Scaling Amplitude to $300M ARR with Co-founder and CEO Spenser Skates

So the problem with software today is that me, Spenser, gets the same set of code for a website as you, Turner, as the same as, you know, anyone else out there. We're all running the same piece of software, but that makes zero sense because there's so much data in how we interact with that software and what our preferences are that that software should be adaptive to who we are as end users. And there's really small ways that people are starting to customize software on a per-person basis. Like you might have a favorites list or a recent list or, you know, recommended algorithm recommended feed. Yeah.

You open Netflix and it like, yeah, you kind of knows what you might wanna watch.

That's right. That's right. But it should go a lot deeper than that. Like the UI, like, okay, first features that you like should be surfaced to you. Like that's a really easy one. Features you don't use should not be surfaced to you. Or, and then you go kind of a level deeper than that and it's like, okay, actually the whole interface you could customize and change on a per-person basis. And then you go even further than that. You could actually build features that you think will resonate with a particular user. And all of a sudden you, you kind of take this to the, to the limit. You get to the state where software is being completely customized on a per-user basis. And that's what I'm really excited to call that self-improving products at Amplitude. And that's the vision of where we're taking this company is that we want the ability for software to be customized on a per-person basis.

Welcome to The Peel. I'm your host, Jordan Novak, founder of Banana Capital. Today's guest is Spenser Skates. Co-founder and CEO of Amplitude.

We do product and marketing analytics to help you understand what your users are doing in your product.

Our conversation gets into the importance of data in product design and company building.

The problem of data in your product is a lot deeper, more complex, and more important of a problem to the success of your digital product than businesses give it credit for.

Lessons from scaling Amplitude to $300 million ARR. Like when to go multi-product.

Really what you're architecting the entire company around is a pipeline to a, to a set of buyers.

What he's learned from closing 3 acquisitions.

There's a reason, you know, they say 90% of acquisitions fail is because it's so easy to screw this up.

How he's thinking about AI.

B2B application SaaS has not figured out how to crack AI use cases yet.

The future of user-responsive software.

And so we ended up making massive changes to the interface. In about 4 months, we got 40% more signups as a result of it.

Stories from the early days of Amplitude.

One of my mistakes was spending too much time building the product.

How to construct your board as a startup.

I have learned 10 times more from him than I have from anyone else about what it means to be a great CEO and what most people got wrong about founder mode. Great executives can harness how can we make this vision that you have a reality.

Before we jump in, a quick thank you to Ann Lee Skates, for suggesting some great topics for Spenser. A quick reminder, I publish new episodes of The Peel every week, exploring the world's greatest startup stories just like this one. Tune in next week for a conversation with Sam Ross at Numeral on all things startup growth. Now, a quick word from Numeral. This episode is brought to you by Numeral. Numeral is the end-to-end platform for sales tax and compliance. They handle registration, ongoing filings, and their API provides sales tax rates wherever you need them with all the integrations you need. Numeral sports over 1,000 customers in both the US and globally, and they pride themselves on white glove, high-touch customer service. Plus, they guarantee their work and they'll cover the difference if they mess anything up. They're fresh off a fundraise, raising an $18 million Series A from my friend Chetan Puthugunta at Benchmark. If you want to put your sales stack on autopilot, check out Numeral at numeralhq.com.

That's N-U-M-E-R-A-L HQ.

Numeralhq.com for the end-to-end platform for sales tax and compliance. Thank you, Numeral. And now let's talk to Spenser. Spenser, welcome to the show.

Awesome. I'm really, really excited to be on this. I mean, what's the name of the podcast, by the way?

It's called The Peel.

The Peel. You're not in the capital, so no, I listen. I see it on Twitter. I see it on TikTok. I, you know, it's all that. It's good.

Yeah. You've been starting to kind of ramp up your tweets lately, haven't you?

You've noticed. I was going to do it subtly, but Yeah, I'm trying to figure out how to build a following. One of the things we've been— that I've been frankly really poor at at Amplitude is getting the word out there. Um, to this day, there are competitors of ours who are much smaller that in certain circles have a bigger brand and it drives me fucking up the wall. And so I'm like, you know, we're learning how to go from salesperson to marketer. Um, and so a lot to learn from you.

Yeah, well, talk about sales per, or sales, like you're a public company, you do a lot of sales. Like what's kind of the, I guess really, like really quick, what is Amplitude? And then just like what's kind of the scale of the business for someone who's just not familiar?

Yeah, so Amplitude's a digital analytics platform. So we do product and marketing analytics to help you understand what your users are doing in your product, where they're getting stuck, how you can make a better product. We believe that data is critical for building great products. Um, and so we've built the best-in-class product analytics. Uh, in the last few years, we've added a whole bunch to the platform. So we've added experimentation and A/B testing. We've added session replay. We've added activation, which is like targeting users, and we've added guides and surveys. Um, all of which are products that work much better together with kind of good data on your customers. And so we've built out this whole suite. You know, we think of it as the— we call it digital analytics platform. And it's to— so we sell that to customers and we've been doing that for the last 13 years. In terms of scale of company, we're about $300 million in ARR, public company, went out to the public markets about 3.5 years ago in 2021. And yeah, so, so, you know, now it's about how do we set our— we're early days in this category and I'm thinking a lot about how do we scale here to $1 billion in ARR, become the market-leading player for what we do in digital analytics and this next generation of, of helping companies build better products.

Nice. And I think like the customer base is just the logos I remember seeing in like the, I think the investor relations stuff that you have, it's like Zoom, IBM, the NBA, like, I don't know, maybe FanDuel, Calm. I'm looking at the sheet now to make sure I get these right. Capital One, Burger King, the company that runs Burger King. Like, it's just a pretty, pretty broad, just like big businesses, like corporations that use and have some kind of app or digital product. Web— can you use it for your website too?

Yeah. So any, any, any time a user is using a digital product. So if you're using a website, if you're using a mobile app, If you're using an IoT thing, if you're using VR or AR, um, you know, if we have our product is in airline kiosks as an example.

Oh really? Okay.

It's also in, uh, for Ford, they use this to track, uh, they have a smart console where it's like all the different controls, like your, um, you know, your radio, uh, your air conditioning. And they use it to track all the different interactions that a user makes with their car so that they can figure out, okay, what's most used? How do I make that experience really good and seamless? Um, and so anything that's a user interacting with a product, um, we support and we do a great job of. Um, and so we want to, um, yeah, we, and our thing is to help companies build better products. And like you mentioned, we started out like a lot of, uh, small companies selling to tech startups, but we work to large tech companies and then now to non-technology companies. So we have everyone from Walmart to NBC, uh, to Ford, uh, to the NBA., to DoorDash, Atlassian, Intuit, Square, uh, all customers of Amplitude today.

Yeah, I remember seeing, I think you have 42 customers that do over $1 million in business with you. I was like, yes, that's like you would just think of the product like product analytics like that can't be that significant of it. But it seems like people pay you a lot of money to use the product.

They do. One of the things I've realized is that, um, I mean, so, so first, just for a $300 million ARR company to have 42 customers over me, a million, that is an outlier. Most customer bases are way more, you know, you, you, your average customer size can be way smaller. The reason is, is because the problem of data in your product is a lot deeper, more complex, and more important of a problem to the success of your digital product than businesses give it credit for.

Like, how so?

