, like, that has been proven correct, but at the time, yeah, it just wasn’t. It just wasn’t a universally accepted thing. You know, if I could go back to that time, you know, probably the biggest thing I would encourage myself is to be even more bold. Like, Code Red may not have been enough. And so I look at a lot of these companies today who are, you know, kind of riding the fence, and I’m like, you. You kind of just gotta. You gotta go. You gotta go. I don’t know what hard decision you’re facing. I don’t know what thing you’re waffling back and forth on, but I can guarantee, like, in your gut, like, in that pit of your stomach, you probably know what you need to do. And it just sounds painful to do it, but you kind of got to go.

 

Darren [00:12:38]:

Yeah, I’m sensing there’s, like, this quality of really strong conviction and almost relishing in people not misunderstanding you, but maybe underestimating. I don’t know, am I onto something there or just. I’m just wondering, like, what.

 

Wade Foster [00:12:55]:

I don’t. It doesn’t feel that way at the time, I’ll tell you that. Anyone who has, like, gone against the grain and has, you know, any sort of desire to, like, please people, which I think if you’re an entrepreneur, you kind. You’re. You’re trying to build things that people like you don’t necessarily love running against the grain for all that long. You want to run against something that people are just like, yes, let’s go do more and more and more of this stuff. You know, I think there are. It’s nice to be proven right. That doesn’t always happen. Sometimes you just straight up get it wrong. That part is nice, for sure. But, you know, I think in those Moments it is difficult to say, hey, I’m going to do this unpopular thing. You just don’t see it that often in society. Like you don’t see it with politicians, you don’t see it in your communities, you just don’t see it all that often to say, hey, I’m going to do an unpopular thing, even though I think it’s the correct thing to do.

 

Darren [00:13:45]:

Yeah. So I’m wondering how long did it take until you were no longer running against the grain and you had some wind at your back? And maybe we can start now on this three year journey of what you actually did, what you’ve learned, where you are today and what kind of still remains.

 

Wade Foster [00:14:00]:

Yeah, it’s like the classic, like early adopter curve. Even to, to this day, there’s like a pocket contingency of the company, maybe 10% that are out in front, that are pushing, and then there’s a group of people that are more cautious and saying like, hey, is this the correct thing to do it? They’re, you know, evaluating the technology more, you know, methodically, and then there’s a group of people who are just sort of like resistant. So I think that’s still true even to this day. But, you know, the sort of like, hey, is AI important and is it here to stay? I’d say we got over that hump. Probably took maybe a year, I would say. So it did take some time before you really saw a lot of people lean in. But every step of the way there’s like a still a re underwriting of it. You know, we saw in fall of 2025, November of 2025, really, this latest batch of models and they got way, way, way better. And all of a sudden new ways of working emerged from that. And so that created a new moment where it’s like, hey, engineers, are you actually using these Cogen tools? Like, are you still writing code by hand or you not? And like before that moment, you kind of still did do a lot of it after. It’s like, actually it should be a lot different now. And so there definitely has been, you know, shifts every step of the way as just more capabilities emerge.

 

Darren [00:15:16]:

Yeah, I’m also as the little I’ve read about you, or maybe more than a little that I’ve read about you, suggests this interesting combination of urgency and patience. As I read a little bit about your story, I was really struck by how much patience you had. I was expecting somebody was just like foot to the floor. You got, you’re in, you’re out, which is Sort of the vibe you get from a lot of people that are trying to push the envelope here. And what I was starting to take away, even in hearing in this conversation, is the wisdom to know things, take some time. So talk a little bit about. Am I pointing to something there that resonates?

