Michael Grinich. WorkOS.

Michael Grinich: If you go pitch an idea for a startup and everybody thinks it's a good idea, it's actually going to be a bad company to build because there's too much consensus. As a founder, your job is to go and incubate new functions. And then right at the moment you start getting good at it is the moment you hire somebody. It's really the only role in the company where you consistently are allowed to be bad at whatever it is you're doing. If you're the dinosaur, you don't know you're the dinosaur. What can we decide by tomorrow? What could we build by like Wednesday or Thursday? And so the cadence that we actually operate on, the week, is not the shortest timeline. It's actually the biggest timeline in a sense.

Nakul: But is there a cost to moving that fast? What do you give up? Michael Grinich is building the company powering the enterprise adoption of AI. Nearly any AI company you can think of, OpenAI, xAI, Anthropic, Sierra, Cursor, the list is endless, is a customer of WorkOS, but I'm especially excited to have him here today because of our personal relationship. The dream of a VC is to partner with someone extremely talented and ambitious and be on a long journey with them through good days and tough ones. My journey with Michael is one of those. He's unbelievably smart, ambitious, tenacious, urgent while being calm at the same time. Working with Michael has made me better at my job. With WorkOS, Michael has built an accelerating 50 million ARR business with no CRO, no VP of sales, no marketing department, just three sales reps. It's a company built by developers for developers. We'll dive into all this and more today. Michael, welcome to the show. So, let's start at the beginning of WorkOS. Before WorkOS existed, you were an application layer builder, building an email client, and were pretty design obsessed. How does one go from a design-obsessed application layer builder to a middleware product builder for a product that you yourself have described as most would think to be boring?

Michael Grinich: Yeah, WorkOS isn't exactly the thing that kids wake up in the morning excited to build. You want to be an astronaut, you want to be a firefighter, you want to go build infrastructure for enterprise SaaS companies. It came from that experience. So, I had started this company to try to change the way people communicate. I had been obsessed with email for years. And back when I was an undergrad at MIT, everything went through email. This is before Slack. So, it was the tool that everybody used to do their job. And so started a company to change how people communicate. Built this product, and I was the quintessential engineer product-minded guy. Building a product that was for me, I had these sketchbooks full of ideas and ways to change how people interact with their mailbox and do it on teams and collaborate. So, raised much money, hired a bunch of great people, built this product that we just obsessively crafted. I've told this story before, but I spent a couple days in a studio of an electronic musician making the notification sound because you hear that all the time for email. And so spent all this time on the product experience, launched it, and then people loved it because the product was really great. But then when we tried to actually scale it to customers, to actually businesses, was the first time I started talking to procurement and security and IT and all these people that end up being the ones to buy software. My friend now, Ilan Frank, who did this at Slack, he was the enterprise product lead at Slack. He calls this the difference between the user and the chooser. It's these two different worlds. And because we hadn't built anything for the IT person, the chooser, we essentially didn't have any of the enterprise capabilities needed to go sell to these companies. And it meant that they shut us down. They said, "If you're touching email, we have to get rid of this." It was a huge blow to the growth of the product. That was very frustrating. After spending years pouring my heart into the design of this thing to essentially have it not succeed commercially because of this set of missing things. And that sent me on this journey to go learn what does it take to become enterprise ready and to grow up market. And WorkOS as a company is actually just a reaction to that, to help other companies solve these problems. We've been doing it for a while. Like I mentioned, the company's seven years old now, but now what we're finding is it's this amazing fit for AI products that are all going up market and growing extremely fast. So, we're a pre-AI company, pre-AI infrastructure company that's now been accelerated based on the AI wave.

Nakul: Walk me through that year. So, you took a year off between Nylas and WorkOS and you had this notebook of ideas. I remember you'd mentioned deepfake videos and tech to de-radicalize people, all of that. So you had this, but also you were probably processing what happened in your last startup and what do you want your next startup to look like? What was that year like? What was going through your head? Was it primarily focused on what's next or some philosophical things also changed for you as a founder?

Michael Grinich: Yeah, I wasn't even sure I wanted to start another company, to be honest. I spent a couple months in Thailand doing Muay Thai and wandering around. I got really into aerospace and so I went and saw all the space shuttles that still exist and went to a rocket launch and just trying to follow my nose. And I thought about joining a company or a growing startup or even going into VC. I had some friends who were interested in hiring me to do that. But this problem just kind of stuck with me. I mean, of all the different ideas I had, I kind of was still bothered by this thing where I had felt like we had done all the right things to build the product, but ultimately it wasn't successful. And when I saw the convergence of this space, I knew if I was going to do another startup, I wanted the market to be enormous. I think there's no way to outrun a small TAM. You just get stuck. And now that I do a bunch of small angel investing, I see this all the time. Small TAM kills companies. So I was like, it needs to be an enormous TAM. I wanted to find an area that I could actually sell to developers. And the reason why is we had built some developer stuff at Nylas, what the company eventually turned into. And I just really liked it. I love building APIs. I mean, what engineer doesn't love building for yourself? And so that was extremely compelling. And the consumption-based business model behind it, I had seen products like Twilio, what they had done and that business model slow to start and then accelerates on the backside. And then I wanted to find an industry that hadn't had a market defining moment. And my gold star there, kind of North Star, was Stripe. So before Stripe, FinTech was not an exciting thing to work on. It was not exciting from a lot of different dimensions. It certainly wasn't appealing to go build. And post-Stripe, it's very exciting. There was Robinhood and Coinbase and all these other things that were built in that way too. And so, I wanted to find a market that had all those characteristics. And funny enough, IT had that. It was very stale. Even the word IT doesn't really mean anything, and it hadn't had the Stripe moment as part of it.

Nakul: One of your criteria is that the idea should look bad. Can you double click on that? Why should it look bad?

Michael Grinich: I think when I was intersecting these ideas, like giant market, hasn't had a transformational moment before. And then the third one is I think it needs to come from some personal insight that you have had. You have to have some unique information. If you're just sitting around a table trying to come up with ideas and all you have available to you is the internet, so does everybody else. And so, you're not going to have any alpha in that. You have to find something where you have a unique insight. And the best place to draw from that is your personal experience. So, whatever you've had. And I had that with the email client with the enterprise-ready set. So, you do all these things. I think actually if you go pitch an idea for a startup and everybody thinks it's a good idea, it's actually going to be a bad company to build because there's too much consensus. And what that means is there'll be too many competitors early on, and you won't have enough singularity to build the organization around. It's sort of like, I have this lemon tree in my backyard. And if you don't go prune some of the buds, you won't get a full size lemon because you get like 10 tiny lemons and they all kind of taste bad. You need one big lemon with all the juice in it. And the way to do that is to discourage competitors, to have something that everybody else thinks is a bad idea, but because you have this unique insight, you see it as a good idea.

Nakul: Actually, let's double click on that. How do you know though, because sometimes a bad idea is actually a bad idea. So, how do you discern between a bunch of these so-called smart VCs telling you, "Hey, this is not a good idea." How do you have the confidence to say, "No, I see something that others don't"?