Let me, I'll give you a really simple example. So One of the inspirations for Amplitude was Facebook found that the best predictor of long-term engagement was how many friends you added. You might say, well, it's obvious, the whole category is social networking. But what was crazy is no one figured that out before Facebook because they didn't have the data to do so. And it was only because Facebook had built out a data science team that they found that insight and they architected a lot of the success of their company around that. Um, and so we do, we help every single company out there do the same thing. And it's like existential, these companies. If you can figure out a little bit better of a product market fit than your competition, you'll win. You'll win the entire pie. You know, it's, if you're either in the technology business, you're either first place or you're completely out of the game. Um, and so that edge that you can get with understanding your users a little more deeply is massively meaningful. Let me give you, let me give you one other example. DoorDash figured out that the most important predictor of whether someone was going to do a repeat delivery was whether their first delivery was on time. So did they give an accurate delivery estimate and did it show up within the boundaries of what they said? And what was interesting, you might, you know, and you can look at that and you can be like, well, that's obvious. Uh, but it wasn't because you might think it's just about speed. It's fast. Exactly, exactly. Their biggest competitor, uh, a decade ago was Uber Eats, and Uber Eats was focused on deliveries in 30 minutes or less. And they were like, okay, we're going to optimize the app around 30 minutes or less. But what DoorDash figured out is people were willing to wait longer, you know, 40, 50, sometimes even longer if you were accurate. And that expanded the range of options they could give people, and you could make the trade-off as a user. It's like, hey, do I want something in 25 minutes or am I willing to wait for the, you know, the restaurant I really want for 45 minutes? And as long as you were accurate about that, you were happy and you would come back. And so, you know, it, it's, it's one of these things that's more subtle that you would not have figured out without understanding the data deeply. And, and so, and, and that, you know, what was that insight worth to DoorDash? Well, I mean, They are the, the number one food delivery provider in all the United States today. Um, and it would have been a tiny fraction of the size had they not figured that out operationally. So, so these questions are existential to these businesses, and we are in the very early days of figuring out how data can be leveraged to build great products. And so we're providing the key infrastructure that allows you to do that. And it's valuable to people. It's incredibly valuable.

Yeah, I think I wanted to ask you why it's so important, why product analytics is so important. I mean, you really answered the question. So then I think just kind of like wrapping up kind of high level about Amplitude, I think it's like $50 billion market or something like how do you, how do you think about like what the opportunity is?

I mean, it's, it's massive. I mean, it's, it's, it's tens of billions, no question in that. Right. You look at the companies that are paying us millions of dollars. Most of the, most of the non-technology companies are still not fully deployed on us. And so, you know, they're just like NBC. We're only in a few applications. Um, we're still not in the main application, Peacock. Uh, Ford, we just got into the, the main driving console, which is great. Walmart, we only work with the mobile marketing team there. So it's early days for a lot of what we do. And so, you know, if we are successful in building that out, I mean, that's easily a multi-billion dollar business, if not tens of billions. I think the analogy we look at quite a bit is we think of ourselves as the next generation of what Adobe has built out as part of its marketing cloud, where—

Yeah, that's a big part of Adobe's business.

Totally. Yeah. It's, you know, the marketing cloud's about a $5, $6 billion business overall. The data part's about half of that business. And so it's very substantial, and we think of ourselves, You know, they've done a great job building for the internet as a marketing channel. We are building a data— the data infrastructure for the internet as the product and value delivery itself. And that's going to be, you know, that's so much more valuable. Now it's early days in the market, you know, we're $300 million ARR, so a lot for us to execute and do, but that's very much what we're aiming at.

So you have like multiple products that you offer at this point. Um, what did you start with originally? Was it mobile analytics? Yeah.

Using an app? Yeah. So very, very early on it was branded as mobile analytics. Then we started, we got a lot of requests where people wanted to tie the mobile and web experiences together. So we called it mobile and web.

That makes sense.

And then we realized really what we're about is the switch from internet as a marketing channel to internet as the product delivery, as product and value delivery. So the application itself, anything with a login with a user that you doing multiple interactions over time, that's what we want to track.

And then how did you think about when was the right time to start adding new features, capabilities, or like entirely new product lines? I know you kind of call out that some of these, when you talk about some of these like non-internet native businesses, like some of them aren't fully deployed, they're not even using all the products yet.

Yeah.

How did you think about when you should add a new product line?

Okay, so very simple equation. It's almost formulaic for SaaS companies like us. As you get, as you pass $100 million in ARR, that's when you go from one product to multiple products.

Why is it $100 million ARR?

Um, you can generally get to critical mass with selling a product, single product up to about $100 million. But in order to get beyond that, you need to sell a broader set of capabilities. And so if you look at the history of every single SaaS company that's gone from $100 million to $1 billion, It's they added other products or other product lines selling to the same buyer through their channel. So, you know, you look at Salesforce, you look at ServiceNow, you look at Workday, you know, you look at Adobe and their acquisition of Omniture. It's the exact same thing. And so the dynamics of a market, at least in, you know, for cloud SaaS, is such that single product generally, once you own it and you can get to $100, maybe a few hundred million, but to get from there to $1 billion, you need to go multiple product. Because there's other pieces of value in order to complete the picture of what you're doing for your buyer.

It's kind of the whole adage of, you know, you start as a product company and then you essentially become a distribution company. Like you use your product to build up distribution with customers. And then because you have that relationship, they trust you more. You can help them solve different problems and it's probably more profitable for you to like when you spend all the time, the investment to like acquire a new customer to instead of just selling them one product, you sell them multiple. So in theory, if you have a good product and you serve them well, everybody wins. You make more money, your customers get a better experience, et cetera.

This is actually one of the hardest parts for great technologists to deeply understand is that the— is how critical the distribution problem is in a business. And so you can create a better mousetrap or a better widget and that will be great, but it's actually, if you own the channel, it's much easier to, to go through and, and scale it up. And so the technology, the specific technology product matters less. And really what you're architecting the entire company around is a pipeline to a, to a set of buyers. So, you know, if you're Workday, you're built up a great relationship with chief HR officers and they need a system of record, uh, for managing their employees. And that's what that is. Or if you're Okta, you're doing that with CIOs. Or if you're Salesforce, you're doing that with, with, with sales leaders. And then, you know, you're going to bundle all sorts of things. And it's actually pretty it's, it's, it's actually easier, uh, and that's the hard part of scaling the company. It's actually easier to, um, attach other pieces of products or other technology through that same channel that you've already built. Like we just did this with guides and surveys where we acquired Command AI, uh, about 5 months ago. And they had done a phenomenal job of building a world-class guide and surveys product. They had this thing called nudges. So, so one of the, You know, I, if you've been to any website recently or any app, one of the annoying things is you'll get what we call a pop-up party where you'll just get all these notifications. Hey, do you want to accept these cookies?

Hey, look, check out this new feature.

GDPR thing reminds you, uh, oh, reminder, you have, you know, notification here. And so you get a pop-up party and it's like, okay, this, you know, it feels like we're back to the 1990s when the internet first came up and like, you just got like pop-ups all over the place, except now these are even worse because they're in the app. Oh, you can't just have a popup block.

It's on your phone and you're like, you like Uber Eats. I feel like it's like a classic where there'll be like a bonus and then there'll be like a little slider thing and you're like, all this stuff, the X is like, you know, tiny thing and it's hard to get rid of.