 

Wade Foster [00:15:59]:

I think it was John Wooden, legendary UCLA basketball coach, who had this saying, you need to go fast, but don’t hurry. And I think what he means by that is every day you have to wake up with a sense of urgency that you don’t want to waste a minute. You’re always looking for, like, you know, how do we move the needle, how do we push things forward however little we can this day? But recognize that it just takes time to build stuff. And, you know, I think that’s something that I’ve started to appreciate more and more the longer we’ve done this. You know, we’ve been in business now almost 15 years, and I’ve seen so many people come and go. One of the things that if you’re an entrepreneur for any, any amount of time, you know, you have, I don’t know, investors or partners, and people love to text you, like your competitors, like press releases and news and stuff like that. And you’re like, did you see this? Did you see this? Are you paying attention to this? Which, it comes from a good place. They’re just trying to make sure you’re on top of things. Of course, any entrepreneur is like, yeah, I saw it. I’m on. I’m deeply aware of what everybody in my space is doing. So I have this, like, 15 plus years of people doing this. And if you go back and look year by year, you can see the companies come and go, come and go, come and go, and very, very few have staying power and longevity. And so I’ve really come to appreciate some of the, like, legendary entrepreneurs who are just gone for such a long period of time. You think about, you know, Bill Gates run at Microsoft or what Zuck’s doing right now at Meta, you know, what Jeff Bezos did at Amazon. These are people who went for decades on the same problem. And, you know, they had bad years, they had bad quarters, they had periods where things didn’t work out the way that they’d hoped. But it’s really hard to compete against somebody who is like, I’m just not going to give up. And I’m just going to find a way to keep compounding over and over and over and over again. And, you know, if they’re good, which in the case of all those people we just mentioned, they’re all. They’re not just good, they’re great. They can have bad quarters, but very unlikely they’re going to have bad decades. Like, they’re just going to be really, really excellent across long periods of time. And so I. I don’t know, I just try and keep that, like, duality of, you know, wake up and just do the. Do the next hard thing, do the next important thing, but also recognize that, like, hey, we’re. We’re playing a long game here, and if we can. If we can play that game effectively, we’ll just outlast everybody.

 

Darren [00:18:26]:

So we’ve sort of assumed the benefits of AI adoption as if they’re a given. But I’d love to hear your articulation of, like, why was it so important for people to adopt, you know, to get to 100% adoption? Like, what are you seeing as the benefits of that? Within Zapier?

 

Wade Foster [00:18:42]:

Yeah, I think you get a couple things. One, when everybody has a fluency with AI, you definitely have better ideas, more novel ideas for where to take your products and what should you build for the company? Which is really important because I think one of the most impactful ways, and I don’t feel people give this enough attention, is how can you actually build, like, AI native products, and how can you actually build services that are better than the prior generation of services? If you do that really, really well, your growth rate and your revenue can be immense. And we’ve seen that in the last couple years where, you know, some of the companies that have done this exceptionally well, their growth rates look far, far different than the growth rates of the best companies of the past. You know, Anthropic is one that’s in the news all the time right now, where they went from, you know, one to 14 in, like, 13 months, and then 14 to, like, 30 in, like, six weeks or something like that. I’m probably getting the numbers wrong, but it’s directionally correct. And that’s just insane to see that kind of growth at the scale that they’re at. So, yeah, I think if you really understand how to build with AI, you have this ability to transform the types of products you build in a way that just is really good for your customers. And of course, the more obvious one, and I think this is the one that people talk a lot about, is how does it change how you operate internally? Do we get more done? Do we do things that we couldn’t do before? Can we improve our cost structures? All that sort of stuff? Becomes possible if you’re leaning into the technology really, really well. I won’t sort of sugarcoat it and say like, oh, it’s super easy, like everyone just like learns how to use AI and ta da, your company’s like magically more productive. I think there’s a lot of learning curves to that companies are figuring out on how to do that. But I think when everybody is building with it, you get a lot more energy in that direction versus a lot of companies feel like they’re fighting internally just to get people to even agree to try to do this stuff.

 

Darren [00:20:38]:

Let’s pick up on this. How do you compete with AI native companies? And maybe just to reference a book that I see in the background there, which is Seven Powers and Hamilton Helmer who was on the show, talks a lot about counter positioning. And I think there’s something about people and companies and founders that get to start from scratch and truly build something that’s AI native, have counter positioning power relative to incumbents that have decade plus of entrenched code and ways of doing things and their very kind of survival and interests make it very hard to tear that apart and compete. Sounds like you’ve been thinking hard about that question. I’d love your take on it because it’s also another big question that’s coming up in the conversations I’m having is we’re at a competitive disadvantage relative to the 15 person company that isn’t saddled with all of the things we’re saddled with. How have you thought about that?

