Interview with Wade Foster, CEO of Zapier

The next thing that I think is where you really see the big unlock is when folks are now starting to identify opportunities to automate work that you weren't doing before. These are when you're now like solving net new problems. This is a problem that like either you couldn't do it because like, ah, there's just no way we could hire that many people and do it reliably enough or what have you. Or it's just something at like your scale, like you feel like we just don't have the resources to go, to go tackle that. And this is where I think folks start to really get a bunch of leverage.

Welcome to the Exponential Scale Podcast, the show where we talk about how tiny teams with big ambitions are doing more with less Scaling Smarter, Not Harder, and Building Big Businesses Without Building Big Teams— unless you count the bots. I'm your host, Ron Schmilzer, and, and today's episode, we're talking with Wade Foster, co-founder and CEO of Zapier. So now before we jump in, I need to mention to everyone listening that today's guest and his company is responsible for more accidental employees than any founder in Silicon Valley. Not human ones. I'm talking about the army of invisible workflows working quietly throughout your business while you sleep, shower, or pretend to be productive on LinkedIn. And Zapier is basically the unofficial HR department for micro teams. You hire no one and Zapier corrals 8,000 apps and tells them how to behave. So if the core idea behind the Five Laws of Scalemaxing— and if you haven't heard that, please do listen to our podcast on that— is don't hire more, multiply more, then Wade and his company Zapier are the folks handing out the multiplication tables. So micro teams everywhere are stitching together million-dollar operations powered by one actual human and a bunch of Zaps held together with hope, caffeine, and Google Sheets or Airtables if that suits your fancy. And if you've ever said, oh wow, I thought a real person was doing that only to learn it was Zapier, congrats, Wade is already part of your company. So today we're gonna be talking to the OG here, the real source to talk about automation, AI, and what it means to build a business where the bots outnumber the humans, which according to the manifesto is not a bug, it's a strategy. So if anyone was here before AI became everyone's second language, it's Zapier and Wade. So Wade, really thrilled to have you on the podcast here and would love to share your insights with everyone.

Yeah, thanks for having me, Ron.

Yeah. So I'll just quick intro and then you can introduce yourself as well. So Wade Foster, for those of you that don't know, is co-founder and CEO of Zapier, the most connected AI orchestration platform. Zapier connects teams, tools, AI, and agents across more than 8,000 apps to automate critical workflows with flexibility, scale, and control. And prior to Zapier, Wade worked in marketing at Veterans United and IdeaWorks. He has degrees in industrial engineering and business administration from the University of Missouri Columbia. So Wade, thrilled to have you here.

Yeah, thanks, Ron.

Yeah, so first, maybe introduce yourself a little more and tell us what you're working on. And of course, Uh, why we're all here.

Sure. So yeah, I'm Wade. I'm one of the co-founders, CEO at Zapier. Zapier is now helping over 300,000 customers automate their work, some 10 million+ users. It has been 14 years in the making. So to your point, it started as, you know, helping people with integration. You know, about 5 years in, we started working on workflow and helping people automate these, you know, more deterministic workflows end to end. And, you know, in the last couple of years, it's all been about agents and AI and how do you sort of mix, you know, probabilistic workflows into the, into the equation. So we are having fun helping folks do more with less, I guess, as the proverbial saying goes.

Yeah, exactly. And I think that's something that's been a repeating topic here that we've talked about. There are these deterministic tasks, the things that must happen in the exact same order every single time in a reliable way, which is sort of what we thought about traditionally, programmatic or automated or rules-driven. There's lots of lots of ways of thinking about it process-oriented. But of course, AI adds all that variability, you know, the things that required human cognition or creativity or variability, right? So let's talk a little bit about the automation component because I think people don't realize still how core a lot of those repeatable tasks are. So maybe for you, as you've seen, especially in small organizations, you know, teams that are under 20 or so, you know, how are you seeing them specifically with using automation to outperform organizations that may be significantly larger than them?

Yeah, it's, you know, we have a number of these companies in our customer base that have, let's call it, 20 or less employees that are doing—

you would—

you wouldn't be able to tell from the outside looking in. You would think this is like an army of people sort of getting this work done. Now, I think what separates these companies is, you know, they figured out how to operationalize automation for their core product and the core value prop they deliver. So yeah, I'm thinking, and usually it starts with maybe a founder or an owner-operator, or maybe like an early employee who's kind of like, you know, maybe they're like a project manager or a COO or operations lead, or like, you know, they can kind of have like a wide variety of these titles. And what they tend to do is they tend to take every single task that comes their way and they get really good at breaking it down and saying, okay, what if we actually automated this? So that is, that mindset is pretty different than I think, you know, certainly what you saw in like Silicon Valley in like the 2010s where it was an era of, you know, blitzscaling and like hire more people, figure it out later. And these folks don't take that approach. They say, hey, we're not going to hire till it hurts. We're going to put automation in through and through. You know, I'm thinking of one company who even in their core product line, every single prototype that goes out the door, it starts with the Zap. They're like, we're going to build a Zap for this every single time. And then that'll test like, hey, is this idea even a good one? And a number of those ideas scale on Zapier. Forever. Some of them, you know, they go rebuild in their own codebase, you know, at a certain scale after they've proven out its value. Some of them don't work, and then they're just happy that they didn't waste the time going and building it, you know, in a production way. And they've got this big old table where they're keeping track of all these automation experiences— experiments that they could go run. And so the end result is that they're scaling by just launching automations over and over and over again. Instead of, you know, hiring people at the end of the day. And so it really, it's almost hard to like, I need to come up with like a name for this type of organization because they sort of look and feel like an evolved species. It's not something I think most companies would recognize when looking at them because they do look and feel and operate quite differently compared to most.