It's the worst. So, and so anyway, so, so we're, we're starting to get what we call popup blindness, um, as users. And they actually figured out, okay, well, we actually want to give you a notification at a contextually relevant time. Okay, so if you, for example, they can detect if you are confused as a user and then show a popup that says, okay, hey, it looks like you may be confused, which do you want to do any of these things? Let me help guide you to how to do it. Or alternatively, you can, if you're a power user and you want to know about the latest and greatest feature, they can customize a popup that's to you. Or if you're very frustrated and you have what's called a rage click where you're just clicking around and nothing's working, they can also, uh, guide you through that specific experience. And so they, they had done a phenomenal job of building, um, this technology. Um, you know, and they've been going for about 4 years. The company was about $2 million in ARR. Um, fast forward to, uh, you know, we got to know the founders last year. We, uh, acquired the company and they joined Amplitude in October., and we've put out, we had them, they worked with us to rebuild their guides and surveys product on top of our platform. And now we're selling it much, much bigger scale, uh, much, much faster than they were doing that on their own because we've built relationships with thousands of companies already. And, you know, including a bunch of the ones that, that we just talked about. And so, um, yeah, it's, it's, it's, as a technologist, it pains me to say it, but the harder thing is, is to build the distribution channel. Yeah.

What do you see? Well, so going back to our earlier conversation we had this before we started recording, I mean, I put Crunchbase on blast where I was like, I can't— I hate their new redesign. So hard to use. According to Crunchbase, you've done 3 acquisitions. I was going to say, I don't know if that's true or not.

True. They got that one correct.

They got that one correct.

Okay.

So you've done a couple of them. What do you think is like crucial about doing a good acquisition? Like if I'm a founder and maybe I'm like nervous, I'm like thinking about this and like, I want to acquire this company, but I've never done it before. You know, I, I want like, I want to ask Spenser. He's done it a couple of times. Maybe he's done some things well, done some things wrong. Like, what have you learned about doing acquisitions?

So you have to have the most important thing. This is advice I got from, from Brad Rentschler, who architected Omniture's consolidation, you know, what became Adobe Marketing Cloud. Uh, he said, firm on vision, flexible on details. And what that means is that you have to be really clear about what the vision for the combined company is, and you should not deviate from that. Now, how you get to that outcome, that can vary. And, you know, but, but let me tell you where this goes wrong for, for most companies. Is because— so, so when we, when we acquired Command, the Command team, we were really clear. We didn't want them to run separately. We wanted them to fully integrate with Amplitude. We want them to rebuild their product on the Amplitude platform, and we wanted to do that right away, very aggressively. A lot of companies who are acquiring smaller companies will say, hey, well, you know, we haven't decided what we're going to do. Um, or they're afraid to say, hey, we're going to tear down your product and have you rebuild it and integrate it, or we're going to fully merge the teams because they're like, well, hey, you know, maybe you want to keep running independently.

Um, and so I take your deal, somebody else's.

Yeah, exactly.

Yeah, exactly.

And so, but that, that's a fatal mistake because what happens is that it's ambiguous around where you're going and why you're acquiring and what the value is. And so you end up, you, you, you end up in this limbo. And then when you actually do try to force whatever value is, you know, a bunch of people, hey, that's not what we agreed to, or that's not what we signed up for. Much better you qualify that out at the start. If, if they're not down to be integrated into the larger company, then, you know, that's okay. Hey, we'll move on and we'll find someone else. Who is willing to do that, but you're not going to get the value by leaving it ambiguous and just letting them hang around and, you know, hopefully by osmosis, letting the two companies merge. So my biggest advice is just be really clear on the reasons. And there's kind of three general reasons. One is team. The second is technology and product. And the third is business. And so for us, you know, we're, we've very much focused on the first two where we don't want to acquire another business and integrate. That's a whole mess. And, you know, it's a very, very different sort of, that's like a private equity roll-up playbook. Um, but we do want to acquire great teams who have built great technologies and have that come onto our platform and be fed through our distribution channel. And so that's something that we've been very upfront. Um, now to the Command team's credit, they embraced it. They're like, yes, we are excited to do that because the, you have these thousands of customers and the sales team and this marketing engine, and that can get us out there. And we can realize our vision of getting our nudges and guides and surveys products in the hands of thousands of companies much, much faster than we could do on their own. And that's what makes it worthwhile for both sides to make the acquisition happen because there's, you know, there's, there's not just, there's the cost, there's the distraction, there's the merging of cultures, there's all the risks that are associated with an acquisition. And you're doing that because now all of a sudden you can leverage this technology, which was just sitting in a, in a much smaller pond to the, a much wider, wider base of users.

It sounds like you thought about this a lot. Are you trying to do more acquisitions? Like, is this something you're excited about or?

Absolutely. Yeah. We're gonna be more aggressive. We're gonna be more aggressive with them. And so we'll, you know, we, we're interested in looking at, uh, companies doing interesting things in the data space in AI. Um, that's something we're looking for things that are synergistic. Where a lot of, uh, having the fact that you have the amplitude analytics and the data set that we do, uh, will match well with them. Um, and so yeah, we're, we're, we're very, very— we're, we're going to be doing— we're going to do that more and more. I, I will say, like, I, I— even on the, the other two acquisitions we did, I think the ClearBrain one was moderately successful, and then the iteratively one was not as successful. And I think we— with the mistake we made with iteratively was keeping the code bases separate for too long, where we delayed that by a year. And that, you know, it just caused, okay, there's multiple versions of the same thing. Which one do you use as a customer? Which one do we recommend? And then, you know, they get even further apart. And then it's not until you unify them into a single thing that you get the value. But by that point, it was like, you know, you missed out, you missed the boat in a lot of what you could have created together. And yeah, and we'll talk about— that was our fault. I want to be clear, Iteratively team was awesome. ClearBrain team was awesome. Like, we, you know, multiple folks from, from those companies worked at Amplitude for many years. And, you know, I'm still in close touch with, and, you know, they helped us be successful, you know, with our activation product or with targeting or with other pieces of the platform. That was just us learning how to, how to do them. And by that, so by the time we met the Command team, it was great and they were very, And there's, it's, it's different, you know, you can't be a founder anymore. Um, when you're joining a larger company, you still have to, you have to work within the broader machine and there's not always things you can go as fast at changing. Um, but you also have much more leverage, you know, cause you have all these resources that you didn't before.

Yeah. Plus it's probably interesting, like you, you know, you, people might say like you give up your upside or whatever and like, you know, instead of being, creating your own big public company, you know, you get all the the compensation and like the value you create for yourself from that. But also you switch from, I'm thinking Command AI, they're probably like around Series A stage and they're like illiquid private stock that's monopoly money valued from VCs. It's like completely worthless valuation at the end of the day versus it's like, okay, you probably gave them stock in Amplitude or cash or, and so you instantly actually get real appreciation. Like when, when you make progress and the company is worth more, like it's actually real versus—

Absolutely. Absolutely. I mean, yeah. And then, you know, combined the, the, the real magic is that now combined, it's like both companies are more valuable as a result. So Amplitude is more valuable because we're able to sell a broader set of products. Um, you know, Command is more valuable because they get to the leverage of the distribution from Amplitude. And so, you, you know, the, the stock of the combined companies goes up a bunch. And so they, they've, they've been, they've been amazing during that. And we've seen that, you know, we've already seen that in terms of their ability to help accelerate our business. Nice.

And it reminds me, um, prior guest to the show, his name's Rahul Sidhu. He had a company called Aerodome. It was like a drone first response. Basically, you call 911, somebody responds, the, the first responder can immediately send a drone within 90 seconds to get to the site and just like basically, that's way faster than a cop car. Yeah. Yeah. And so it was going really well. Like, he's like, I didn't really want to sell the company, but he had an opportunity to join Flock Safety. Oh, yeah. Like, a leader in public safety. Yeah. You know, thousands of customers. I'm going to get these numbers wrong, but it's like really big business, had all the distribution, and he's just like, it'd be kind of nice to just like get like switch the stock from Aerodrome. I become a part of Flock Safety, I get their distribution, I still participate in all the upside. I feel like it's a, that's a good way to go in some cases.