 

Wade Foster [00:21:36]:

Yeah, I do love the Seven Powers framework. I do think counter positioning is probably one of the best weapons that people in startups can use today because AI causes business model transformation and business model disruption as much as it does anything. And so you can look at particularly public company tech, SaaS, stocks, and they’re in a really difficult position because they’ve got all these structural disadvantages that they have to change and they’re a public company. That makes it very difficult for them to do that because it might make their financial statements look worse to Wall street in any given moment and they will get hammered if they make those changes. The example I most commonly hear is things like seat based pricing. Seat based pricing has been a stalwart for these companies for forever. And you know, if companies are not buying as many seats or you know, seats isn’t the right way to consume this. If usage based pricing is the correct way, you know, there’s a lot of reason why the customer base might not want to buy as many seats anymore. And so all of a sudden you have to reshape your pricing model. And that drives, like, the entire business. It drives the entire, like, foundation of revenue contracts inside of, you know, some of these companies. And it’s one of those things where they can lip along and do price changes for, like, raise prices, you know, here and there and kind of juice earnings quarter by quarter for a few quarters. But it doesn’t actually fix the structural problem, which is that they need a fundamentally new way to go sell the product. And eventually that kind of catches up to them. You know, you think about, like, the AI transformation of these businesses, it’s not obvious to me that the way you ship software, the way you sell software is going to stay the same. In fact, I think it’s fundamentally changing. You know, we’ve gone from a world where code was expensive, and when code is expensive, you have all these rituals that happen before you write code. You recognize that code is expensive, engineers are expensive. And so you do all this customer research, you do all these data analysis, you debate the roadmap, you align on the work. All this stuff happens early in the process, and eventually you align on something to go build. And once you do it, you hope that you got it right, because code is expensive to build. And now an engineer goes out and sets aside to go build that thing. But now we, we exist in a world where code is cheap. Engineers can wake up and just simply decide to build all that stuff, and it doesn’t cost all that much. Now why is that important for internal counter positioning? I don’t know if this is a counter positioning thing. It’s just, you know, these companies that have hundreds, thousands of people employed in them, folks whose job it was to do this particular thing in a particular way, they are incentive for that system to stay the same way. Whereas a small company can just never hire those people. And they can just be like, all right, we’re just going to do it this new way. And so, you know, it creates this headwind that an incumbent has to go fight up against. And so it gives the upstart, you know, maybe more advantages than they would usually get being an upstart. So I don’t know, it’s a really interesting time to be a startup because I think you have, you’re going up against these companies that are struggling now. The flip side is also too, like, I’ve not seen every income like in the past. The meme is that, like, incumbents are like, big and slow and they sort of don’t pay attention to the trends with AI. I don’t see that like, I think everyone is like very keenly aware of the importance of AI on their business. The incumbents have been very, very fast to jump on this stuff. And you know, one of the other big challenges is that it’s harder than ever to get distribution and get attention. And one thing incumbents are great at is they’ve got a marketing channel, they got a sales channel, they know how to get in front of people. And so they might have these other disadvantages, but they have these things going for them. So it’s a, it’s a really just interesting time because the playing field has shifted and it’s not yet obvious to me, you know, what the, who the winners and losers are. It’s going to, it feels very, very early still.

 

Darren [00:25:36]:

And to make it specific to Zapier, like how are you thinking about and what have you done? And maybe one of the things was three years ago you made that decision to start early. I’m just curious, as you’ve thought through this strategic dynamic, you know, the choices that you’ve been making to not lose competitive advantage and maybe gain it.