Yeah, that's actually something that we've been struggling with here or grappling with in that sort of the existing terminology for entrepreneurship and for scaling doesn't really fit well with this new model. You know, that's why we started using the term micro team because, well, it's not really a small business. It's not really a startup in sort of the, the traditional venture-backed, you know, high hypergrowth mode. And, you know, it's not— it's definitely not like a large enterprise. It's entrepreneurship. So for now Micro Team seems to stick with a sort of an ideal that it's much smaller than expected. But this idea of scaling through the power of leverage basically is this idea of calling scale maxing, which, you know, comes from a bit of a, you know, I guess Gen Z term, you know, the looks maxing, grind maxing.

But it's like—

Yeah, yeah. And it's like, it's this idea of like, well, let's— how much can we scale? How much can we push the scale with, with the kind of the team that we got maximizing, squeezing every little bit of value out of it? So it's cool that you're thinking about that as well.

Yeah, it's— it is. You're right. Like, these folks are odd. Like, you look at them and, you know, they might have 20 folks in the org or something like that, but, you know, the revenue that they're pulling through, the customer bases they're pulling through, they're actually just much, much larger. And so— but they don't look like it when you actually go look on the inside. You're like, this doesn't— this doesn't look like a classic enterprise-scale company. But in many ways, they're They're punching in that weight class.

Yeah, it's interesting. And I think because of that, every member of the team that they do have is highly critical and highly valuable, which I like. Uh, it's sort of felt in sort of this Silicon Valley-esque, uh, you know, rapid growth, hire fast, and then I guess fire fast phases that people felt a little more dispensable, you know, where we felt replaceable. But when you have these micro teams where everybody is really squeezing the value, everybody really isn't. You know, that truly replaceable. And I like that, you know, that adds the value back to the human, interestingly enough.

Well, yeah, I think also it's more fun to work on these like smaller teams. Like everyone, you just know each other on a more intimate level. There's a deeper sense of ownership and a responsibility. Yeah, I think all of us have been on teams— well, maybe not all of us, many of us have been on teams at some point that just got— there's just too many cooks in the kitchen and there's this diffusion of responsibility. There's not enough work to really go around. And what happens is like kind of imaginary work starts to spin-up. It's like, it's not the most critical thing, but you know, I gotta go find something important to work on. And so you kind of go invent more things, whether that's actually going to help the organization or not. And so I do think like constraints, and one constraint is keeping the team small, like makes the, all the work you're doing is all impactful.

I like that. So let's talk a little bit about, about some of those, these patterns that some of these organizations might use. And you said one of these is, it's more like the skill of identifying the high-value repeatable task that they can optimize and squeeze. And maybe the first iterations or 2 or 3 aren't the best, but, but they're working quickly towards that. So this is this pattern maybe that you've seen, cause you've seen a lot of automations and you see a lot of things that work. So, you know, for those founders that are, that are drowning in these repeatable tasks, when you look at these high-performing users, maybe what patterns or systems do you see that others haven't quite yet figured out?

Sure. You know, there's a couple of ways that folks coach on identifying automation. And, you know, I think the most basic one, which does work, but, but I think it limits your upside, is to simply just look at your personal daily tasks. So you can go look at your calendar, for example, and just, you know, look over the course of the last couple of weeks, like, what are the things I do every single day, every single week? And which of these tasks, you know, might possibly have could be either partially automated or fully automated, and then you can start to systematically go automate those bits. So that's one way. Now that'll probably help you with your own sort of personal workflow and you'll improve that way. The second way is to then apply that for your team or your organization. Go look at the things that are tasks that, you know, maybe sit across individuals or teams and say like, hey, where are these handoffs that are happening and how do we actually go automate those end-to-end. And in both of these cases, you're automating work that you're kind of already doing. The next thing that I think is where you really see the big unlock is when folks are now starting to identify opportunities to automate work that you weren't doing before. These are when you're now like solving net new problems. This is a problem that like either you couldn't do it because like, ah, there's just no way we could hire that many people and do it reliably enough or what have you. Or it's just something at like your scale, like you feel like, ah, we just don't have the resources to go, to go tackle that. And this is where I think folks start to really get a bunch of leverage. This process is more like what you would think an entrepreneur goes through. It's being very observant of their environment, paying attention to where are there like customer problems or needs that are being unmet. Where are there novel ways to go solve that problem? And a lot of it is just about, you know, having lots of ideas and being very observant of the problems around you and then figuring out, hmm, you know, if I built a little thing like I could probably solve that problem in this, like the shortest distance possible. And, you know, it includes this set of automation or this tool in and around it. And, you know, I think the way in which I would coach folks on that is like, It's just like, get better at observing things, like get better at like paying attention to the things in your world that you're like, oh, like if, if only, you know, this thing sucks and if only we could do X or, you know, boy, wouldn't like it be incredible if we could solve this problem? But by golly, there's just no way to do it. And then instead of saying, by golly, there's just no way to do it, stepping back and saying, well, well, well, what if, what if we, what if we did this? And so anyway, it's that, that aspect is, I think, harder to coach and teach people on. But the teams that figure it out are the ones that I think have the biggest opportunities.