Michael Grinich: I think ultimately that's a judgment call. It has to be based on your own personal faith and conviction. Good ideas, the best ideas always start off looking as bad ideas. I think Paul Graham wrote a whole essay on this. Also, bad ideas start off looking as bad ideas and they stay bad ideas. And so you have to have this conviction as to why your idea is actually not bad. And I think the way to break it down is, what are the beliefs you have or the things that need to be true in order for that bad idea to actually be a good idea? What are the assumptions underlying it that will allow it to grow? And for us, I think at WorkOS, it was that the incumbents wouldn't build this as a product, that we were going to go after it with more of a developer-led angle versus an IT-led angle. And that we thought the explosion in SaaS and new applications was going to continue to grow. The market was going to get much, much bigger. You want to build in growth markets. So, I think founders should be able to articulate that, but ideally a good way to do it is to base it on personal insight. I even had this experience with the WorkOS idea. I just thought somebody else would build it. It was so obvious to me as a need that I spent a while looking around being like, "Certainly someone else must be working on this." And I couldn't find anybody. And I was like, "Well, I should just do it then."

Nakul: And then you also have this concept of minimum awesome product versus an MVP. That is a more well-known term. Can you talk about that as to what you mean by a minimum awesome product, why it's important? Is MVP a bad idea?

Michael Grinich: The reflection of the minimum viable product, which is like ship early and ship often and get stuff out and get feedback from customers. I think the value of the MVP, the minimum viable product, is to get feedback as soon as possible for engineers that otherwise might be in their heads and stuck building stuff. And it's generally the right sentiment. But the counterpoint here is that people are really busy in the world and you really only get like one shot. Especially with developers, their first experience with your product ends up becoming a defining brand moment for you as an organization. You don't have that many of them. And so for us, when we build product, we actually try to make something that's notably impressive. It doesn't mean it needs to be done, like feature complete everywhere, it's just a certain part of it has to be impressive. It has to be different and has to be unique. And I think if we can't discover that or articulate it, we usually don't understand the market well enough. We don't understand the product well enough. What we're going to build is not going to be unique enough that maybe we shouldn't build it. I think we're only really in the business of doing things that other people aren't doing versus just following along.

Nakul: So, one of the things I really want to dive into with you is your org design philosophy, because it's pretty unique for a company of your scale. So, WorkOS now has about 100 employees, only one PM, a very small design team. More than half of the people are engineers and really engineers are responsible from product design to launch messaging. Correct me wherever I'm wrong. There's no VP of sales, there's no CRO. What's the org design philosophy here? It does seem like it's a company run by senior ICs and leads, not VPs.

Michael Grinich: Yeah. There's no head of engineering today or head of sales or head of marketing. We just crossed 100 people a little over a month ago and about two thirds of the team is engineering and design and product in some way. I think it's been a reaction to going and hiring a lot of these people and then finding actually the biggest impact on what drives stuff forward is really high agency, super capable ICs that can actually do things end to end. So, today we have five engineering managers that report to me that are driving forward sometimes multiple product areas with a bunch of engineers working with them on those functions. I spend time directly working with engineers. I meet with almost every engineering team weekly. I meet with our very small marketing team doing content marketing and events and product marketing. But a lot of it's done by people hands-on. Not a lot of meetings about strategy or performative—

Nakul: How many direct reports do you have then?

Michael Grinich: It's a good question. I'm actually not sure how many. Probably a couple dozen or so right now.

Nakul: And that doesn't spread you thin?

Michael Grinich: No, because people that are really experienced and really senior don't need that much management. They need alignment. They need information. They need context to be able to make decisions quickly. They need fast communication, but they don't need management in the sense of heavy career coaching or figuring out what to do. Accountability is a bit different. I've also found that you can't really make people work hard. People that don't want to work hard and drive, you can't force them to do it or motivate them some way. It's just this internal drive. And so I think our structure and operating model is actually just as much a reflection of the type of people we hire as it is the type of choices we've made.

Nakul: And so maybe double click on your recruiting philosophy then. What do you look for when you're recruiting engineers with this org design in mind?

Michael Grinich: Yeah. So engineers, in particular, definitely have to build stuff. They have to build stuff end to end. We look for product engineers that can go talk to users and actually figure out what to build. But we've also found we can teach people that. It's actually not as important that they've come from a product engineering culture. Probably the most important thing we look for in engineers is curiosity. And it seems obvious like you'd want curious people, but there's a lot of jobs actually including engineering where you don't need to be that curious. You can just repeatably do the same thing again and again and have it work. So, we look for people that have come across situations where they weren't prepared for it or they didn't have information, but they were high agency and they're able to go solve those problems.

Nakul: In your assessment, do other organizations hire VPs because they think they need to hire VPs? Why are other organizations not saying the same thing like, let's look for high agency, curious people, product engineers who can own end-to-end. Everything you just said seems beautiful. Why are other companies not adopting this philosophy?

Michael Grinich: It's a good question. I don't know. A couple years ago, I think it was when Paul Graham wrote that founder mode essay and Brian Chesky was talking about it. Even before that, people talking about the PM role kind of dissolving and engineers doing it. We at WorkOS, we were kind of like, "This is just the way we've been operating. This is just what works with us." So, I think there is maybe more acceptance around that as a role. I certainly think that there's a big piece of it where companies at earlier stages look to mimic what later stage companies do or try to emulate what successful organizations do, and they back propagate it. The biggest example of this I find is people having typical board meetings. A board structure and a board meeting needing to govern a public company is completely different than what you need to do at a startup when you're finding product market fit. To apply that is kind of insane, but yet we still do it. And so, I think the same thing happens with executives. The type of executives that are amazing for scaling later stage operations to come in and build something zero to one is not going to be the right fit. And so it's just kind of mis-hiring. Unfortunately, people don't really write about this. I used to think that post-product market fit, everything got easier because there aren't blog posts about it. Everyone writes about pre-product market fit but not post. I think the reason why people don't write about it is the advice doesn't generalize as well. Pre-product market fit, talk to users, eat, sleep, write code, that's it. Post-product market fit, you have a thousand things to figure out. Every company's a little bit different in that way. But the era of the builders, the era of the high agency IC I think is now, especially with AI and especially with the acceleration happening in the industry.

Nakul: So, we want to dive deeper into this AI and the future of engineering org, but let's touch on that here. How has your recruiting changed in this AI era? Are you looking for different things in the people you're hiring?

Michael Grinich: We look for the same stuff. Yeah. Our recruiting hasn't changed. We get more random applications because of AI, and we even use AI to filter them out just from efficiency. We look for people to use AI natively in the company. And so, when people interview with us on the engineering side, if they're not using any AI in how they're writing code, it's actually a red flag.

Nakul: What's the signal that somebody is truly AI native versus just using it on the peripheral or showing it in an interview?

Michael Grinich: Yeah. I think asking people how it's changed the operating cadence of their life, which sounds like a giant question, but there's some people that are not using it at all. They tend not to apply to tech startups. And then there's some people that are still using it casually. They're like, "Oh, I use ChatGPT once in a while to refactor something or reformat it, and that seems like a cool effect." And then there's people where it's completely changing their capabilities, what they can build. Sometimes you see the bags under their eyes because nights and weekends, they're cranking on stuff. They're imagining building agents—

Nakul: You want to see the bags under the eyes.