 

Wade Foster [00:25:56]:

Yeah, I mean there’s a. Without sort of giving away like internal strategic stuff. There’s been a number of things that we’ve done that I think that have been important. A couple of years ago, we made a massive price change where we made Zapier much more affordable for our customer base. We made it so that every plan could have a pay as you go option so you could do usage based billing without having to flip over to a higher paid subscription. We took all these buy Zapier apps that we had that felt like folks shouldn’t have to pay for those things and we gave them away for free. Every plan had a limit on the number of zaps you could build. And we said, no, every plan is going to be unlimited zaps now for free. You can just do that. The end result of this was we gave away something like 30 million in ARR to our customers. But yeah, that’s one of those things that I think a public company would find it really difficult to do because that’s a lot of money, you know, to do those types of things. But we said, hey, we want to take our lumps now and we want to make sure that we’re positioning ourselves well for the future. You know, just last week we launched a new product. We launched the Zapier SDK. That was like a very different thing. Zapier is known for our no code, ease of use, things like that. Well, the SDK is a tool for coding agents and the power of it is that you can go install it inside of Claude code or Cursor or Codex and hook it into any of the 9,000 plus apps that we support on the platform and now in a governed way have access to all these different tools. You don’t have to be pasting API tokens in plain text into these environment files and things like that. So you know, that was a specific tilt into thinking who do we actually serve? Like, who is our actual audience in this new world? You know, we’ve got more things like that where we’re sort of stepping back and trying to understand like where do we have a competitive edge, where do we provide like durable value and you know, where do we have the right to win and let’s go work on those types of areas. There’s many things we could do, but are those things that we are uniquely good at doing?

 

Darren [00:27:58]:

Yeah. You’ve mentioned Public Company a couple times and I think it points to the importance of structure to enable the conditions for you to make those kind of forward looking, longer term bets. And I know that you, whether intentionally or otherwise, have built Zapier in a way that’s different than most venture backed companies. How much has that helped you in terms of orienting towards the future, being able to, you know, make decisions that other companies, given a different structure might not be able to make?

 

Wade Foster [00:28:33]:

Well, I think it has helped a lot. You know, I think we were able to just think a lot more long term. You know, we can sort of do what’s right for the business and you know, we don’t have a lot of the, you know, stress that comes from somebody saying, you must achieve x percent number next quarter or else I’m going to, you know, take away your funding and you’re not going to be able to meet payroll. So instead we can run it like an actual business and say, hey, like we think that this is the thing that is going to be best for us over the next five years and even if it requires us spending more money here or doing, you know, reducing our revenue in this way, if we believe it’s the correct thing to do for our customers, we can just decide, let’s do it. We don’t have to worry about convincing other people that that’s the correct thing to do.

 

Darren [00:29:17]:

Yeah, yeah. So maybe we kind of go back to the AI adoption and just sort of complete the swing. Where are you today? What have you achieved? I know some of the things that you’ve been public about that. I’ve been really both appreciative of your generosity to share those and really impressed by some of the AI fluency rubrics that you’ve put out there. But bring us up to the present moment in terms of this three year effort that you’ve been leading.

 

Wade Foster [00:29:43]:

Yeah, I mean, 100% of the employee base is using AI Daily now to accomplish their goals. We have this AI fluency rubric that you mentioned. We launched it last year. We just released a V2 of it. You know, from what I’ve heard, a lot of companies ranging from smaller growth stage startups all the way to massive public companies, are now using that rubric themselves to help shape their own internal transformation. We use it for evaluating new hires, we use it to help with performance reviews. It’s just a really good way for raising the standard and raising the bar for what it looks like inside these companies. I think the next frontier for us is what I would call moving from individual AI to institutional AI. And this is one of the things I see a lot of the companies that are on the frontier of adoption. It feels like the next hurdle a lot of individuals will cite, hey, I am much more productive because of this technology, because of the way I’m using this. But if you go look at it from a company lens and say, hey, are we actually, you know, delivering features faster or are we reducing our cost, or are we improving our customer experience? Like, is business metrics changing as a result of this? You know, the number of examples, there are much, much fewer. And I think the big challenge here is that this is where it’s not just about an individual saying, hey, I’m just going to use it more to do all the same things I’ve already done and do it faster, better, cheaper. The real challenge is you have to change the institutional operating system. And that means redesigning the org chart, redesigning systems and processes, you know, taking humans out of the loop here, maybe adding them to a new place. You know the thing I mentioned before about how code has gone from expensive to cheap? Well, the entire software development lifecycle needs to now change because of that. It’s. And so this is where we’re spending a lot of time now is paying attention to. Okay. Because we have this like amazing technology and everyone really understands how to use it. Now, how do we actually just fundamentally like flip the table over in terms of how these, like really important systems and processes work internally so that we actually have more than just individual acceleration with AI? We want to have institutional acceleration with AI.