Yeah, that's interesting. It's kind of funny. A lot of these ideas come full circle. We were, you know, back in the day when we were doing integration tasks, we're talking about business process management, business process reengineering. Used to have to hire consulting firms to come in there and untangle your systems and figure things out. But a lot of what you're talking about is more of like, these are soft skills. These are things that people need to learn how to do. It's ways of thinking and You know, it's interesting. It's something you're definitely not taught this in school by any means. And maybe this is inherent talent. Maybe just some people are just more talented at being observant and sort of distilling these things. Maybe that's what separates the high performers from the lower ones. Who knows? But I think you're definitely on to something here. And this is sort of a question I've asked a number of technology leaders like yourselves, which is a related idea around planning. You know, the world is moving so fast. It's really hard to come up with, say, a 1-year plan, let alone a 2-year plan. A 2-year plan. And the funny thing is, of course, technology firms like yourself, we're just so used to moving quickly that, you know, we don't think in those long terms. But then you go to like an insurance company or a bank and they're talking about their 2-year plan. Like, wait a second, the world is going to be a completely different place. This kind of ties into the idea of identifying these opportunities for automation. Really, it's about squeezing this timeframe considerably down. So, you know, maybe, maybe from your perspective, you know, how do you even think about planning? You know, in this environment where even if you ask some of these technology leaders, what are you going to do in a year? And honestly, they're not so sure either.

So yeah, spot on.

So yeah, getting into that. So like, you know, AI is really trend— AI is adding to the automation piece. We haven't even really talked too much about AI here, but what did you— what capabilities did you once assume were too complicated to automate that are really now suddenly possible?

Yeah, I think the, the biggest thing we benefit from with AI and specifically large language models is we can work with unstructured data in a way that you couldn't before. You know, in the past, if you had like big blobs of text, if you wanted to categorize things, if you wanted to make decisions based off of that with reasoning and logic, I just like, wasn't really possible. Or if you like, if you really had to have a like fairly simplistic approach to even begin to attempt. And even still, you didn't get like great results out of that. Nowhere is this more obvious to me than meeting recorders. So we've seen the rise of meeting recorders in the last couple of years. They're everywhere now. And in the past, there was hardly anything that happened with meeting recorders other than like you'd get the transport back or the recording back. And now it was up to you to figure out, well, if you want to do anything with it, great, go do anything. But for most of us, I think like meetings were these ephemeral things. You'd sit down, you talk to people, maybe someone would take notes if you were particularly diligent. You had one or two action items, you went and followed up on those things. But beyond that, the discussion, the dialogue, any insights, they were kind of just left to human memory. But the reality is your, your meetings are chock full of like great information that are highly actionable. So here are like a handful of ideas that I have learned from our customers you can do with a great meeting. So one, you can use meetings to create coaching bots. So if you are, you know, a sales rep and you're getting your, your meetings are recorded all the time, you can have an LLM that's maybe trained on a particular sales methodology. And after every single meeting, you can get notes, you know, did you do good discovery? Did you do, did you, you know, do a good demo? Did you do a da da da? And it can coach you on that stuff. Infinitely patient, available for every single one. Fantastic. Similarly, for internal meetings, you can do something similar. So like there's this great book, The Five Dysfunctions of a Team. You know, I run that on all my meetings to be like, hey, like, how did you do against like the Five Dysfunctions of a Team? Coming back to that sales call, well, similarly, if you're talking to a customer, one of the things you can generate the other side to perhaps automatically generate a case study, you know, if they're talking about why they value the product, this, that, or the other, you could again have an LLM say, hey, structure this conversation in the form of our case study template. It's a great way to get stuff out there. If the customer talks about your problem or solution in a new way, maybe they're using particular keywords, new vocabulary, new vernacular. You can automatically go generate a landing page that is using those keywords and then you could generate ads against those on Google or Facebook or what have you. So as you can see, like there's all this just context that sort of sits inside of there that in the past pre-LLM just kind of would get lost to time. Most people just wouldn't do much with that. And maybe if you— and if you did, it was like you were just like you were probably, you know, top 1th, like 1% of the 1% in sort of like operational efficiency. But I find that most of us are just too lazy to like actually be that efficient with it. And so that to me is where like the most exciting new use cases are, are being able to go work with this unstructured data and do and create interesting things from it.

That's really interesting. Yeah, I still think even now, making effective use of these recorded meetings will still probably put you in the top 1% for right now.

But I do think that is going to change as people start to wake up and go like, hmm, there's more that can be done with this information.