Totally.

Where it's like you just get stock, you get stock in like a great asset and sometimes your investors are excited about it too. It's like, oh, Flock Safety, we've been trying to get in this company for years and the founder wouldn't take our money, but now we get some.

So. Absolutely. Done right, it can be great for all sides. Now, I will say like, I have to warn people, like there's a reason, you know, they say 90% of acquisitions fail, uh, is because it's so easy to screw this up. On all the different elements that, that we just went through. So, uh, but we've, you know, I think we've looked at companies like Datadog and, and others that have done this well. And I, I think, you know, we've, we've figured out at least something, you know, there's an element of randomness too. And then in terms of, you know, how the teams mesh together and, and how much the— normally you can get the technology and the product and the market risks down, but then it's like, you know, people, people are just random.

So it seems like alignment seemed to be like a big thing just to get right. Yes.

Which is like, which is not what you would I think that's not the sexy part of it. No, no, like all this, you know, you can have a million bankers giving you fucking all these synergy models, like spreadsheets and like, oh, you can eliminate cost redundancies. And if we just, you know, triple this company's growth, you know, this will be worth, you know, this much. And like, you have all these bankers pitching you on all these crazy things. But the key thing is, uh, is you need to be clear in your own head about why you're doing it and then agree on that.

Between, between both, both sides. Yeah, that makes sense. And I know you, you're like thinking a lot about AI.

What's—

I don't know, I don't know, currently the land, just like how are you thinking about it at Amplitude right now?

It's so first thing to be really clear about, it's very, very early days for AI, particularly in B2B SaaS. I think we've seen amazing breakthroughs in code generation., in application development, um, in, you know, obviously all the text and image generation stuff. Um, but if you look at it within B2B SaaS, it's like, it's still very, very early days. Like, I don't think I can name a B2B SaaS company, frankly, that has done AI well.

What is the best— is there like a best example you can think of?

No, that's part of my point. Most of them It's like, okay, we have a chatbot and that's a wrapper around, um, you know, the UI. And it's like, all right, this is literally more work. Well, one, you're just regurgitating help docs at me. Two, it's like, this is more work to try to, you know, I can just create new ticket here as opposed to trying to like type this fucking thing, like telling me to create a new ticket. Like, why, why would I, why, why would I use this thing? And so I think it's still very, very early days in terms of the impact that AI is going to have on applications, application SaaS. I'm not talking about infrastructure. I think there's been a lot of great examples of doing amazing things on the infrastructure side, but on the application layer, like, you know, I haven't seen much. Honestly, look, I'm going to hate on Salesforce and I respect Benioff and, you know, they built the greatest SaaS company of all time and pioneered the whole category for us. But like, you know, you look at, I don't know if you saw the Super Bowl ads on the AI stuff that Salesforce have. It's like, it's Matthew McConaughey was doing some stuff. Yeah. Like he's like, oh, it put me outside in the rain eating at this restaurant and gave me food I didn't like. And it's like, That's not going to ever fucking happen. And even if you spend millions of dollars in Salesforce, Matthew McConaughey is not going to have a different outcome or that, or like he's taking the wrong route to the airport and it's like, that's not even AI. This is clearly fucking made up. Like these are not real use cases at all. Yeah. Um, so, so, so I say that because I was actually talking with one of our board members about this, um, is that he's found, yeah, I think AI somehow, like B2B application SaaS has not figured out how to crack AI use cases yet. And so it's very, very early days, um, on it. Um, and I, I don't know if that's the, it just disrupts too many of their assumptions about the workflows or how value is created or how the use cases are. I see a business model. It's still early. Like I'll just talk about it for Amplitude. Everyone is expecting this magic insight finder, like, hey, we're going to tell you how to increase conversion in your application. One of the challenges with that is that, um, is that a lot of the context is not actually in the data already by default. So what do you mean? Well, so, so like what different properties mean, like, okay, so, so that Facebook example, you had to, you had to understand, or that DoorDash example, you had to understand, okay, this particular variable means delivery time, and this is what you want to look at, and you want to see how it's correlated here. And part of the value is having a human go through that process and set up that experiment and ask that question. And somehow, if you're just feeding it to them, it's, it's, it's not perceived as the same value, and it doesn't register in the same way. And so it's like, okay, well, what needs to be true about how that insight is packed up and how it's generated in order to, um, in order for someone to see it, understand it, value it, and then spread it, spread that insight around the company. And so there's still, there's, you know, that's just a small example from the Amplitude space, but there's still a lot, there's still a lot of figuring out about how is this going to generate value in application layer B2B SaaS that I think we're in the very early days of. So now I want to be clear with all of that said, like we're all in on it. Like we're going to go really aggressive. That's part of why I said we're going to look at acquiring teams. You know, my co-founder Jeffrey's working on it, a bunch of folks from the command team. Are working on AI at Amplitude and we're looking at how to staff it up because there's— we're figuring out exactly what the product paradigm is that works. We've done a lot of great experiments. We have a chatbot, we have Ask Amplitude where you can create charts and it's actually pretty effective at doing it and you can iterate with it, with the chatbot on what that chart looks like. We have things where we'll automatically detect anomalies in your data and surface them to you. We're just starting to build an agent where we will create variations of your website for you to run A/B tests on, like what we call feature experimentation agent. We're going to— we're creating a bot that will automatically create guides and surveys for you. We're trying to figure out this insights piece where we figure out, okay, how do we actually create insights that are valuable and surface them to you in the right way? But anyway, that all said, so we're all in, we're being really aggressive on it. Um, but it's, it's still very early days for how it's going to impact this space.

Yeah. It seems like to your point, when you talk about like, it sounds like the, the meme that I hear is like the software now does the work for you. Like it like starts to replace labor. So it sounds like what you just described is there would be a person that has to set up these A/B tests on the website, measure them, but you're the Amplitude product just does it for me. Yes. And so instead of me having to do this manually, I just see the results of the test you did automatically. Okay.

Exactly, exactly. So, so, and I wanna be clear, we just have, you know, we have some prototypes and we have an early demo on it and it's not like we're figuring, we're figuring this out. But the good part is we've been able to build out a team on this and we're going big on it and I'm spending quite a bit of time myself on it. So I, I, I, I say all that because my, my point is I think the transformations that I think we've seen in other spaces have yet to happen in SaaS. And so I think we're going to see there's an opportunity for companies to take the lead there.

And that's one of our aspirations. It's pretty crazy. You've probably seen, um, I think Stripe with their annual report that they put out, they show these chart of like the growth rate of a traditional SaaS company or traditional SaaS startup and like the growth rate of an AI native company. And they grow like twice as fast or whatever, just like they get, they hit whatever revenue milestones based like in like half the time of traditional.

So yeah, I mean, you see ones that get to $100 million in ARR in like a year.

A year.

It's like the new meme, right? Which is, which is absolutely nuts. But then, then it's like, okay, that stalls and that, you know, half of it churns out the next year. And so it's, it's a very different— the adoption of these products, it is nuts. The, the customer demand, like, so we showed a, what I'd call a concept video, not even a demo, a concept video for what this AI could do. And at our customer advisory board, we had a customer advisory board where we got a bunch of our top customers. And at the end of it, we had them vote on what they were most excited about. And by far it was the AI agent, even though the vast majority of what we showed them was fake and they knew it. They're like, hey, if it can even do a fraction of this, we're in, you know? So it's like there is a lot of demand for this. And I think the companies that figure out like how to tap into that can do incredibly well.