 

Darren [00:31:55]:

Any early hypothesis that you’re holding in terms of the unlock there, or are you still in the.

 

Wade Foster [00:32:01]:

Yeah, you know, I think the. The biggest opportunity that I’m paying attention to is the, what I call, like the coordination tax. As organizations get bigger and bigger, you start to have more and more humans that are helping just coordinate the work. And, you know, a lot of this is product team works on building XYZ feature. Okay, XYZ feature is done. And now there’s like a fleet of humans that say, okay, we got to go take, you know, what has happened and we got to write, you know, documentation, help documentation for a public website. We’ve got to do new sales decks. We’ve got to do training material. We’ve got to have a launch campaign. We have to basically repackage the same fundamental information in 5, 10 different ways to go to all these different sets of constituents inside the organization and help them all understand this fundamentally same bit of information. But the reality is that translation layer, that’s what AI is really, really good at. These are large language models. They are really good at taking some corpus of information and spitting it back out in a slightly different way. And so you can hook an LLM up to your code base and say, hey, every time fresh code goes out, I want you to look at it and tell me, what is this feature? Why does it matter? And rewrite it for all these different stakeholders. And so all of a sudden there’s like all this coordination tax that’s happening internally, where people are going to meetings and cross functional stakeholders and da, da, da, da. And you can cut down on all of that stuff pretty immensely. And so that, to me, is one of, I think the biggest challenges companies face as they scale is they pay this really heavy coordination tax. And it’s not particularly fun. The people in those jobs don’t often feel like that’s the most valuable place that their time can be spent. They would rather be spending time building and shipping products, selling product, marketing the product like there’s other areas that they rather spend time. But it’s a necessary evil that a lot of these companies fall prey to. And so I think, you know, if you can sort of really reduce that coordination tax, you can also redistribute those humans to the things that also. That. That also, like re, accelerate the business, which is more product out the door, more marketing campaigns, more sales efforts, like the things that really touch an impact, the core business that you have. So that, to me is like, I think one of the things that I’m most excited about, I think you’re also

 

Darren [00:34:23]:

pointing to something related but different, which is the sense that I get in even my own personal experience, but certainly listening and hearing from others, which is there’s something about when you remove the tax that you’re referring to. There’s a element of creativity and energy and joy and maybe it’s a little too much to say kind of human flourishing that is enabled, and it may be, I think, the greatest promise of all in terms of like what, you know, what the next paradigm of doing creative work looks like together. Is that too far fetched of a notion? And, you know, what’s your experience been with that?

 

Wade Foster [00:35:07]:

Well, I think you’re onto something like, you know, I don’t know about you, but when I talk to my friends, my colleagues, the folks that are most AI pilled, as they say, and you ask them, are you working more or less? Most of them cite, I’m working a lot more. And then you ask them, why are you working more? Are you working more because someone’s breathing down your neck to tell you to work more? Or what’s going on there? And oftentimes folks are saying, I’m working more because I want to work more. I’m having fun. Like I like doing this stuff. There’s a lot more side projects, there’s a lot more dabbling, there’s a lot more experimentation going on. And I think it’s because the technology just makes it a lot easier to do these things that were hard to do before. In the past, you might have been like a master at one particular discipline. I was really good at this thing. But you still have ideas about how you might do some other thing that you’re not an expert at, but you kind of just don’t even have the baseline level of skills to do it. So you never get a chance to, to. But with AI, you can be, you know, competent very quickly at some other thing. And so the ideas that you’ve had in your head, you actually can start to bring more of them to life. And then you get in this like loop with the LLM where you’re just like, I want to do it, I want to do it, I want to do it. I’m going to do more and more. And the way the feedback mechanism loops is it’s so quick. The feedback is so, so, you know, it takes a short prompt is instant. A longer prompt might be, you know, I don’t know, five, ten minutes, the longest I’ve seen. Or, you know, we’re starting to get things that run like an hour or two, but mostly, like, you have a really fast feedback loop, which is very different than working with humans. Like, if I delegate a task to you, you might come back to me at the end of the week and say, okay, I’ve got it done. But with AI, you get this almost instantaneous feedback loop, which just makes it more fun to keep getting into the work. And so it does feel like we are tapping into a deeper well of creativity that folks have where they’ve maybe had these ideas or they’re now having more ideas and they’re able to bring more of them to life because they have an outlet that allows them to express those ideas.