That's good. Well, good. I'm hopeful because, you know, you want meetings to be productive. So, you know, outcomes like that make, make it more than just we all got together and said things and Nobody can remember the next week what we just said. So I'm hopeful as well. So let's talk a little more about, about sort of these, these tiny teams that are punching above their weights. And you know, what, what do you think in terms of a capability that you, that from your perspective, every micro team should adopt within the next, you know, 12 to 18 months? It's hard to think that far, but that you think maybe very few people are really doing today, but again, you see some of these outliers and you say, oh, these folks are doing something really interesting and unique here.

Yeah, well, you know, I think one, meeting recording, like meeting recording has so much interesting exhaust that you can do interesting stuff with it. So I would start there. Yeah, I think second thing is your customer communications. So, you know, if you have like an inbox, you know, that you're doing customer support or sales discussions, etc., almost all of these have some amount of automation that could, can, can happen with it at minimum. You should be like tagging, categorizing, triaging those emails that are coming in. At maximum, you can perhaps start answering some of these directly with confidence within, with an agent. And, you know, I've seen most teams are able to get somewhere at least to like 50% are able to respond to. I've heard some orgs are up to even 75% depending on like the complexity of the queries that you're getting from your customers. So that's, hugely common amongst teams. You can obviously run this on your personal inbox too if you get a lot of emails. I have one running on mine. On customer acquisition side, almost everybody is running— or I say everyone— the most sophisticated teams are doing some sort of like internet research and like lead qualification and lead scoring now in a more sophisticated way. So say you get like a typical, you know, contact information, so name, email address, and oftentimes they'll send that out, run a like LLM to do like deep research on the account and say like, hey, tell me all this stuff around it. Then you can hit an LLM that says, hey, I want you to score this lead based on these characteristics, you know, good, better, best, or bad, or, you know, bad, good. And then, you know, based on the outcome of this, I want you to, you know, add it to a CRM or I want you to alert the sales rep, or maybe I want you to generate a follow-up email or a follow-up text message to go out to the account. So that's another one that's super common. Seeing a lot of folks start to do, you know, like kind of basic internal search. So you can, you know, hook all your tools up via like MCP and other things like that. And you can build internal chatbots really easily now to answer questions that you might have. You know, almost a naming convention we're seeing amongst our customer base is, you know, if they have— if you use Slack, you'll have like an Ask Data or Ask HR or Ask Sales. Because almost every team has this like Slack channel where, you know, people just come in and are like asking teammates like basic questions like, where is this? Where is that? Help me find this. Help me find that. That don't actually require like a human to be answering those things. These are just typical like search or generative search type queries. And many of these things can and should be answered by like a chatbot or like a chat agent at the end of the day. So there's like a handful that are exceptionally common. I know I'm leaving out many, many, many other use cases, but I think the point here is that no matter the organization you're in, there's so much opportunity to be using and investing in these capabilities.

Yeah, I think that's great. And it feels common. And the funny thing is it isn't yet common within most organizations, but I think this is the expectation that these things should be common. So I think it's good to mention it.

Yeah, spot on. I mean, we're at the beginning of that, like, you know, early adopter curve where there aren't playbooks written. So there's so much discovery and invention that is happening at the moment. And this is a big reason why I encourage teams to do show and tell around this stuff internally. It's not enough that folks are, you know, paying attention to LinkedIn or X or YouTube and trying to learn externally. The internal show and tell is so valuable because you're getting to see what your teammates do in the context of your own company's data. And that part is so valuable because it's, it's just much more like relevant to, to, to the team. And so doing that kind of like internal show and tell can really help get people's wheels turning a lot faster.

And I do love a good show and tell. A lot of our listeners know I used to run these demo events called Tech Breakfast, and it's just like there's nothing better than not having PowerPoint and actually showing something. Oh, it's fantastic.

I know.

I love that. So let's talk a little more about this. You know, one of the things that people are trying to do as part of the sharing, this collaboration, is there's a lot of sharing of templates templates and of workflows, and there's a lot of demand for it. Man, I see this all the time on LinkedIn. Like, I did this thing, now send— message me a comment and I'll DM you the template. It's fantastic. I don't really know how many of them actually respond, but it's fantastic. So, so there's a little bit of a— there's just a lot of hunger for this right now. There's a sort of a tension, if you will, some— sometimes between building your own workflow and trying to get prebuilt ones. And so how do you see that evolving, you know, especially with AI? Do you imagine a world where perhaps maybe workflows are self-configured, that there's like marketplaces that really optimize for, you know, template exchange, that sort of thing?

Well, it's a good question. I, you know, I think we're still trying to figure out this exactly. To your point, I don't think the current way templates are generally done by most folks are actually all that impactful. I think partly it's because the context of how I do it it might inspire me at like a macro level, but at a micro level, like the implementation is just very different depending on like the organization, right? And so I suspect what has to happen here is, you know, we're— we probably want to use templates to focus on the inspiration. Like that's the job to be done here is to get people excited, but it's not there to actually be like copy and use it exactly as the way this other team does it. Instead, it gets fed into you know, like a Copilot or an AI assistant or something like that, that helps you go build a version of it for you where it's like, hey, can you take this template, this idea, this concept, and adapt it to my organization? Ask me questions about how my— like, what stack do I have? What problems do I have along this? And then help me configure it in the way that I want. And I think that's— there's something— there's a missing link somewhere in there today for— based on the way most of these like template experiences work.