So when you think about like all these technology, big technology waves, things that get a lot of hype, like with crypto and Web3, like if I'm on your board, I'm like, Spenser, we need a Web3 strategy. Like you can say, you know, we're not doing this in Amplitude. Like, this is not relevant to our product or whatever. Like, it's pretty like if you're Delta Airlines, it's like, no, we're not going to give you an NFT with each ticket. Yeah, but with, with this you can say like, yeah, like there's actually like customer, like we can solve problems for customers with some of this stuff.

Absolutely.

So it's like the, the, the extreme like demand that just is there from people that you can, you can use it to solve problems like it's, it's so much more, um, I don't know, visceral and real than maybe some prior, like false start of these, like new platform waves. Like you always see people put up presentations of like the eras of the web are like desktop and then internet and then mobile.

And then like Web3 was one for a while, but now it's like, you know, just, we're crossing that off. The funny part on crypto and, um, is that, uh, You know, it's like, there was this post, I think it was by Moxie Marlinspike, where he dove— he was a technologist that dove into how OpenSea and a bunch of these other blockchain-based systems worked. And he was literally like, "Okay, the goal is decentralization, but one is everyone is using, you know, OpenSea or one of these other providers. So it's effectively a centralized database. And then if you just took the part where it was on the blockchain out and just used a regular database, this whole thing would run like 10 times more efficiently. Um, and so, and like you get the same exact same results and frankly people, you know, could, wouldn't even be able to tell. Um, and I always thought that was, that was, that was a funny one with the, with the crypto wave. So, and that jury's still, still, still, still out on exactly how that's, how that's going to net out. Yeah.

I'm actually, you could probably, did you get pitched of like Amplitude? You should put all the data on the blockchain to be like safe.

Yeah. There'd be, there'd be no reason to do that with us. It's like our, our data. So we are in the petabytes of data. And so to like, to put that on a, like, you're not going to like, you can't transfer that on, you know, over the, I mean, if you're in like gigabytes, like megabytes or even maybe gigabytes, you can do that stuff. But once you start getting into into terabytes or petabytes where we are, it's like, no, like that's not happening.

So it's like across the whole entire customer base, all the, I think I saw a number in, um, the website, it's like 18 trillion interactions of the product that Amplitude has like tracked.

Yeah. Yeah. We have 5 petabytes of data, you know, and that's growing, you know, that, that's growing like 30% year on year.

So that's just anytime. Anyone that uses a product at Amplitude is covering, like any—

Does any action.

Yeah. Okay. So it's huge, huge, huge. So when I'm on my phone, I scroll up and down and up and down.

I just made like 10 more actions. Yeah, exactly. It's like, just as a comparison point, ChatGPT 3.5 was trained on about 500 gigabytes of open source internet text. And so this is, you know, we're at the 5 petabytes scale, which is, you know, 10,000 times larger.

Wow. And that's, I think, the big opportunity in just generally with LLMs and AI is like data, just like having data to use. Massive amounts of data.

Yeah.

Yes. So then how do you think about that with where you're sitting?

Like what kind of opportunities do you have? Oh, that, that's part of what makes us so excited is that we have the opportunity to use this data to help generate insights for our customers and feed that back in their products. And that's a very unique position that we're in is that we have the data and you know, you can train— customers can train models on it and then they can feed that back into recommendations for their end users about, hey, here's a special deal that I know this group will resonate with, or, hey, I know this set of customers will like this different product experience better. And so that's a really exciting thing. And I think very, very little— there's, there's very, very little research has been done on that. There's, there's these things called large action models, which is like Can you replicate the actions of users at scale?

But even those have been trained on quite limited datasets compared to what we have. Interesting, huh? Yeah, it reminds me of, do you know, have you ever come across a product, Ada, customer service? I haven't. So I had Mike Murchison, the founder, on a couple weeks ago. And so their product is, they have one of the larger, like, AI in, like, being used AI application products. Uh, and it's basically just people use it for customer service. So like your whole, and they're sort of these like AI agents, these like, like non-human employees that just live in your customer service, like inside of all your customer service that you deliver. And so if like, it, they, they sort of like replace humans. So instead of like having me and you are like on the customer service team responding to people, It's just like an AI that just does this stuff. And then if there's like a, a break of like, oh, there's some issue where like we can't do this refund to this person in like APAC because of like currency, I'll get a ping and I like jump in and like take over for the AI. So it's an interesting example of like AI in production that's being used at scale. And it's interesting because you kind of talked about the opportunity that they have where this example of doing like a password reset. So you and me were on the customer support team. Someone emails and they're like, I can't log in. And maybe like the first wave of this pre any kind of automation at all is like you respond to them and you'd be like, oh, here's a link to reset your password and you manually type it out. That became automated. And then eventually it's like in the app, there's a button to like get an email to reset your own password. And then there's a, the next iteration of it is actually figuring out why their password got messed up. And it's because in the last update of the product, there was like a bug in how the, I don't know, how the password got stored or transferred into the updated database. And so it logged out their account or whatever. And then the next iteration is, okay, well, can you actually then figure out the bug and fix the bug? So the current process is like, all right, there's a bug in our app. We need to go back to the design team, product team. We need to incorporate it in the sprint and it might get updated in 2 months or whatever. So can you actually then just go fix the app? So in the course of Somebody needed their password reset to a minute later, like you have an update out and like there's no bug in the app anymore.

I, this is something I'm really excited about with software, which is what I think of as user responsive software. So the problem with software today is that me, Spenser, gets the same set of code for a website as you, Turner, as the same as, you know, anyone else out there. We're all running the same piece of software, but that makes zero sense because there's so much data in how we interact with that software and what our preferences are that that software should be adaptive to who we are as end users. And there's really small ways that people are starting to customize software on a per-person basis. Like you might have a favorites list or a recent list or, you know, recommended algorithm, recommended feed. Yeah. Like you open Netflix and it like, yeah, you kind of knows what you might want to watch. That's right. That's right. But it should go a lot deeper than that. Like the UI, like, okay, first features that you like should be surfaced to you. Like that's a really easy one. Features you don't use should not be surfaced to you. Um, or, and then you go kind of a level deeper than that and it's like, okay, actually the whole interface you could customize and change on a per-person basis. Um, and then you go even further than that. You could actually build features that you think will resonate with a particular user. And all of a sudden you, you kind of take this to the, to the limit. You get to the state where software is being completely customized on a per-user basis. And that's what I'm really excited— call that self-improving products at Amplitude. And that's the vision of where we're taking this company is that we want the ability for software to be customized on a per-person basis. And the exciting part is we have the data on people's preferences and what they like and what they don't and where they get stuck and what they want to use. And so we actually, if we can inject ourselves back into the building of the product, we can do that. Now the cool thing is because of the platform we have, we have the analytics, which is people's preferences, and then we also have hooks in with through feature experimentation, through guides and surveys, through activation and targeting where we can actually proactively change the experience. And so we're just at the cusp of starting it now to start to do this. And that's, that I think is a magical unlock in What software can be for, for each one of us. And, and that makes it so much more valuable. Like I was, I was just using, um, as an example, like this, my Samsung smart TV and the thing is like stupid as hell. Like I hate saying that because they're one of our customers too, but like to get, uh, I, when you download an app, I probably want to open the app right after I download it on the smart TV. Like if I'm downloading HBO Max, I want to open it up so I can actually watch the thing, but then like you download it and you have to go to a separate screen and then it's not in anywhere in your favorites and you have to go to apps and you have to search for it. It's like, what? And then eventually I figured out how to like pin it, but it's like, why the hell am I doing that? Software should be able to reactively figure that out, uh, for me. Um, so anyway, so that, that's the future we're excited about building and enabling. And the key part, you know, obviously we have in that is, uh, is with the data that we have on what users like and what, what their preferences are.