Michael Grinich: Totally, yeah. I mean, there's been this funny thing the last few weeks of CEOs coming back and writing code and building stuff. And my GitHub contribution graph is turning green from all the side projects I've been building and working on. I think you're looking for people like that that are curious. How can you not be obsessed with this thing? It's like the biggest technology transformation. If this you're not going to be curious about, you're probably not curious about a lot of other things in your life. So, it's pretty clear when people are AI pilled or not. And we look for people that want to lean into it, that want to do more of it, that want to figure out how to build a company with this as the foundation versus reluctantly being dragged forward.

Nakul: So, let's talk about the operating cadence because it's not just a company of super ICs, but the first 60, 70 people were all distributed, partly because it started during COVID, maybe partly intentional. What's the operating cadence like at the company? What meetings matter? What's the weekly rhythm?

Michael Grinich: Yeah. So, the company is still very remote. We just opened our second office in New York. So, we have one in San Francisco, one in New York. But across those two locations combined, it's actually only about 25% of the company. And even people in San Francisco aren't necessarily commuting in at the office every single day. We open offices to have more space for onsites, for events, for community engagement. Can't even describe how much pizza we've had ordered to our office in the last six months, but we still operate the company as a remote organization. It's a distributed organization since the beginning, since pre-COVID. We run the company really on a week-to-week cycle. So not two weeks, not sprints, not quarterly or half-yearly.

Nakul: So, every week matters.

Michael Grinich: Every week matters, yeah. So every single Monday, today's Monday, we had a staff meeting this morning, and we go through the entire business top to bottom. We break it into two pieces. The first staff meeting is the go-to-market side of it, which is sales, marketing, events, finance, kind of deal pipeline review. And we go through everything that's there. And then the second part is engineering, product, design. Go through all of the projects, all of the teams. And between the two of those, it's about two and a half to three hours total. So, it's not a crazy amount of time. It's like a half day. But on that cadence, we're able to reevaluate the entire organization every single week and create this really strong alignment across all of the team managers and leaders. And the meetings there are a combination of team leads, managers that are in them, as well as really strong ICs, staff, principal level folks that are driving forward projects. So, we do that every single week. Those meetings are recorded. They're public to the whole organization. The notes are open. It's not some secret decision making cabal. It's just a high context, high decision making environment. And then throughout the week, teams will meet. There'll be follow-ups from that. There'll be things that we build and ship. There's subteam meetings. We have a marketing meeting throughout the week. There's multiple deal pipeline reviews as customers are moving through, things around recruiting, other project meetings. And then every single Friday, we do a full company all-hands for an hour where we kind of do a similar type of thing to that first staff meeting where we're running through updates. It's a bit more condensed. It's the whole organization. So, it's like the highlights. But my favorite part of that is demos where I'm showing off stuff that we're building and surfacing it in front of the whole team.

Nakul: Is there a central doc where everything comes together and everybody stays in sync given everybody's distributed?

Michael Grinich: Yeah. Every Friday, the all-hands, there's a document that goes with it that we run through. We prefer to do documents instead of slides. And so it has a lot of textual content. Not all of it is read through. It just is there, it's available for everyone. And we cover the business top to bottom, revenue numbers, customers that are going well, things that are not going well, candidates in the pipeline, people we just hired. We do everything. So, one of our operating principles is radical transparency, talked about internally. And I think it's kind of similar to an AI agent. If you don't give it the context, it can't actually make decisions and reason. And so we try to provide all this context to people. And for people working at WorkOS that just joined, if they're an IC on the team, that all-hands meeting, it's dense. It's an hour. It's like you really got to pay attention. It goes fast, but it's so information rich that we find it's just this injection of information about what we're doing right now. And then when we need to change, we can. We can adapt the whole organization to it.

Nakul: What's the operating cadence or one operating cadence that outsiders might not know about WorkOS that brings it all together? Is that the Friday meeting? Is that something else?

Michael Grinich: I think it's probably the speed at which we operate, yeah. That's usually the most surprising thing.

Nakul: Is that a core value or an operating principle?

Michael Grinich: Yeah, it's one of our operating principles, yeah. It's "move fast with intention," we say. But probably the speed. Nat Friedman, who's one of our angel investors, I'm a huge fan of his. He told me once, I was on a call with him catching up, and he used to say, "A week is 2% of the year." And I've adopted this as well now in the company. And so on a Monday during the staff meeting, especially for new people that come in, it's a default thing to say in many organizations, "We'll follow up on this next week." And we have like an allergic reaction to that as a team. We're like, "No, no, no, let's follow up tomorrow. A week is 2% of the year. What can we decide by tomorrow? What could we build by Wednesday or Thursday? What could we ship by the end of this week?" Long-term success is just short-term things stacked on top of each other. And so the cadence that we actually operate on, the week, is not the shortest timeline. It's actually the biggest timeline in a sense.

Nakul: But is there a cost to moving that fast? What do you give up?

Michael Grinich: I think so. It can feel thrashy sometimes if you change direction too frequently. You also give up some amount of legibility across the organization. One of my favorite books is this book called Seeing Like a State. And one of the tenets of it is that the state, like the government or an organization, is designed around creating legibility in systems. The perspective of the state is to impose something that isn't actually the best for the citizens or the output of the organization, but the best to be the most reasoned from a viewpoint. And so I think in terms of moving fast, one of the things you can give up is this feeling of knowingness about the whole organizational legibility, knowing what everybody is doing. If your core tenant is you need to have perfect clarity in what all the tasks people are doing, your fastest you'll move is the communication speed across the whole organization, which is slow. So, you have to push decision making locally or you have to get really involved, like really hands-on. For me, rather than waiting for engineering managers to surface things up, I'll go work hands-on with engineers to shortcut it. And building an organization where that is actually exciting and a good way to work versus something that's scary or not a good way to work. I think that's just culturally a hard thing for folks to do.

Nakul: But is there an intensity there that people self select into? It seems like culturally that comes with the speed.

Michael Grinich: Yes and no. Engineers always like moving fast. I've never met a— I mean, there's things where we don't move fast, like when we make API changes or infrastructure changes. There's a project we worked on that took us over a year to roll out, because we were very methodical about how we did it, to not have any issue with our customers. But generally speed is something people are drawn towards. I think the place where people get stuck is this idea of "hire smart people and get out of their way." That's this phrase that's often used. And I don't think you should hire great people and get in their way. I don't think you should ever get in people's way. But there's a version of this that's like, hire great people and unblock them as fast as possible and give them deep context and help them be really successful. That is much more hands-on. And it requires also managers to have a harder job. Managers have to be driving forward a ton of the context. They have to be really high agency themselves. The job of the manager is way harder in this type of environment, way, way, way harder. And there's just not that many managers that can do that. Realistically, I think it's the reason why they're more impactful and generally they get paid more than ICs too. So, it is self-selecting in that way a bit, but we've also found we can teach people to operate like this.

Nakul: So, let's talk about brand. Developer companies typically do not invest in brand the way you have. You have MCP Nights, especially in the last couple of years, it does seem like there's this internal phrase you guys have mentioned, WorkOS is everywhere, and it's really been everywhere. When did you decide WorkOS needs to be the brand that it has become today? What was the impetus? And was that something that was core to your philosophy when you got started or you learned along the way?