 

Darren [00:37:14]:

Yeah. I want to move us in a moment to the bigger question of culture, because my guess is that this doesn’t happen in the absence of an intentional culture that you and your co founders and leadership team have cultivated and curated. Before going there, just one quick question on the AI adoption. You made the decision, I think, last year to name Brandon Summit your chief people officer, also chief AI officer. And I’m interested in both of those. The importance of having a single person that has that kind of responsibility and then their choice to make the chief people officer and Brandon, that person in particular.

 

Wade Foster [00:37:54]:

Yeah. So, yeah, I think this is. There’s a couple things to. To know here because I’ve been asked this a number of times. I think a lot of people’s initial instinct is like, oh, I should just make my people officer, my AI person, and be done with it. And that’s not exactly what I’m recommending.

 

Darren [00:38:09]:

I get it.

 

Wade Foster [00:38:10]:

You know, I think we resisted having someone in that role for quite some time, like almost three years, in part because it felt like AI is something that all of us need to go figure out how to use. It’s not just one person’s job. And I saw a lot of companies struggle with this where they would have these AI work groups or these AI councils or these AI guilds, and, you know, the guild was in charge of figuring out AI. And so you’d have one or two or three people in there that you’re really excited about, and then you go talk to the rest of the company and the rest of the company’s doing nada on that stuff. And you, you know, you sort of have that dispersion of responsibility and that’s the result you get. And so we said, no, no, no, no. This is something that everybody’s going to be, be working on. And that approach is what got Us to, you know, 100% of people using all this stuff. Then we started to have a new set of problems emerge. Okay, we’ve got 100% of people using this stuff. We’ve got individual, you know, AI transformation that has happened. We’ve got millions of ideas for how to go even further, but they’re not things that an individual can do. We’re trying to go solve these institutional AI problems. These institutional AI problems require rethinking these core systems and processes, rethinking how the org charts work, rethinking a lot of these concepts from the ground up. Not technology problems, not like AI adoption problems, but org problems. And Brandon’s team had already been doing a lot of this stuff exceptionally well inside the people Org. And you know, what better than to take the person who’s already leading the charge on these things and just so happens to be the HR person who thinks a lot about like, you know, org design and systems and culture and all these things and say, hey, come help the whole org out with this entire problem. So for us it was like a really, really natural fit. But I think, you know, depending on your company, you might have a different set of problems. Like maybe you do have a technology problem and you need somebody that’s more tech forward to take that role. Okay, great, have them go take it. So I think this is one of those cases where you, as the CEO or leader of the company, have to take stock of what is our bottleneck, what is a problem that we are uniquely facing and how do we go solve that to figure out whether it’s you or somebody else who needs to go carry the mantle on what needs to happen next?

 

Darren [00:40:28]:

Yeah, this personal institutional distinction I think is so important. I think it will continue to be. I’m, I’m looking forward to maybe even having a follow up conversation with you at the right time on that culture. Can we talk about your philosophy around the importance of culture, how you build it, and then what you’ve done that you would consider distinctive that might be helpful for people to hear?