That's pretty cool. One actually super interesting point that I heard you, you've mentioned before. Well, I mean, we're talking about, you know, I don't know, innovation. I actually hate the word innovation, but your weird language was like sometimes you would think if you want to be more innovative and move faster, you hire more engineers, you have more people on the team. You actually think if you have too many engineers and like too many people working on things, you actually become less innovative.

How does that— Oh, one of the challenges is most people think it's like, oh, more engineers, the more innovative, you know, the more output you're going to have. So it's like linear. So if you have, you know, 10 engineers, you get 10 output. If you have 20 engineers, you get 20 output. You know, it's like a linear curve. The reality is actually the opposite. It's not only is it not linear, It's, it's, it's, it actually slopes down at some point because really, how does that happen? The complexity. So a lot of things happen as teams get larger. Ownership gets diluted. Specialization increases. The number of dependencies between individuals grows, you know, as the super linearly, so as the square of what's happening. People don't have context. Um, this is actually an observation Fred Brooks had in The Mythical Man-Month, um, which was about the development of OS/2, uh, way, way back in the day, um, where adding software engineers to a project that is late will make it even later.

Really?

Okay. And because the, the, it's, it's not like a, it's not a strict function of the number of engineers working on it. There's the knowledge of the code base and how to make an impact and how to make changes in order to deliver what you want. And so. You have to be very, very deliberate with how you add capacity to a team because in the immediate adding to a team, you're going to slow it down. And then if you don't set up boundaries, if you don't set up the mandates thoughtfully and well, then you inevitably slow down. And so it's no secret, like the most SaaS companies at our stage between $100 million and $1 billion, their rate of innovation slows down dramatically. So their ability create new products and put things out there. And so we know in our category that we are still in the early days. And if that happens to us, we'll die. We'll be dead. And so, um, we need to keep— so we've been very intentional about how do you limit this, the rate of growth of the team so that you keep up the ability to, to where you give really concrete areas of ownership to people, um, and you hold the bar high in terms of what you expect them to do, and the knowledge transfer does not get lost on the current state of the system and how to improve it and innovate on it. Yeah.

So it's not stop hiring engineers.

You still need—

probably are growing the team, right?

Yeah.

It's just you need to think about how you design it so you don't— I mean, is there a way to actually increase, you know?

Oh yeah. Yeah. So, so the trick is, is that you're not going to linearly increase, but you want it to like at least be going up, you know, so it's increasing, so it's monotonically increasing as opposed to going down. Um, and there, there's a number of ways. So one is like new innovations should be done in a team that is siloed from the core team because processes that you'll normally run, like you'll do a quarterly planning, but for a team that's building a new product, quarterly planning makes zero sense. They have no idea what they're gonna build next quarter, you know, let alone next week. Instead, you wanna do something like a weekly check-in where you set the direction and that's it. And you don't plan any longer than a week. And so that's a, that's a very simple example of processes for a large team are not appropriate for a new piece of innovation. And you need to be able to keep both. You need to, in order to do a new innovation within a company, you need to be able to keep that group siloed from the existing systems and processes, which does not happen by default. By default, what happens is an engineering leader or product leader says, here's the process we're running. I want everyone to hear for to this because it's what makes their job easier, but it's actually not the thing that makes— they actually need to apply a different one for different teams. Interesting.

Yeah, I can definitely, I can see how that can be extremely difficult to navigate, especially when you, when we just talked about when you're in that 100 to 1 billion phase is when you need to add new products and you basically need to continue to be innovative. Yes. But then it's harder, I guess. Like, do you see a lot of companies— I don't know, like they die in that phase?

I don't know if that's like a fair— Oh yeah, a lot of companies stall out. I mean, you look at, frankly, most of the peers in our category, like a lot of them have either slowed or are shrinking or are flat. And so absolutely, absolutely happens all the time. And, you know, I mean, I know B2B SaaS the best, but, you know, I'm not surprised if you look across lots of companies, you'd see the same thing.

Yeah. And you said something that I thought was, uh, I would not have thought this about founder mode. You remember when founder mode was like the big thing and everyone's talking about it and like, oh, I'm doing founder mode now or whatever.

Yeah. What's your opinion on how that's kind of played out? I, you know, the, the most interesting part about it was how many, how few founders embraced it and how many backed away.

From it. Cause I thought people were saying that like, oh, I'm doing founder mode now.

Right?

Like that was kind of thing.

Like I'm going into founder mode. Well, if you look at it, you, you, I, you can't find any company at scale that's say 100 employees or more where like there's been very, very few founders. Even Brian Chesky himself, you know, did, did an interview where he backed away from a bunch of the concepts, which is funny. And brought out. Um, the reason is, is because the reality at, the reality at any large scale is the company is the success of a company is the output of the team. Even Jobs, you know, very famous, you know, extreme example of founder mode was said his best product he ever built at Apple was the team. Um, and so the key, the key part, so I think a few things, the first is the key, the key thing that the first key thing I took away from, from founder mode is that you have to, um, you are responsible for the success of the business and you cannot delegate those details to anyone else. And, you know, there might be other, you know, manner, there are other, other ways of, of doing it, but as a founder, like you have to be in the details of the business. And for us, that's engineering and, and sales., for other companies it may be something else. Um, the second is that the right executives, and this is what I think a lot of people get wrong on it, the right executives can set you up so that you are empowered to make and drive decisions which are important for the business. And so let me give you, let me give you the most recent example of this I just experienced. So we just had François Agenstadt our chief product officer, join us a little over a year ago. And one of the things he helped me do is, is he set up a 2-hour product review, uh, for me every week with a place, uh, that where I was trying to drive an improvement in, uh, the, the user interface of Amplitude. So the user interface of Amplitude had gotten very calcified over many, many years and like our signup process was 12 steps long and it was a total mess. And we, you know, we asked you for like 20— we asked 20 questions as part of the process. We had you verify your email. We had to like type in all these different fields and remember them.

And it was just— and your average customer probably just wanted to like get in.

Yeah. See if you like the product. Try it.

Yeah, exactly.

And so I'm like, all right, we got to cut out all these screens. We got to cut out email verification. We got to cut out all these things you need to remember and just get the signup process down to 3 screens. So create username and password, put in your data, you get charts. That's it. But in order to do that, nobody owned it at Amplitude and ownership was distributed a lot of places like, well, this team owns account signup and this team owns this marketing funnel and this team owns the data setup. And it was like, okay, for this to happen organically, it's going to take way too long. I just need to be very deliberate about what I want to drive and rally the team to do it. And so we ended up making massive changes to the interface. About 4 months, launched that last year, the Amplitude Made Easy, um, and dramatic difference. We got 40% more signups as a result of it on a weekly basis, 40% more companies sending us data every week after signing up. Um, and so massive step change in the adoption of Amplitude and the ease of use that would not have been possible had I not been in the details driving it. And that's not like a lot of people actually had the same aspiration I had to, to, to make, make that change to the experience at Amplitude. I remember talking to our head of design at the time, and she very much wanted to, to make this change happen as well. But it was like there's just too much coordination required, and unless I was driving that as CEO, it was not going to happen. So it was a kind of simple example of how building the right executive team can set you up to be very effective in the details.