Michael Grinich: So, brand has always been very, very important to me. And I think generally design, like product design, product experience. I just never really wanted to build anything I wasn't proud of. And so a big part of that was the brand and communication of it. So, even from the very beginning, we were obsessing around the visuals and look and feel and the communication style of things. And to me, brand and design, it goes down into the copywriting, the text that's on your homepage. And it's very challenging to get right. It's one of those things that when it all locks in together, it feels good and you kind of know it. But up until that moment, it's awkward and doesn't feel good and it's painful to create. Brand has always been really important. I would say in the last year, maybe two years, we've invested a lot more in it as we've scaled. And one of the reasons why, I think, we started having a few more competitors that started just trash talking us, honestly, especially with VCs. They were saying we weren't doing well, and just stuff that wasn't true. And before that, we were pretty heads down. I mean, we're an infrastructure company providing authentication, security stuff. We're not the star of the show. We're not in the limelight. We don't need to have the biggest attention on us, but it really bothered me that someone else was telling our story instead of us. And I thought it did a pretty big disservice to the amazing engineers we have that are just laboring over this and suddenly someone else gets to go say that we're not doing well. And so part of it was just my reaction to that, saying, "I want to tell our story a bit more broadly." And then we found it ended up being really effective because customers got more excited about it. I started doing these customer weeks where I would just celebrate people had launched on the platform and started using our stuff and it has just compounded. If you look at our customer base today, it would be like a top decile VC firm. It's incredible. And the pattern across all of that has just been this relentless focus on the product and then the marketing has followed from that.

Nakul: It's like that famous Michael Jordan thing, "and I took that personally."

Michael Grinich: I did.

Nakul: Sounds like that's what happened to you?

Michael Grinich: Yeah. I mean, I'll try not to—

Nakul: Don't mess with this guy.

Michael Grinich: … start cursing. But it's a little bit of like a mama bear. It's like you came after this thing that we've been laboring over, trying to create greatness, and we're going to defend it. We get fierce around that thing and fiercely supporting our customers too. We are the best product in the category and we'll stand by that. So, I think that's part of it. On the brand side though, like you mentioned the MCP night, there's another piece of this that's just kind of experimental. So, MCP. About a year ago, I started getting kind of obsessed with it. It's this new protocol that came out from Anthropic, people were demoing it. I go to tons of meetups and hackathons and evening events. And so we decided to do an event all about showcasing MCP. And we got hundreds of developers to show up and it was this amazing super meetup/mini conference. We ended up doing three of them last year. The last one, it was our holiday special. It's 700 people at the Regency Ballroom. I wore a tuxedo and was standing on stage talking about this AI model protocol stuff. It's weird. And so, there's a piece of our branding and sort of event marketing that comes together that I think for us is just like a celebration of technology and builders. The company is made for engineers. We love making things for other builders. We will do stuff that's just cool and interesting to do. And at another type of business, you might have to justify it or you might have to say, "How many MQLs did we drive with this type of thing?" To some degree, we just follow our nose on things that's really interesting and great to do. And that has led us to discover what products to build, it's led us to find other customers, partners, it's led us to hire people. It's just an expression of our love for making things for other builders.

Nakul: So, it is an engineering driven company. We've established that in this conversation also, but we are in this moment of time where the engineering profession is dramatically changing. And so let's talk about the future of engineering. First of all, how has AI impacted the productivity per engineer at WorkOS? Have you seen tangible?

Michael Grinich: Oh, absolutely. Yeah, we're moving way, way faster. I mean, I thought we were moving fast before. Now we're moving way faster. I don't know how to quantify it. I mean, we try to measure stuff, lines of code, speed of PRs and stuff, but it's tangible. I mean, it's clearly not just us, the whole ecosystem is moving faster. It's also more fun. People can build things and express ideas and discover things and learn things much quicker too. Even me personally, I've been spending weekends hacking on stuff and doing things that otherwise would've taken me three or four months and I can do it in a weekend. And that's wild. I mean, that's just still hard to wrap my head around. So, it really is changing a lot of stuff. We're rebuilding our entire kind of infrastructure and code base around enabling agents to author code on top of these systems. We're thinking about new products we build for agents in addition to people. We've been doing that for a while, but I think now it's really in the forefront. And even how teams operate. I was talking recently with some folks on our team about we only have one PM across 50 some engineers in the organization because engineers are able to do a lot of that product work. I actually think if engineers are able to move faster, we might need more PMs.

Nakul: Yeah.

Michael Grinich: Actually might go the other way because we need more people thinking just about product. If one engineer can be like 10 engineers worth of impact of the previous era, maybe the ratio changes.

Nakul: Do you feel that the number of engineers you have today is now forever going to be stable?

Michael Grinich: No, we're going to keep growing just because we want the platform to do so many other things. I mean, in the last year and a half, we've launched six major new products. It's this explosion from just this auth and identity stuff to now we have a product that does fraud and abuse detection. We have a product that does integrations, an encryption product. We launched feature flags. There's more stuff coming in a few weeks. We got tons of surface area. And so, just even having one or two people, usually a good kind of team is two to three. So if one person's sick, you have a backup working on it. It's more fun to work with other people. But I think we'll have way fewer than companies in the past. Today, we're a team of 50, 60 engineers. It feels like a team of 100-some engineers. And I thought to get to this point, the company would need to be two to 300 people. We just passed 100 last month. And so I do think companies generally will need fewer people, but the company headcount growth rate might be similar to the past, just the businesses will grow faster.

Nakul: Yeah. So, that brings to the question of what do you think is the future of engineering orgs? If you were starting a new company today, do you think you would answer that question differently and say the maximum number of engineers for an engineering org would have been 20 or something?

Michael Grinich: I think it depends on your product. So I think there's a ton of products out there which are not selling to developers. What they're selling ultimately is not technology in the sense of bits to technologists. They're selling a service enabled by technology to another business. And those businesses won't need to hire teams of dozens of engineers. Today, there's a lot of companies that have 15 to 30 engineers, 40 engineers, that's their terminal state, and they might have 150 people in sales. That's not uncommon if you look at the ratios of stabilized organizations. Those engineering teams might be able to be a lot smaller. Maybe it's like 10 engineers or something like that. In the same way, companies don't have enormous numbers of accountants on staff or people doing compliance. You have tools that help enable it. So, I think that will reduce significantly. At the same time, the ambition of what companies can do and the scale of your impact and how much more is demanded by the market continues to increase. So, I think it's actually one of these Jevons paradox types of things where engineers will be able to do more and therefore people will need more engineers. But the type of engineer is changing. You'll need engineers that sort of can manage multiple agents and work across streams. This idea of a product engineer that can do things end to end, I think, is really high leverage with AI systems. Ivan, from Notion, he talks about AI being the manager of infinite minds. So, I think the way organizations operate will be pretty different. And the era of— We've already seen this with front end engineering. There used to be people that just did front end engineering, take the Photoshop document and turn it into a website. That's toast. That's gone. And I think there's going to be other pieces like that.

Nakul: How has AI changed your operating cadence at WorkOS?

Michael Grinich: I can consume a lot more information. Well, first of all, I can build stuff again. I can build stuff. I hate webpack. It's this underlying technology used for packaging and compiling and transpiling languages. Yeah, it was my first experience like a year and a half ago with Cursor. It just fixed this whole thing for me. And so it's enabled me to get back in the place I love, which is making stuff and building things. So, that's one thing—the joy of engineering is there. I use it a lot as a deep analysis thought partner. I will pipe into AI tons of different resources we have as a company or things I'm thinking about. I use it really frequently to help synthesize my thoughts around memos that I write to the company internally. That's a pretty unique thing. I used to have that by talking to other people and I found I can actually get even higher density through different AI systems. And then the last thing is building products for AI and agents. No longer is it just building things for people, it's building things for people and machines.