Yeah, there's a lot of demand. That's exactly what people want too. They want the, they want the general flow, but they want it to work for them, for their own infrastructure, their own data. So yeah, I think innovation here is a great demand.

Yeah. And if you've, I mean, if you've tried any of these templates, you probably, that's probably the first reaction you had is like, well, this looks great. You know, Ron, you built an amazing template for your podcasting setup, but my podcast works a little different. You know, I don't, I don't do mine on Zoom. I do mine on Riverside. You know, I, I like to do clips for TikTok and you like to do them for LinkedIn. You know, it's just like there's all these like, okay, yes, it's sort of the same. Mostly like we're doing a very similar process, but there's enough differences that I couldn't just like, like just carbon copy your exact flow and expect it to work perfectly for me.

Yeah. Wade's spitting truths here.

Exactly.

That's exactly what it's like. Zoom and LinkedIn. That's exactly right. And for you, probably Riverside and TikTok. Yeah, it's where the audience is. It's where you're your infrastructure is, but you're exactly right. You're exactly right. So one of the things you did was you went public with your own 2026 AI transformation strategy, which is a rare move. We're starting to see more of this. You know, OpenAI punched their little red button of disaster. I'm saying it that way because I saw somebody did this generated AI video of like Sam Altman actually hitting like some fire alarm. And it was, it was kind of a funny video, but you know, why did you decide not only to come up with it, but but share it in a very broad, broad way. Radical transparency in this open source style approach. Yeah. For your own internal transformation roadmap. And, you know, what impact do you think it might have? Not just obviously for yourselves, but you're— there's an idea here that it would impact others and the broader ecosystem.

We— so for those who missed, like, we shared our like 2026 AI transformation strategy externally. So today Zapier has That basically everyone internally is using AI daily as a driver. And we're trying to figure out, okay, how do we get even more impact from this? And one of the things we felt like we could do is share publicly how we're going about it. And in doing so, one, hopefully it's helpful to others. People can look at it and say, oh, that's interesting. Maybe Zapier has figured out a thing or two that we haven't and we'll go try those. But also the reverse is important too. Maybe we can learn from other people by putting it out there. The internet is usually not shy about telling you that you did something wrong or you could be smarter or better at a thing. And so by putting it out there, it actually, you know, helps us get better at our own AI transformation approach. And I think this is really helpful, particularly at this moment in time, because we're at the beginning of that early adopter curve. There's no playbooks written. There's no, you know, here is the best practice, the correct way in which to go about these things. And so the only way folks learn is from this you know, storytelling, exchanging notes, exchanging ideas. And so a big part of putting it out there was, was simply that. And, you know, I'd be lying if the third motivation wasn't, you know, we sell this like our product helps people with automation and change, etc. And I felt like, well, shoot, if we're just talking the talk, like, that's not that impressive. We need to figure out how to walk the walk as well too. So let's go show people how we are actually doing this. And as a result, you know, I imagine a number of folks will look at our product and say, Well, Zapier used these tools to go do it. Like, let's go give it a go ourselves as well too.

Yeah, that makes sense. Part of that show and tell, but in a different way.

Yep, 100%. It's just like, you know, we don't have that many ideas, but when we have a good one, we just kind of rinse and rinse, wash and repeat.

Yeah. Well, I think that's, that's part of what people are really interested in. I loved it. And for our listeners, I'm going to, I'm going to find that, that transformation roadmap. I'll link it in the show notes so you guys can read it and follow up more on and take a little more detail, no matter what size organization you are and kind of where you are on this journey. But one of the great things I loved about it, because the sort of the punchline, if you will, of the whole transformation roadmap was this idea of moving from just adopting AI, using it maybe in a casual way or in an augmentative way to like this transformative approach, which ties this thing we talked about earlier about these micro teams that are punching above their weight. There was something different that they were doing. I think it ties this idea, well, they transformed something. They transformed the nature of their operation. They weren't just thinking, oh, I'm going to do this the standard way, but just make it faster or whatever, squeeze some cost out of it. And I think this is one of those mindset things, right? So micro teams often, they get stuck sometimes in this adoption phase. And for the ones that are truly transformative, they're crossing that bridge into transformation, and they're operating with that 10x leverage rather than 10% improvement. So from your perspective, like, you know, when you see the light bulb going off, maybe you've seen that in your own organization. You know, what are like, what is this, this mindset shift perhaps change that, that they need to make to say, oh, I'm not just going to use this tool, I'm going to change the way I do things?

Yeah. So, you know, I think what I have seen at Zapier and what I have seen from our customers that are moving the fastest on this AI transformation tends to fall— have like a couple milestones that are, that are particularly important. And before you're able to transform the organization, you have to transform the people in the organization. And that means at the individual level, not at the team level, not at the company level, etc. And this is where hackathons, show and tell, have been so useful because you can start by saying, hey, I'm a CEO and here's my memo. We're going to be AI first. Let's go.

Rah, rah, rah, rah, rah.