So the narrative around founder mode without— before me hearing you say this, I might have thought that you would have just said, I'm doing the onboarding, everyone else, like, this is my thing, I'm taking care of it. It's just getting done. What you're saying, what actually happens is that you have a team that like assisted you in doing that and you probably had other things.

Oh yeah, I would not have been able to do it myself.

Absolutely not.

Okay.

So it's— you really think that the best way to do it is, you know, you're— you still care about the product, you care about the business, whatever you're trying to drive results, whatever. But you, you set up the team in a way that actually allows you to do that and and not have to, you know, be solo person founder.

You're the only person in the business. No, no, no. You, you can't, you can't, you can't. Great. Like, I, I think a few things on it. One is that, um, great executives can harness, um, your, can work with you to figure out, okay, how can we make this vision that you have a reality?

Yeah. Cause your CPO might actually be way better than you at all this stuff.

You're like relying on him to like, you're using his expertise to get things done faster.

Totally.

He actually did a good job because I, you know, I was frustrated about this. And so I'd go into product development leadership meetings. I'm like, what the fuck? Like, how do we change this? Like, it's not acceptable. Like, it's like, well, this team controls this, this team controls this. How do we do it? How do we prioritize it? I'm like, man, I'm going crazy here. And what he did a really good job is like, okay, what we need is a forum that's dedicated to this, that has the right stakeholders in it, and that the context is set about what we're trying to drive. So let's set that up for you, Spenser, and then you can go in and drive. And that's been incredibly effective. So anyway, that's a simple example. I think the other part about it that people miss is that it's perceived as micromanagement, but that's actually not what I'm trying to do. I'm trying to work with everyone in that room as a peer. Like, I want them to tell me, for us to debate ideas back and forth so that we're coming to the best one. Like, one of the things actually that I pushed at the start towards eliminating that we came back on because I saw it was actually the right thing was user templates. Templates, so guides for creating a funnel or guides for creating a retention chart. And I was like, man, templates, no one wants to click through a wizard, have a series of steps they have to do. What they should do instead is we should just show them the result out of the box and they can mess with it. But the team ended up coming back to me and showing me a whole bunch of data that showed that, um, they were actually some of our highest used and most engaged with features and people who used it retained a lot better. And it was like, oh, okay, this is actually, wow, people actually value this. Um, because maybe I'm trying to oversimplify the experience a little bit too much in this direction. So that was a great example of like, yeah, you know, it's not like I'm not, I'm not, I wasn't religious about any parts of my view. You know, I really wanted to debate, um, with the team about how do we create the best experience.

End experience for customers. Yeah. It's nice that you're a, you're a company and a product around helping people use data to make decisions. You use data too. Yeah.

It's good that you're still doing that. The part that's so funny, well, I'll tell you, you know, this is like the eating your own dog food part, is that even then it's hard. It's like, it's so hard. It took us a while to tease out the fact that the lift we had got from the changes in the product process were 40% because we,. It was like we did a marketing campaign and then we rolled them out with a big announcement and launch. And so we saw like a spike up where it's like 2x the use and we're like, yay. We're like, okay, but how much of that is from the launch versus how much of the last is from the improved signup process? And like it took us, it took us like some, some, you know, it's like you had to get really detailed nuance. And so, um, yeah, so that's part of why I say the space is so early because even Amplitude is as easy as, as focused as we've been on making it easy, it's still way too hard to use these data tools. And even we're the poster, we're the poster child that of, hey, we drive self-service among your organization.

Yeah. Well, and speaking of like using data tools, I know like back to the original founding of the company, you were working on something else.

Yes.

And, and that's how Amplitude came to be. Like how, how did this all kind of start way back in the day?

So we first started, so we first started, my, my co-founder Curtis and I first started this company called Sonolight, which is a voice recognition app where you could talk to your phone and send and receive text messages. And it was obvious to us at the time, this was 2011, it was obvious to us at the time that you wanted to look at the data to figure out whether users were successfully using your product. We wanted to know, okay, if you have a successful first voice recognition event, what's the impact on retention? We ended up building out all this data. You know, like a lot of engineers with hubris, we built out all this data infrastructure to help us answer the question and spent way too much time on it. But we actually found out some very interesting things.. And the crazy thing to us is, as we talked to other companies, no one else had, um, no one else had access to this data. And so we're like, okay, maybe a company can come out of this. So when it came down, came time to wind down Sonalight, uh, we ended up building, um, you know, we ended up, uh, saying, okay, there's a huge opportunity on mobile analytics here. And then mobile analytics became web and mobile and then product and then, you know, digital analytics that, that we are today. And you were in YC?

Yeah, we did, we did Y Combinator, uh, winter 2012.

Okay.

And I think there's like an interesting story. I remember seeing like you, you were going to like give people back the money or something if they wanted.

Wow.

Yeah. Well, I don't remember exactly the full story.

I just like, so what, what exactly happened there? Oh, um, so, so wow. You have a good memory. This is a long time ago.

So we, uh, we had fundraised from a few angels.

This was before, after, during? This was, so, so we did YC, um, we went to Demo Day. I did a demo on stage of Sonolife, really impressive demo. Like I had my phone, I put it in my pocket. I'd like do this huge thing about talking to my phone and, you know, like I'm from the future. It's amazing. Um, and we actually got the most press coverage. It was us and this company called Pair or Couple, I think, um, that like had like, it was like a texting app, you know, or sexting app between a couple. Um, and, um, we ended up raising a little bit of money. We raised like, we got $150,000 just from being from YC and then raised $120,000 on top from 3 angels, including one Adam Draper. And after doing that, we kind of realized looking at the data that like, It's going to be hard to make a successful scaled consumer company out of what we're doing at Sonolight. It's just, it's just the, the voice as interface wasn't quite there. And it was, it was not clear if it was going to take off. And so, um, we said, let's go do something else. And so we were like, well, hey, we raised a bunch of money and we don't feel right. And so we should maybe give this back because like, we're not like doing the thing we originally said we're going to tell all these investors. Um, and so we ended up, um, we, we asked Paul Graham and Y Combinator and they were like, no, like, what, what are you talking about? Like, don't give the money back. Keep the money. Like, you know, they've already written it off at this point. And like, better, better you have it to keep going on what you're doing next.

And, and by the way, they're investing in you, not really in the idea at this point. Did you not really appreciate that at the time?

Like you thought they like had this thesis on voice recognition?

Yeah.

Yeah. I don't know. They wanted to invest in that for some reason. Really, really, they're just. I remember one of the angels, I won't say who, but one of the angels I was convinced invested in us because we were the lamest company out of the batch. Or like they invested in like 3 or 4 companies, all of which, one of them including us, and I was convinced they were investing in those because they were the lamest companies out of the batch. Was like the valuation low or something? Like they thought like— I don't know. Maybe I think we were all on the younger side too. And so that was part of the appeal. It's like, all right, these kids will figure it, you know, figure something interesting out. I couldn't tell.

Okay.

We ended up pivoting that company into Amplitude. You know, of course, all these years later ended up doing incredibly well. So, you know, thank goodness for them and thank goodness for us. We didn't return the money back at the time. Yeah. On a per-share basis, I think it was—

they invested at $0.04 a share in the seed round.