Nakul: And beyond you, have you just mandated everyone needs to be AI native at WorkOS? Is that a top down mandate you've given, or how has it changed the rest of the company?

Michael Grinich: Yeah, it is. It's actually become one of our operating principles, AI native. And I would say the way we mandate it is not like, "You have to do this or you're fired." It's more that we all see this being a transformative impact to the organization, and it's something we actively have to push on. So, the whole idea of our operating principles as a company is there are behaviors or things that we want to focus on that we won't naturally converge to. We have to intentionally put energy to. For example, build for developers. That's our number one operating principle: developer joy. Most companies don't build for developers because they don't sign the check. And so if you care about developers, you have to make it always a focus in your mind to stay in that space. So, I think AI is something similar to that for organizations that have been built up before. Maybe not today for a company started by a bunch of Gen Z founders that are all AI built from the very— It's just they're swimming in it. It's just native. But for us, we have to rethink things like as we scale our sales team. How do we build that in an AI native way? When we interview people, we look for it. It's in our performance review cycle. Of course, it's in engineering, but it's everywhere in the organization too.

Nakul: Are you expecting leaders to report how, like non-engineers, so your sales team or customer support team or developer marketing team to show some semblance of, "Hey, we've pushed AI nativeness across the board in our function"?

Michael Grinich: Yeah, definitely. We expect to see that across people. And I think if I have a conversation with anybody on our team, and I say, "Walk me through how you're using AI or what you're experimenting with or what you're building or how it's changing your work." And they don't have some answer to that, that's going to raise alarm bells. Or if they're saying, "I think we're doing enough," I think that raises alarm bells because of the speed of how much this is moving. I've became a student of technology and business through starting companies. It's sort of my hobby as well as my job. There's never been any disruptive technology that's been at this scale and moved this fast. There's been things at the scale before, like Industrial Revolution, printing press, the internet, but nothing that's happened in this compressed time interval.

Nakul: It's insane.

Michael Grinich: It's totally insane. It's so fast that it's unnatural to reason about. And so you have to force yourself to re-experience it. A very natural thing for engineers to try something, evaluate it, come to a conclusion, and then have that form the basis of how you have other thoughts. It's very unnatural to say, "I know I looked at this a month ago, but I'm going to forget everything I realized and come at it with beginner's mind," like a tabula rasa type of thing. And that's very uncomfortable for people to do and especially for managers to do as well. So, I think it's just as much of a cultural change as it is a technology change.

Nakul: On that point about how fast everything is changing, doesn't it feel sometimes crazy? You feel like we're right in the middle of a science fiction movie or something. I don't think even in my future I will ever be as excited as I am right now in this two to five year period. I don't think it is possible to be more excited.

Michael Grinich: It's wonderful. If I had known this was going to happen when I was in college or maybe a little bit earlier when I was in high school, I actually think I would have been more excited about my life and my career. Not that I wasn't excited, but I'd always liked building stuff. I liked the internet. I like making stuff and played with Legos as a kid. I'm a creator, like a maker at heart. But if I had known that this was going to happen and this was coming, it's the golden era for making stuff. And the AI models, it's a human computer interaction breakthrough as well. It's not just writing code.

Nakul: Totally.

Michael Grinich: The machines now are more human. We can make these things that truly feel like what science fiction was in the past. It's this unlock. And that to me is just absolutely thrilling. It's like I have a hard time sleeping because I just am so amazed at how incredible it is and plus the ability to go build stuff. It's like someone just gave me a pile of super Legos. And all of that together at the same time, I have this startup that I was building that's like this enterprise API infrastructure thing that's now powered by AI, which in hindsight makes sense, but previously was not at all designed for this. So more than anything, I just feel lucky. Lucky that the thing I had is now in the middle. That's just not a stroke of genius, that's luck. And to have the convergence of those two things, it's like a video game plus toy that you can't put down and then the dopamine hits you get from it. I mean, it's just so cool. I mean, you can have it do anything. And I think going back to what you were talking about at hiring, I look for people that have had that moment, that I've had that realization or are leaning in that way. And I think it happens for everybody a little bit of a different way.

Nakul: You're 100% right on this, that some people have had that moment and some just haven't. And—

Michael Grinich: I call it being AI pilled. It's kind of like in The Matrix where he takes—

Nakul: Actually, you just said something that's pretty powerful.

Michael Grinich: … takes the red pill. And I've even talked to people on our team and people from many other companies about how to AI pill other people, because it's a thing you can actually—

Nakul: Can you AI pill other people?

Michael Grinich: Totally. Yeah, absolutely. You absolutely can. Not 100% of the time, but almost always.

Nakul: But if they have not AI pilled already before meeting you, is it saying something about their curiosity?

Michael Grinich: For engineers, for sure, yeah. I think for engineers, if you're always looking for a new way to adapt your craft and you're not seeking this. Generally, it's not that they won't become it, it's that their energy will be more of a Luddite, which teams—you want to have this balance of people that are always looking for the new thing and then people that are also on the other side that are a little bit more conservative. If you have too much on the one side, it just becomes total chaos and you're really building everything all the time and it's not a good balance. So, I think even from our side, from our team, we have people that are, I wouldn't say more conservative around it, but they're trying to find the right ways to apply it or not to introduce security issues or not introduce performance issues. But even on the security side, just a month ago, one of our security engineers started using the GPT-5 high reasoning model to find security vulnerabilities in our product, and actually was able through a set of different prompts and through a bunch of prompt engineering to actually identify stuff. Very weird, small edge cases, race conditions, the sequence of things that need to occur for that to be a vulnerability was pretty extreme, but it still found it. It found these logical issues in our product. Wasn't ever exploited, but by solving it made it much more secure. And so I think even with that, with people that you maybe wouldn't think of as the people driving it forward, getting them AI pilled often can lead to the biggest outcomes. I'm finding that if you talked to me a year ago, I would've probably said, "We'll need to go hire a bunch of junior engineers that are fully using AI." I've actually found the most senior engineers are the most impactful with it.

Nakul: Understand, yeah, yeah.

Michael Grinich: It's like a super tool for them on top. So, you definitely can AI pill anybody, but it's such a transformative technology. Most people have never experienced anything like it, and so they need a few reps to get it in.

Nakul: So, what's the new bottleneck in company building?

Michael Grinich: Knowing what to build, I think, is the hardest thing. Today for software engineering, you can make pretty much anything, you can make it really fast. And so the question is, what do you do? What do you do next? What do you build? Where do you go? What is your strategy? Where's the stuff moving? I think that's one big piece of it. I think the other thing is understanding your customer really well. These are things that have always mattered in the past, but the company that can understand its customer better than the competitor will be able to consistently deliver a better product that's a better fit for them sooner. And so it's deep empathy with the customer and connection to them in that way. I think it's those two things. And then speed goes back to, can you move fast today? If you're not using this stuff, even if you're moving 10% slower, that compounds over time, you'll be smoked by somebody else moving quickly. And so speed is everything.

Nakul: So, let's shift gears towards culture. So, you describe WorkOS as a place of low oversight but high expectations, right? And that sounds great on paper, but in practice, how do you avoid micromanagement then or do you not?