But if you stop there, the company is going to go, great, Wade, love to hear it. I got goals I got to go hit. Let me go hit those goals. And maybe if I can do some AI stuff on the side, maybe I'll dabble a little. But for most companies, that's not enough. Like, you have to you, you actually have to like create space and time for people to put their hands on the keyboard and say, we're going to actually learn how this technology works and we're going to figure out the use cases that might be impactful. And so those hackathons, those show and tells, those create the individual learning moments where people can go from not having used the technology to now figuring out, oh, here's a bunch of ways that I individually can use it to be better at my specific work. I call that raising the floor inside the organization. But the trick is, even though you've raised the floor, it doesn't mean that you've actually transformed the company. You know, oftentimes you're still just using it as a personal assistant to do the job you've always been doing, just better. But the real opportunity with Zapier isn't to just do the job you've always been doing, but better, but it's to do entirely new types of jobs, entire new types of problems to solve them in unique ways. And so for us, we felt like, okay, we've got everybody in the company excited about AI, using AI, they have it as an assistant. We feel great about that. But while the floor has been raised, we need to go raise the ceiling now. And that's where we've been focusing on saying, okay, now's time to go look across teams, across departments, across leaders and say, what have these like workflows actually need to be torn down and rebuilt again in this new way? What problems have we been avoiding or simply chosen not to invest in because we couldn't solve it? It wasn't economically useful to do it pre-AI, but now it's actually possible. And those types of problems often require more than just hackathons and show and tells to inspire people. It requires real change. It requires real investment in new types of tooling, new types of jobs, new ways of working. That it, you know, it's not the type of thing that like I could just show up and say, okay, today I'm just going to totally fix this. Like you have to put some elbow grease into it to get it working in a new way. And so that's where I think, you know, it's really valuable to kind of think through and break down what does it actually mean to change your organization and how do you want it to look differently? Because if you don't do that, I find the companies that struggle often say, we want to just be different with AI. Well, different how? What does that mean? How do you want to work differently? What problems are you actually going to solve? And if you don't kind of step back and think through those different milestones, those different stages, Yeah, I think you can find yourself confused and frustrated and ultimately not moving and benefiting from the technology as much as you could be.

Yeah, a lot of times it does come down to change management.

I mean, honestly.

Yeah. And humans, you know, the funny thing is we're not that good at change. It's interesting.

Yeah, I— it's, you know, what's— there's that old meme where it's like, who wants change? And everybody's hand goes up. Who wants to change? Everyone's hand stays down. And I had some truth to that.

Yeah. You know, one of the things that you're doing there to help with that, because that's, you know, that's, that's human psychology and that's behavior and that's motivation and things like that. But if you can't grow, if you don't change, right? You can't— there's, there's no way to, there's no way to do that. And I think one of the things you're doing is you're developing a digital twin at Zapier and AI, this thing called AI-powered value engineering. So this may be stuff that our micro team's never really heard of. So maybe you could talk a little bit about that and how you see these AI agents, you know, helping becoming teammates. And I guess there's some near future vision perhaps for what a Zapier agent may be.

Yeah. Well, so I think a good way to think about this is let's consider your— a traditional, like, support organization. So inside of a support organization, let's say you've got, you know, 50 people that do customer support. And, you know, a customer might come in with a question about a particular thing. And, you know, if you looked across the entire support organization, you'll have an answer to that question ready, eager to go. But maybe only 10 of those folks of the 50 in your support organization actually know it at that given time. Well, that kind of stinks if the customer ends up talking to one of the 40. But this might be true of almost any question that might come in. It's like, yes, the comp— like, is the— I get this question from new people joining the company all the time. It's like, do we have data tracking this? Like, well, yes, we do. But that doesn't mean we're using it. That doesn't mean we're benefiting from it. Just the simple fact of like having it is not usually enough. So what the digital twin exercise is doing is trying to just systematically Go catalog all the things that we actually know and code those in a knowledge base somewhere so that when somebody asks a question, you don't have to ask the, you know, you don't have to track down one of those 10 people who knows the answer to that thing because that knowledge has already been encoded in one spot. So these digital twin exercises are really just trying to take the hive mind, the like the storytelling out of our heads and get those documented really, really well. And you might think, you know, oh, that sounds impossible. Like, that's like a huge exercise. Is everyone just supposed to sit down and like blab all the things they know? The answer is like, well, no, not quite. The job of these digital twins is actually just grow them as you need them. You know, you don't need to sit down on day one and have every last bit of knowledge encoded. But what you can do is say, well, you know, every time we answer a question of a customer, Let's go check out our digital twin and see, did the digital twin already have the answer to that? If not, great, let's add it to the digital twin. Now they know too. And then if you just make a habit of that where you're just logging the answers to these things in all the places where you're already doing the work, over time you're gonna have an AI that knows a lot more than any one given person knows. And so that's how I think these digital twin exercises can be really effective for organizations. If they take the time to encode the, like, the knowledge that they, that the company already knows, but it's just not evenly distributed across everybody.

Yeah, it's going to sound familiar. A lot of people worked in this knowledge management area, so they didn't maybe—

Exact same concept.