 

Wade Foster [00:40:49]:

Well, gosh, I thought a lot about culture over the years. We’ve done a lot of unique things. We were one of the very early fully distributed teams. Never had an office. We started working that way in 2012 and really haven’t looked back. And so we built a strong culture around asynchronous work, written communication, documentation, things like that, Strong default to action culture, strong bottoms up culture, all those types of things. Turns out some of those decisions have actually been really important. I Think in the AI era, the fact that we have a strong writing culture, the fact that so many things are well documented, turns out to be very good for LLMs. So all this context that we have about how we work and, you know, what works, what doesn’t work, gets ingested by these tools and helps the next person come along figure out, well, why did Zapier choose to do this? Or how does this system or process work? Is there a standard operating procedure for this, that or the other? All that stuff is much easier, much more readily accessible with a chatbot or an LLM that sort of sits on top of that corpus of information. So that has been a massive advantage for us from a culture perspective. Yeah. I think the other thing I’ve been reflecting a lot on lately as it relates to culture, is I heard another fellow CEO who shared this with me and said, you know, you have a strong culture when you have a winning culture. And so if the company is being successful by whatever metric you deem to be successful, then your culture is strong. If you are not being successful at your goals, doesn’t matter if, like, people like the culture, you don’t have a winning culture. And so that’s one of those things that I’ve reflected on, especially, you know, as we’re in this, you know, intense period of, like, creative disruption, creative destruction with AI, where I’m like, hmm, what does it actually take to win in this era? And do you have the right, do we have the right cultural ingredients to actually win, not just enjoy working in this period, not just to do some interesting, cool things, but to actually come out on top? Much more challenging question to answer that. I think a lot more people, when they think about culture, they think about, is your glassdoor score high? Like, do people enjoy working at the company? Is your retention rates good and not as much on, like, will you actually build products and services that durably last for your customers? And yeah, I think that CEO is right when he sort of described it like, the best cultures can have lots of different variables. Lots of them can potentially lead to that outcome. But if you don’t have that outcome, ipso facto, you don’t have a strong culture.

 

Darren [00:43:31]:

As you look at what is going to be required to win today, where your culture is, what are the elements of it that are like, yeah, that’s absolutely going to drive winning, and where do you need to evolve your culture? And then I’m also interested in how have you, over time, when you’ve discovered, wait, there’s still some gaps or there’s something that needs to shift how you’ve. How have you actually intentionally evolved your culture?

 

Wade Foster [00:43:56]:

Yeah, I think we have a lot of the right ingredients. You know, we have these five key values. Default to action, default to transparency, empathy over ego, growth through feedback. Don’t be a robot, build the robot. Those are our five values. So they’re all like, they all feel like the right ingredients. The question is, do we live up to them as best we can and do we manifest them as best we can? Now, I already mentioned the default to transparency. I think that is a huge asset for us right now in the age of AI. So much of our work, whether it’s the conversations we have in Slack or the meeting transcriptions that come out of all of our Zoom calls, all of that stuff is legible to AI. And so that means we are able to build this corpus of knowledge that can act as a coach, an assistant, a decision maker, all these different things for so many systems and processes in this future. And that compounds organizational speed. So you start to get that, that default action gets much easier because you’ve sort of leaned into this, you know, transparency moment. You know, I think the one that’s been more challenging for us is that empathy over ego. I think this one is, if done well, actually can serve you really, really, really effectively. But I think done poorly, it can hold you back. And so here I like to borrow from. I heard this from Ravi Gupta, who was a. A partner at Sequoia. He has this amazing line which is the best leaders are both demanding and supportive. And the magic happens when you get the intersection of those two things. It’s easy to be one or the other. You know, 10 out of 10 demanding. You can sort of imagine what that looks like. You probably hit your goals, but you might be a tyrant and people don’t want to work with you. The flip side, you know, 10 out of 10 supportive. It’s very easy to imagine what that looks like. Pizza parties every day. You know, people are having a good time. You know, you’re high fiving, but you may not hit your goals and you may not be successful. It’s the combination of the two that is really tough. And I think most of us, if we think about the leaders, the mentors, the people in our lives that sort of actually do both of those things together, it’s probably a shockingly few number of folks that actually manage to do those things successfully. Maybe you can think of a coach that you had growing up or maybe a teacher who was putting their hand over your shoulder. And being like, yeah, I know you messed up. You know you messed up, but I also still believe in you. They managed to ride that line just perfectly. And I think the best folks do that very, very effectively. And so that’s kind of what our empathy over ego value is supposed to get at, is this, like, you know, demanding and supportive component. But I think human nature, and especially as organizations get larger and larger, the default is to lean harder and harder into supportive. And so you can have a culture of kindness that permeates, but it doesn’t speak the truth. And so you have to find a way to marry those two things successfully together. Where, you know, I can tell you, Darren, hey, you’re messing this thing up and not demoralize you. And you actually believe that. I’m telling you that because I actually believe in you, and I want you to win, and I want you to be successful, and I’m rooting for you. And the best thing I could do to help you succeed is to tell you that and to build that type of environment, I think is really tough. Like, you know, it’s hard. Again, I think, you know, it’s just shockingly rare to be in organizations that truly embody that kind of attitude over long periods of time. You know, I won’t pretend to say that Zapier is perfect on that dimension. Heck, far from it. But we aspire to try.