Michael Grinich: Yeah. Probably if you ask people on the team, we don't avoid micromanagement. It's not really a dirty word on our team. I think micromanagement generally gets a bad reputation because the way people experience it is managers that come in that have no context, that don't seek to learn or understand anything and just say some random thing. And then the ICs are left being like, "What do I do with that? This person's coming and making a decision. They don't know what's going on." So in order to have that type of hands-on work from leadership, I think it requires the management leadership to be really well-informed, really, really well-informed. I'm talking like reading hundreds of support messages every day, deeply understanding the customer, doing more work than anybody else in the whole organization. I spend a lot of my evenings watching what I call WorkOS TV. I go watch customer calls. I get on my exercise bike, I jailbroke my Peloton so I can use it to watch Gong calls. And I'll spin on the Peloton bike and watch them at 2X, just so I'm saturating my brain with as much as possible. I think if you can do that, you have the opportunity to get in and do more of like micro-leadership. Be like, "This very specific thing you're deciding, I think we should go this other direction because this, this, this, this, this." And if you're able to explain why and provide the context to people, I've actually found that engineers want that. They don't want to build the wrong thing. They don't want to go in the wrong direction. They don't want to be handed a spec and just to go make it. They want to be provided more information and context in it. So, it's really this hyper collaborative environment and then with high agency, low process, people take things end to end. Engineers are responsible for product shaping. They're responsible for writing the change log for the announcement, for following up with the customers that asked about it, for ensuring it's successful. You're not just a code monkey in the middle, you do the end-to-end thing.

Nakul: As this conversation has played out, it's clear that you have a lot of intentionality around culture and all of that. It's a distributed org, senior ICs, high agency at the center of it, but even the way you reframe micromanagement to micro-leadership just now. In a distributed org where people are all separate, how did you ensure this is uniformly getting distributed as cultural norms and the cohesion of operating principles coming together? Whether rituals, is it just over communicating week after week? What brought it all together as new people are joining, but they're not in the same office?

Michael Grinich: Yeah. Part of it is writing it down. So operating principles, living them, it's a reflection of actually how we were doing stuff, not just aspirational, but actually how we're operating day to day. I think another piece of it is just time. As a seven-year-old company, we've just had the ability to continue reinforcing things that work really well. I think it's very hard to build culture super fast. You just don't get the density of the stuff that works. And so I don't know a way to speed run it. A lot of the way that we operate and the cultural practices we have are actually extremely well suited to the AI era, which again, is just happenstance. I think of how our product engineering function is really high leverage with AI. That's part of it. We do some stuff where we bring the whole team together every six months or so. We do subteam gatherings more frequently. That's what we use these offices for. We do a lot of events, but early on we didn't do that. I think our first onsite was delayed due to COVID. We were going to do it. And so the thing that brings people together more cohesively, I actually think is building the product. The center point of the company, the center point of the culture should not be we all go get drinks and do karaoke together. It shouldn't be we all get fat bonuses when we hit some revenue number. It should be about the customer, about the reason what you're providing and why that's impactful to the world and what satisfaction you get out by that infinite game that you're searching. And I think all of our operating principles and the cohesion of the company really centers around why we should exist in the world and the type of businesses we work with. That's the most durable, that's the most powerful over time.

Nakul: I want to talk about navigating some tough times. And some of this is being informed by me having a little bit of a front row seat to your journey. So back in 2020, March 2020, you were planning to launch WorkOS, the first product going out in March 2020 at SaaStr. And then COVID happens, lockdown starts, that event is, I think, gone and the whole launch plan is gone. What happens next?

Michael Grinich: Yeah, we were going to launch it. I'm trying to remember. I think SaaStr was going to be in April or May or something. And so it got canceled. I remember Coachella got canceled and that same day SaaStr got canceled. And so—

Nakul: The two most important events for you.

Michael Grinich: Exactly. Yeah. I've actually never been to Coachella, but SaaStr I've gone to. And I was like, "We're SaaS for SaaS." It's a perfect place to launch it. We were in stealth at the time, we had a small team. And so we just launched it through Twitter and Hacker News and did really well there. If you go back and look at the Hacker News launch thread, it's actually a pretty good time capsule of the company now, six years ago, almost six years ago. I would say that launch was due to us looking at the bank account and saying, "Hey, we raised a year and a half of money and we're going to be out of money by the end of the year. Let's go. Let's get it moving." And so sometimes the market plays your hand, or the market deals you a hand and your job is to play it. And I often feel it's a little bit like surfing. I'm like a pretty terrible surfer. I really like doing it. And especially here in San Francisco, if you go surf out at Ocean Beach, it's pretty brutal. It's a hard paddle out. The wave is heavy, it's cold. You are a small part of the machine that's happening, of the ecosystem. All you can do is control where you are and how you move, but the power is in the ocean. And I think that's like the market. You can only surf the wave that's there or not and deal with the conditions that's in front of you. And so COVID was just like another one of those for us. And there's been big ones and small ones. I think COVID, in particular, we all experienced it at the same time and it had different impacts in all of our lives, but generally the impact was kind of the same. We were all kind of stuck and locked down. It was a giant, shared, traumatic experience for all of society, which typically only occurs during natural disaster or famine or war. It's a very interesting, different type of thing. So, that was challenging for sure. We were already a remote company and so our operating model didn't change. In fact, it was actually identical, but of course, enormously stressful for everybody. Their kids can't go to school. Now, your partner's at home and you're working alone. It's like everything else is different. So, it's very challenging in that regard.

Nakul: Correct me if I'm wrong on this, your plan was to launch in April and then raise a Series A in June-July based on the launch metrics. And then you decided, "Look, I don't know where the world is in three months. I need to do both of these in parallel." I think you literally did a launch and then two weeks later, you were then raising a Series A. Walk us through that.

Michael Grinich: So this is March of 2020, lockdown occurs. People were saying it's going to be two weeks, two or three weeks or something, which of course we know now is ludicrous. When I was actually an undergrad at MIT, my freshman class I did, the freshman consulting seminar class, was about the Spanish pandemic flu. And we actually wrote a report on how the MIT Institute would respond to a pandemic flu. So I went back and read my old paper and I texted some friends, and we expected it would take two years for the world to recover, which is about what it took. So I saw that and I was like, "We're going to go into an economic winter. Market's going to crash." It's just like catastrophic effect on the world. And so said, "We need to launch to get the product out there and I need to go raise money so I can continue to float the company if the market's going to bottom out." We had a great outcome in terms of the financing. I mean, in terms of the partners we worked with, the amount we raised and everything. But for a lot of the firms that I went and talked to, it was the first time they'd ever done partner meetings over Zoom. I mean, I'd been living over Zoom, so I was fine, but they didn't know really what to do. And I think that's kind of the thing too is like when you make a decision, the important thing is to decide and then execute on that when the world changes.

Nakul: In those moments, once you've decided you stop feeling the stress and you're just acting, or does anxiety live in the back when you know this is going to be a hard raise? Most VCs are not used to making Zoom-based decisions.