Yeah, yeah, yeah. And this a long time ago, we tried to do that, like, with wikis, and like, you put a wiki up and you try to do it, but like, you know, you know what happened. It's that it would go up there and people would add to it, then over time it would decay and have old information that was incorrect and nobody would update it. Anymore, so.

Yeah, I mean, it's spot on. And I think that's a great example of, you know, where AI makes some of these old ideas new again, where it's like, huh, that old idea, it was the right idea. It was just, we didn't have like great tools to really get the best out of it. But with an LLM, you can do a better job of detecting stale information, keeping it up to date, actually like logging some of this stuff. Because you don't, now you don't actually have to like rely on humans to be like, okay, I got to answer the question. Answer the question to the customer, then I got to go over to the wiki and log it over there instead too. Instead, you can build automation where that second part just happens as part of the process in the background and then just have one person like, you know, helping like audit and screen it on an ongoing basis. And so I think there's just tons of opportunity like that where you can go look at some of these old ideas you probably had in your company and say, hmm, that thing that we decided didn't work or was too hard or was too expensive, etc. Could we actually do it again? And the answer might surprise you. You might be like, actually, it's kind of crazy easy now.

Yeah, that's a whole big idea in and of itself. It's kind of interesting. We're talking about this. I'm like, wait a second, that's like a billion-dollar idea in and of itself, just having some sort of real corporate or organizational knowledge base that is always up to date and relevant and available and all sorts of good stuff. So that's a big idea. So one last bit here before we get into our rapid-fire takeaways, we always have 3 rapid-fires, is again, you know, from, from seeing AI and automation in daily real-world life, right? These are not pilots, these are real-world implementations, and especially trying to push it internally as well. So it's kind of interesting, you're, you're applying the technology to yourself as a proof point that this really can achieve transformation as we're talking about. You know, maybe what's one thing that some observations that these micro teams that can adopt, the ones that want to be highly efficient for their own operations, you know, as they're trying to run their organizations, maybe what are some of these behavioral or process changes that they should, that they should put into place, maybe even day one before they start?

Yeah, well, I think, I think the biggest thing you've got to do is you've got to put some constraints on yourself. Change takes effort. And if you keep doing things the way you've always been doing them, you're not actually going to get the benefit of this. And so it's really helpful to step back and say, where are the right places to put constraints in? And the common ones that I hear are things like, we're going to pause headcount, like no new hiring can happen, or we're going to pause tool spend. You can't buy new software. And, or at least you have to like, the threshold to like getting new hires or getting new software spend is much higher. And you know, the person that approves those, you put somebody who is both one, high judgment in approving those things, but also two, someone who is skilled at understanding automation, AI, and how to deploy this stuff. And now that person can say, hmm, you actually could solve this. With tools and agents and so on and things like that versus hiring or software. So I think, you know, you want to put some of these like simple constraints and mechanisms on yourself to force yourself to do the uncomfortable act of reinventing the way in which you work, your team works, your organization works. So that's probably one of the more important rituals. I think, you know, if that's the stick, The carrot is things like hackathons, show and tell. It's the more exciting things of seeing, whoa, that was incredible. Like how, like this old process I was doing, like that was taking me multiple, you know, many hours a day or like, you know, tens of hours a week now just got turned into an agent that does this automatically. And I only had to spend an hour building it. Like those types of like jaw-dropping moments, the faster you can get yourself or your teammates to go experience those, you start to realize, I actually want to work that way. Like, you don't have to convince me now. Like, I would prefer to do that than to go like try and hire somebody or try and go buy more software. Like, it is clearly way better to work in this other way. And so I think you kind of want to mix those, you know, some carrots, some sticks to like, you know, push yourself in a direction that you want. You know, you naturally want to go, but, you know, you just need a little, little, little, little, little nudging. It's like trying to build a habit of going to the gym. It's like, okay, some days I just don't want to do it, but I need a little bit of a kick in the pants to get there.

Yeah, I love it. It's actually great timing. People are thinking about their New Year's resolutions.

Exactly.

Sometimes like, I want to, you know, have this, this goal here. I was like, well, it's going to require change.

You want to do it.

A little stick, a little carrot. This is fantastic. We'll have a couple rapid-fire questions here. We always ask the same 3 questions to everybody, whether they're micro teams or technology enablers or influencers or anyone that's sort of in this space. So the first question has to do with tools. I think I may know where you're going to go with this one, which is what's one AI tool or automation you absolutely cannot run your business without? The one that gives your team the 10x leverage every single week.

All right, I'll give you, I'll give you one Zapier answer and one non-Zapier answer. So Zapier is like the, of course, answer for me, but a specific use case, one that I've had running for ages, is I have a simple Zap to help me manage my to-do list. And we run our team inside of Slack. Slack gets pretty noisy across many different channels, and I found myself losing messages, forgetting to respond to teammates and things like that. So, you know, when I first set it up, it was a simple Zap that would take a, a saved message and I would have it shoot me an email and I would just manage it out of my email inbox. Now, as the years have gone on, I've evolved this over time. This is another key thing with automation. The first thing you build doesn't have to be the grandest. You can build a basic thing and then make it better for yourself over time. This is a good example of that where now today still is the same starter, which is, hey, save this message in Slack. But now what happens is That saved message goes and hits another step that extracts the whole thread from Slack. Then it hits an LLM step that says, hey, go summarize everything going on in here into a short one-sentence description of what the task is, and then make sure the action item is spelled out clearly for what Wade needs to do. And then it goes and lands that in a shared Coda inbox that I share with my assistant. And then my assistant helps me go triage a lot of that stuff so she can tell me You know, after a busy day, I can plop into that to-do list and say, here are the 3 things you need to do before you log off today. And then here's the next 10 things that you need to get to sometime this week. And so it's a real— like, it's one of these simple systems that I've been using for, for years now that helps me stay on top of my day.