 

Darren [00:47:15]:

Yeah, you see me nodding, because it’s actually. We call it high care, high standards at Trium. It’s an explicit pillar of our culture. I go so far as to use the word love. I call it love and performance just because I want to be provocative. And I agree with you 100%. So I’m curious to maybe use this particular example to answer the question. As you are sitting here as CEO, you’ve identified this polarity as kind of central to your values and your culture. There’s work to be done. What do you see as your singular leverage as CEO, and how do you use your position, if at all, or what do you do other than your kind of own influence to drive that change in your culture?

 

Wade Foster [00:47:59]:

I’ve gotten a lot more comfortable just trying to model the behavior 100% of the time. There’s been periods, as I’ve continued to learn how to do this job well, where I’ve said, oh, I’m going to delegate that to someone on my management team, or someone else will spot that error and they’ll catch it. Someone else will do this, that, or the other. And sometimes they do, sometimes they don’t. But what I have found is one of the best things I can do is continue to just reinforce why this way of working is so important. And no matter how big or no matter how small the thing might seem, to call those things out and connect it back to these core principles that we have. And so that’s how I try and go about it. And I hope that that way of working sort of permeates through the organization and sort of, like, reflects through and hopefully the best version of it does. I don’t have any other sort of mechanism that I can think of that I do beyond just trying to, like, live and breathe it 100% of the time and not, you know, shy away from it because that’s a rough day or, like, I’m tired or I don’t want to do it or whatever. It’s like just waking up and just committed to, like, do it, do it the right way. 100%.

 

Darren [00:49:12]:

Yeah. I don’t have any other way either. It’s the single piece of advice I give, and it’s great. I mean, we’ve covered so much ground in such little time. But I want to just give you, as we kind of bring our conversation to a close, an opportunity. Anything that we haven’t discussed that you want to make sure we cover or something that we have that you want to put a finer point on.

 

Wade Foster [00:49:34]:

It’s a great conversation, Darren. I didn’t get a chance to chat much about Zapier products, but I hope you all check out the new Zapier SDK. It is an incredible product. If you go to my Twitter X profile, I have pinned to the top how I rebuilt a chief of staff with the Zapier SDK and it has totally changed how I work. I think a lot of these concepts we’ve talked about today, like being able to have all your slack and meeting transcription legible to AI. When you connect your tools to AI, which Zapier can help you do, it makes some of these cultural tenets that we’re talking about a lot easier to enact because you just have the data more readily available to yourself to. To go work on top of it.

 

Darren [00:50:12]:

Yeah. And I’d also point people to your writings and your LinkedIn posts because I think you’re making a really valuable contribution to a topic or set of topics that people are really interested in and finding enormous value. And so thank you for doing that and thanks for an incredible conversation, Wade. I’ve thoroughly enjoyed it and appreciate you spending the time.

 

Wade Foster [00:50:34]:

Love it. Thank you, Darren.

 

Darren [00:50:35]:

You’re welcome. 

 

What a wonderful conversation. What stays with me from Wade is something simple and difficult in equal measure. The willingness to act on conviction and make an unpopular decision before it’s obvious to everyone else. That quality, the capacity to see clearly and move without waiting for permission is, I think, the most underrated thing thing in leadership. And it is very rare. I look forward to being with you on the next episode of one of one. Until then, I hope you live and lead with courage, wisdom and above all, with love.