Michael Grinich: It's still really intense and stressful. I was thinking this weekend I went skiing. There was a big storm up in Tahoe and I went and climbed and skied this thing at Palisades that I always wanted to. And you're looking over the edge and like when you commit to do it, it's not like you commit and then it's smooth sailing. It's like you commit and then you have to be really fully present for the next however many seconds you're through it. It's kind of like that. You have to be really, really intensely in it. But I actually find that those situations where the world changes or you learn something drastically different and you have to make a change, it's actually a really easy decision. The decision is clear because the trade-off is death. If you're skiing a line and the trade-off of crashing is, you're going to hurt yourself, it's very easy decision to work to not crash and to be very focused and present. I think the bigger challenge is when the trade-offs are more like 51/49, how do you make those types of trade-offs or where it's not clear, it's blurry, or where you're getting different types of information. But in situations where the decisions have been clear, the act of going through the financing, like the Series A raising it, was hell. It was really hard. It was very different than every other financing. It was hard to get information. It was hard to converge to the end financing process. There's a lot of lack of clarity around where the market was going to be, all that stuff. But the decision to do it at that moment wasn't a hard decision. It was actually pretty straightforward.

Nakul: COVID, you were still a small team, but since then, if you've had some tough times or tough weeks, can you talk about how you communicate to the rest? It sounds like you're good at psyche management. It's a binary—I have to make this happen. I'm already in the arena, so I have to make this Series A happen, right? It gives you a little bit of clarity, but how do you communicate to your team during tough times?

Michael Grinich: Yeah, I think the— Well, there's types of decisions that take longer to make because you're trying to gather more and more information. And the ideal time to make a decision generally is as late as possible. At the last possible moment is the ideal time, because by that moment, you have gathered as much data as you can to be as well-informed. The painful thing is at that last possible moment, it's like hiring. You want to hire someone at the last possible moment that you need them to do a job, but it's very painful at that moment too. And so, I think that act of information gathering and weighing the trade-offs can take a long period of time. And I'm often seeking information from our team and other people when I'm doing that. What products to build? What is working in the market? What isn't? What are we experimenting with? What type of people are we hiring? People we just hired, are they working out or not? Not everyone onboards successfully. And we want everyone to be successful, but you're kind of just constantly trying to figure out how the organization's health is behaving. But I think once you make a decision, what I have found—I've actually experimented with both directions, like either giving people the information and not, like on personnel stuff, like why we let somebody go or something. And I've generally found that if you have folks that are high agency and that context is helpful if you're just like, "This is why we're doing it this way." There was a product that we deprecated last year that we were very excited about. We put a bunch of energy into it a few years ago. We built up all this stuff and then ultimately it didn't work. And so I decided to spin it down, recycle it. We actually rebuilt it and just relaunched it. Now it's doing a lot better. But I told the team, I was like, "Hey, this thing didn't get product market fit. Here's why. This, this, this, this, this." But it was tough. It was kind of like 51/49 for a long time. Is it working out? Is it not? We're getting some interest. It's not quite enough. It took a while to come to that decision point. So yeah, the ones that are the big decisions, I often find generally the information is pretty clear around it, but the smaller ones are harder.

Nakul: Let's talk about the inner game. So because you've touched on this a little bit, how has your role as a founder and CEO evolved from managing a couple dozen people to now 100 person company?

Michael Grinich: I think one of the things I mentioned around the culture building of the company over time, now that we're seven years in, there's a lot of people in the organization that have worked here for many years, many, many years. And I wouldn't say they all act like me, thankfully, but we all kind of think in a similar-ish way. I think we've built this sort of brain trust where there's an operating way that we know how to interface with each other and can almost simulate what the other person is going to think. And so you can run that in your head and be like, "Oh, I should probably just do it this way because these other two people would suggest this, and I should try it in this fashion." And so that allows us to act like a cohesive organization, cohesive team. It feels a little bit like launching synchronized— I've been watching the Olympics, like synchronized dancers or skaters or people racing.

Nakul: Have you had to change your management style though as the company has evolved?

Michael Grinich: Definitely. I think I've gotten better at it. I mean, I think that's probably more than style, it's just like technique, which you hope by over time you get better and better at doing it. I think also I've just learned a lot more stuff. One of the reasons I like this job, the kind of company building mode, is it's kind of the perfect job if you are just insatiable for learning stuff because constantly— Well, maybe I'll describe it as this. So, there's this management philosophy called the Peter Principle, which I think is in a Dilbert cartoon, and it means people rise to their level of incompetence. So, you stop getting promoted at the moment you're bad at your job. And it's just this idea that in a stable organization, everyone sucks in management. I think there's this thing called the reverse Peter Principle, where as a founder, your job is to go and incubate new functions. So if you don't have anybody in marketing, congratulations, your job is to go figure out marketing as the founder. And so you go spend all of this time trying to learn about what is good marketing, what is marketing for your organization, and you suck at it for a really long time. And then right at the moment you start getting good at it is the moment you hire somebody. And so your job is actually one where you're constantly not very good at it. And it's really the founding role, the founder CEO, is really the only role in the company where you consistently are allowed to be bad at whatever it is you're doing because you're incubating the function. I would never be hired as a VP of marketing somewhere or a VP of sales, but I can act as a bad one for a while to learn enough, because I bring all this other context in the organization, and I'm there to do the job. And so for that, you're kind of always a beginner and always learning. And it's sort of a painful existence if you like to do things well. But on the other side, it's actually really stimulating and you're constantly kind of turning over things in your mind. And it requires having an extreme growth mindset. People talk about growth mindset, where you're looking for adaptation and learning and change, and you sort of lean into things that feel uncomfortable in conflict. I think to do this, you have to have an extreme version of that and be willing to fall on your face all the time.

Nakul: Yeah, yeah, a lot of the time.

Michael Grinich: All the time. It's like getting punched in the face every day.

Nakul: How do you manage your psyche and health when you're getting punched in the face all the time?

Michael Grinich: Yeah. My perspective around it has shifted over the years. I think early on, my reaction to that struggling, the getting punched in the face was sort of one of "why me?" I was the victim of the situation, or it was so hard because poor me, that my team's not strong enough. I hired this person who's not doing well, customers don't like our product, whatever the excuse is. That's one way to live your life, is to have that perspective. I think it's actually just a mindset shift. If you can instead just think of the struggle as being more of an opportunity, like a gift, a privilege. Just the reframing of that actually makes it something you can stomach. And your reality can be the same in terms of what you actually have to do day to day, but it's just such a more pleasant way to go through things. And it turns things that otherwise are suffering into things that— It might be hard. It's not easier to do. It still can be extremely challenging. It's like adding rocks to your backpack when you're walking around if you just have this negative aspect to it. And it's one of the reasons as well when we hire people, we look for optimists. Pessimists can be successful in the world, but it's just so much more enjoyable to do things with optimists around the table. And just they bring things up. They raise people up versus pulling them down. And so I think that mindset shift is actually one of the things that's happened over the last 10 years. And I don't think I have any specific moment it occurred or some mindfulness practice or anything, but just trying to look at those challenges as opportunity versus burden makes me excited to wake up every day and excited for this change.

Nakul: Billie Jean King from tennis has this quote that says, "Pressure is a privilege," which is like a rephrasing of what you're saying in terms of the gift.

Michael Grinich: Startups are like the extreme sports version of business.

Nakul: Yeah.

Michael Grinich: In a tech startup in AI, it's the most extreme, most dynamic, most fast evolving. And if you're the CEO where you haven't done the job before, you're like in a fresh situation every single day, what an extraordinary gift to have. And I think for this job, I don't know if I would ever be hired to do it again.