Oh, I like that.

Now, the non-Zapier-related one is I use Granola a lot these days. Now I do use it with Zapier, so I'm sneaking it in. But Granola itself is very valuable even without it. So Granola is a note-taking tool that helps you in all your meetings. And so I have it turned on for most of my meetings these days and it keeps log of it, which is helpful like any meeting note taker. But what I really value from Granola is it has this feature called Ask Granola and it allows you to ask questions and then it will iterate over top of all of your meeting notes and you can ask it all sorts of interesting questions. So you can say, hey, what's you know, what topic should I discuss in our upcoming weekly exec hangout? And so if it's been in all my meetings from the last week, it might know, okay, here's 3 things that you need to bring up in the exec team weekend. What topics might be good for sharing on social? So, you know, helps me ideate and come up with ideas for things to share on social. What are 3 things that I'm doing great as a CEO? What are 3 things that I'm doing terrible as a CEO? And so it can coach me on the meetings that I'm showing up. And so the idea of being able to ask questions and ask for feedback across a large corpus of meeting data has increasingly become very valuable because it has the context of most of my day that I've been working on. So those, those are two for you.

That's great. Well, it's like, you know, I'll put links to— of course I'll put links to Zapier and Granola in the show notes. That's fantastic. So the second question we usually ask is about process. So what one system or process what should every micro team have in place by the time they hit $1 million in revenue that separates what I call the grinders from the scalers? You know, what would you tell them maybe to stop doing, or what should they start doing today? Hmm.

Once you hit a million in scale, what one system? I think once you get to that scale, you need to start work on encoding the knowledge of how you operate. This kind of comes back to the digital twin work. I think this is really valuable today because it's the type of work that's going to help you onboard agents into the company and humans into the company. Both are going to benefit enormously from having well-documented encoding knowledge of what has worked to get you to a million. What are the ways? What are the principles by which you want to operate? What are the standing operating procedures to do key and critical tasks? So it's not that you need to have every system and process in your organization well documented. That part is something that you're gonna have to always be working on. But I think by a million, you wanna start thinking through, how do I actually operationalize this stuff instead of, you know, just waking up and, you know, white knuckling your way through the day.

Yeah, that's fantastic. One of actually the scale maxing laws, which is you should, you should, you can't scale chaos. So, you know, having processes in place really helps because that's how you can continue the momentum. Fantastic backup on that. Final lightning round question is just a general tip. So if you could give one piece of advice to a founder who may be running a team the size of a dinner table, a dinner table size team right now who wants to mega scale, especially from your background, you know, you've built, you've built it up from, from zero. So, you know, what advice would you have for them?

Well, if you're able to run a team that fits around a dinner table, I think you are at just a perfect team size. Like, there's just nothing better than that scale of team because you have— everyone feels a deep sense of ownership. Everyone appreciates each other on a personal level. Everyone respects the work that each other is doing. There's no social loafing. There's no hiding. It is just a really great spot to be in from a team size. Now, if the goal is like, hey, I want to scale, you know, you are going to be adding more humans to the mix. Just no ifs, ands, or buts about it. If you want to do big things, it takes humans. I know we talk a lot about automation, but I still am like a huge believer in like the powers of humans to help scale to bigger and bigger things. I think one of the massive things you can benefit from is instituting a rule that's don't hire till it hurts. Make yourself understand why you're bringing this person on, what value they're going to bring to the team, what skill sets they need to bring to the team. And only then are you going to go in and say, great, this is the right time to bring somebody in. And that will make sure that every incremental person you bring in is going to punch above their weight class.

Well, this has been fantastic. I really want to thank Wade Foster. You were amazing for this podcast, provided some fantastic insights. So we have a mantra for you for the week, and that is real transformation takes change, and change is harder than buying a new tool, but you can't get the breakthrough you want without the change. So invest in yourself as much as the tools. So if you've loved this interview with Wade Foster, just hit that subscribe button, leave a review, share it with a small team founder friend who still thinks that 10% automation improvement will have the transformational impact they're seeking. Thank you for listening to the Exponential Scale Podcast. Think big, stay lean, and scale smarter. And we'll catch you on the next podcast. Micro-teams aren't the future. They're the now. If this episode got you fired up, hit that subscribe button, share it with your crew, and join the movement over at scalebrate.com. That's S-C-A-L-E-B-R-A-T-E dot com. Till next time, think big, stay lean, and scale smarter. Because you're not a tiny team, you're a micro team with mega potential. So get exponential, and thanks for tuning in to the Exponential Scale Podcast.