Nakul: Yeah.

Michael Grinich: How could anybody be seen to be—

Nakul: Would you hire yourself kind of thing?

Michael Grinich: I don't know. Yeah. I mean, how would you even interview people to do this, right? But to have the privilege to do that, it's not easy, but it's an extraordinary opportunity. And I think about it all the time with investors we have to have the privilege to do that. I feel an immense responsibility to our shareholders and investors to do that, to take it really seriously. It also, I think when you frame it as that learning growth mindset, it's also really fun. There's no speed limit to how fast you can learn. It's just you. It's like athletics, like running. Or cycling or things, I always love these sports where there's no limit.

Nakul: It's you against you.

Michael Grinich: It's you against you. It's you against the clock. That's it. And that's just— For some people, I mean, if you wake up every day and look at that, it's terrible, but in other ways you can look at that. It feels like the most pure. If you frame it in that way, it can be the job of a lifetime.

Nakul: Our next section is a quickfire round. So you ready?

Michael Grinich: Let's do it. Let's do it.

Nakul: Okay. A role you hired too early for?

Michael Grinich: Head of sales. There's this idea that once you reach a million in ARR, go hire a salesperson. And then at the same time, people say, "The first three head of sales you hire are all going to be bad." Which is like, why am I spending time doing this? I actually think founders should run the sales process for much longer than people think, up to tens of millions in ARR. Because sales is where you get the juice out of the market. It's where you learn where to take stuff. It's the most information dense part of the whole organization. The only reason I'm hiring a head of sales right now is because I just don't have enough hours in the day because I have to run all this other stuff. But if I could, I would keep doing it. It's a really, really high leverage. So, I think that's one people should hold off until much later.

Nakul: Red flag you never ignore when recruiting now.

Michael Grinich: Curiosity, yeah.

Nakul: Lack of...

Michael Grinich: Lack of curiosity. Yeah, yeah, yeah, lack of curiosity. You can get at it from a lot of different ways, but if someone's not curious, they're not going to be able to adapt. And the only thing I can guarantee is the world's going to be different in a year or two. And so they have to be curious and I can't teach them. I can't teach them to be curious.

Nakul: Something you've changed your mind on about company building.

Michael Grinich: I used to think that the way to scale was to hire experts that had done it before and plop them in and have them grow and scale. And I don't think that's not useful, but I think it needs to be combined with people that come at things from fresh eyes, and you need both of these ingredients together. The idea that once you find product market fit, just go to this library of people who've done it and throw them in and turn the crank, it's actually not how it works. And thankfully, I think that's actually kind of boring to do. There's a much more creative, interesting piece of it. So, people can learn almost anything.

Nakul: What scares you the most about the next two years?

Michael Grinich: I'm generally an optimist, so things don't, I think, scare me as most. I would say the thing, when I started this year, our management team, the thing I was the most concerned about was that this company was founded in a pre-AI era and even our growth was great pre-AI, but now the laws of physics are different and what you can do is different. And even with all we've done to adapt the business to this, my biggest concern was, we're still holding on to something from that previous era around how we're behaving and operating, we're not letting go of it. And we don't know what that is. It's just so much of a shift. And if you're the dinosaur, you don't know you're the dinosaur. You're not aware of that.

Nakul: So, being the dinosaur or potentially being the dinosaur scares you.

Michael Grinich: Yeah. Or what part of me or part of—

Nakul: Is still—

Michael Grinich: … our operating model is not that? And it requires you to go seek out. It's like that mirror, what is the thing you're not doing that you're scared of, that you're scared of making the change and push through that, reinvent yourself. It's like an unknown, unknown.

Nakul: Okay. Last one. What would you build if you weren't building this?

Michael Grinich: So weirdly, I used to have all these different ideas for stuff and so many of them have converged into this company and I don't totally know why. I mean, I had ideas for aviation, electric airplanes and robotics. And I think voice AI is going to be a huge, huge thing. There are so many industries I think that are slow to pick up the transformation in AI and places where there could be an enormous leap ahead in a small amount of time. And I think those are places where probably I would try to go seed a bunch of companies and get them going. One of them, my dad's a retired physician, my sister's a dermatologist, a doctor. So I see this stuff in the medical world and that's a thing that helps people directly. We build things for other developers building stuff, but that's the thing that changes quality of life. And healthcare in the United States, it's so expensive. I mean, we have the best care in the world, we pay for it. And so there's places like this where there's just such an enormous opportunity to bring everybody up and lower the cost at the same time. And there's dozens of things like that with AI. So, I'd probably go work on something like that, yeah.

Nakul: Closing questions. So, I have the same two closing questions for every guest. So first one is, with everything that's happening in the world today with AI, what do you think our world looks like in 10 years? I know it's a broad question, but take it wherever.

Michael Grinich: I think technology continues to be more and more human. We'll look back on what we use today, and there'll be these clunky devices that we had to tap and stare at in these different ways. And technology will just be more natively integrated with us. We'll have AGI. It'll be a big transformative thing, but it also won't feel like that big of a deal. I think there's versions of AGI we have today and how you search Google. If you went back to the 50s and you showed somebody a Google search box, they would consider that AGI. If you showed them a Waymo, they would consider that impossible AGI. And so there'll be things like that that we'll continue to have across a bunch of different areas. Labor's going to be disrupted in this enormous way, I think for the better for the world. And I think just quality of life will continue to go really well. But I also don't know the future of what does a company mean? If you can have a really small number of people just do something. I don't know. In the past, there used to be all these companies that existed for all these small things that don't really exist, like travel agent doesn't really exist anymore. And it's not like because we lost travel agents as a business, nobody travels anymore. In fact, more people travel. In fact, they have better experiences. So as an optimist, I think things are just generally going to get better. And what I'm hoping is that these tools continue to unlock more creativity for people and more expression for people and more problem solving. And we can just continue to accelerate on that.

Nakul: And then last question is, knowing everything you know today about life and how life works and everything that's going on in the world with AI, what would be the singular piece of advice you'd give to your 25-year-old self?

Michael Grinich: Yeah. So 25 years ago, starting that first startup or had started at a school, I spent a lot of time trying to figure out how I fit in the world based on how other people described the way the world worked. And one block for this was actually around sales. We were talking about doing sales. I studied computer science at MIT, had a CS degree, and I started a startup and the narrative was always like, you go hire the sales guy, certainly you can't learn to do that. Bring in the business professionals, like the MIT guy always goes and hires the sales guy to go run it. And so it was just this thing in my head that I don't have those skills, I won't be able to have those skills. And when I actually decided to learn it, which was during WorkOS during this company, it was challenging at first, but I figured it out and read all the books and met all the people and learned all the stuff. And so I think in a weird way, that was actually this big blocker to me, just learning in that way. And so I would say, don't let others define you. Don't let others, the market or something define you or what you know or who you can be. Don't let somebody tell you you can't run that fast. Prove it for yourself because nobody knows you better than you do. And generally their assessment, they don't even have that much information to make it up. And so why not actually see if it's possible or not? So yeah.

Nakul: Cool. That's a great note for us to end at. Well, thank you for doing this.

Michael Grinich: Thanks for having me. Great to see you.

Nakul: Awesome, man. Thank you.