Okay. And it's like $11.60 or something as of— Yeah, we're like $11 or $12-ish right now. You know, our peak was at $80 at one point. So, You know, they made out just fine. Yeah. You know, got like a more than 200x return on their initial capital.

Yeah. Well, and so I think you mentioned too that it was a real grind when you raised the first round, like starting the company.

Yes.

Can you just like talk us through that and I don't know how it went, how you got it done?

Like just, you know, tell people what it was like. It was brutal. I mean, that was the first time I wanted to— I wasn't sure if we were going to make it. And I felt like, you know, I just don't want to do this anymore. Um, but you stuck with it, obviously.

I did.

I did. Um, I, um, yeah, you know, so we were trying to raise $1 million. We had been working on Amplitude for a year. We didn't have any customers. Um, we had a bunch of people using it, but no one paying us. Um, and really the only credibility we had at that point was we had the vision and we had the team. So, you know, it was myself and Curtis and Um, this was before Jeffrey joined us and I remember I'd just go out every week and I'd pitch a bunch of VCs and angels and get a bunch of nos and then it was like, you know, repeat that in and out.

And so, so what was like the pushback that you got?

Like why weren't people investing? Uh, it was the idea that mobile analytics was too late. You know, we were too late. Flurry had already taken the market and you know, what, what we were doing was, was too late to it. That was one. I think another one was, you know, it's like we didn't have any traction really. So, you know, people were right. I was also not a particularly good— I didn't know how to sell what we were doing and pitch what we were doing. That took a while for me to figure out. And so, you know, I'd go out, raise, try to raise money, get a bunch of rejections each week. Then randomly, you know, I might get 2 weeks later I might get $50,000 to come in.

And it was like, fuck, it's like a long way between $50,000 and a million.

Um, it was like 5% of the way there.

Yeah.

5%. There you go. There you go. And so, you know, of course we do things like, oh, we count, you know, the $270,000 we had already raised under sunlight and be like, look, wow, we're making progress in the fundraise. Yeah. Um, and we're like, anyway, so it was, it was brutal. It was absolutely brutal. I think it was The story I always told people at the time was it took 6 months, but I'm pretty sure if I look back in retrospect, it was probably like a 9-month fundraise. So we were very, very unhot and unsexy at the time.

Now we ended up, we did end up getting it done. So how did you get it done?

Like what was convincing people? I don't know. I still don't know. I think I eventually got better at pitching it. I think the momentum of the round ended up carrying itself. You know, we got to a tipping point around $650,000 where like a bunch of people are like, okay, look, this looks like it's going to get done. I better get in now.

And then we ended up going from, I $2 million to raise $2 million quite quickly. Super quickly. So because there's usually there's like a risk, like if you've raised nothing and I give you $50K, that might be like a month of runway.

And if no one else invests and you get no customers, like the company dies. It's actually really risky.

One VC was like, yeah, I'll give you $200K, but only if you could raise $1 million.

So you're the worst.

I hate when people do that. I mean, what you got to do, like what I do is I tell people I'm in and then I just help them raise the rest of the money. That's awesome. That is awesome. Good for you. Well, and sometimes strategically, It's like, I'm in and like, we'll sign the safe and I'll send you the money. And then sometimes people come in at a higher valuation and you actually get a deal because you like, you came in really quick.

Yeah. No, that's the way to go. That's, that's kudos to you for not putting any entrepreneurs in that situation.

Like this contingent one, it's like the worst. Yeah. I, it's just, I like, I mean, I know you, you said like you follow me on Twitter, like the memes and stuff. I mean, that's just like low-hanging fruit of like making fun of bad behavior.

Oh, there's so much. I mean, VCs are like the, like, I remember I told our first VCs, I'm like, oh fuck, I fucking hate VCs. Cause I was just being real with them for a moment.

And they're like, not us, right? And I'm like, no, no, no, of course not.

Yeah.

You guys are great. So then you, so you mentioned like you didn't actually have any customers. How did you get people to start using it?

Like, do you remember like those days of like a couple first couple customers? Yeah. So I, um, one of the first customers we signed, uh, the very first, so, so before that we just begged people to use it and we just basically give the whole thing away for free. And it was, it was, um, you know, none of those were paying us and I, I kind of knew not many would pay us. I remember arguing with this one startup about like, hey, can you give us $50? It's just like, what are we doing here?

Wow. What are we doing here?

They wouldn't even pay you $50? Yeah. Yeah, they wouldn't. They're like, well, maybe we need to see more value. And it's like, you know, just, just your time meeting with us is like not, it's not even worth that. So it's worth more than that. So like, come on. Like, um, so what we ended up doing was I remember I got introduced to this one company. Super Lucky Casino by one of the investors who was checking us out. They ended up investing. And so I remember I show up to the meeting with the co-founder, CTO of this company, this guy named Brett Terrell, and we kind of go through the demo and we spend 30 minutes showing him our little crappy analytics product. And we get to the end of it and he asked me a question that had never been asked before. He was like, "How much does this cost?" And I was dumbfounded and I'm like, "What the fuck?" Like, "What? What'd you just say? How much does this cost? Oh, okay. Oh shit. Okay. All right. I better come up with a number." And so the first number that popped into my head, of course, was $50 a month. That's what you pay for SaaS, $50 a month. Month. Okay. Like, wait, wait. I remember Patrick McKenzie's advice to always charge more. And I'm like, all right, let me think.

What's— let me double that and go to $100.

So this is all within the course of like 2 seconds. Yeah, 2 seconds. So I'm like, so I'm about to say $100. I'm like, no, no, no, no. I really got to push it here. I'm like, I need to think of a number that is bigger than $100. And so the biggest number I could think of at the moment was $1,000. It's like 10x that, go to $1,000. Okay. And so I like $1,000 a month. With as much confidence as I could muster. And to my absolute surprise, he's like, wow, that's so cheap. Like, great, we'll buy it. And I'm like, oh my God, this is incredible. All the, all my dreams of being a professional software engineer that creates software that someone pays for, I've been fulfilled in that moment. Someone decided to pay me money for software. And so, you know, I now have created a tiny, tiny, tiny amount of value for someone. And so all the years of toil and, you know, all the missed opportunities and, and all the money that had been put into us and belief that had been put into us, uh, has been validated in that moment. Um, and so I was just, yeah, I was like, all right, I've achieved everything I've set out to achieve emotionally of starting a company. The rest is just, it's just gravy from here on. So that, that was the first, that was the very first. And then we started, you know, then we launched the company a few months later and then we started getting on the train Um, we actually found, um, and the, so the reason Brett wanted to buy us is because he had been a product manager at Zynga before and he was used to having these amazing dope-ass analytics with like, you could slice the users all sorts of ways. Did they build a bunch of stuff internally? Yeah. So Zynga had built a ton internally, but he didn't have any of that in his own company. Um, so he's desperate for anything. So he's willing, you know, to use the software from this like little two-bit shop that had like no product or customers or nothing to speak of.

Yeah.

He probably had just like a full-fledged team of like 10 engineers just like all day building custom analytics. Totally. Yeah. He was desperate for this. Yeah. I mean, they were like a, they were like a small 50-person company at the time, but he was desperate, you know, he, he wanted this problem solved and he was willing to pay, you know, thousands of dollars a month to figure out how to get it solved. So anyway, I, Brett, I just want to, I appreciate you all these years later. Thank you for believing in us. And, you know, that, that gave me the confidence to go out and do the rest of this business. Did he know he was the first customer? Yeah. Yeah. Or I don't know if he knew at the time he was, but yeah, you know, pretty soon he figured out and we took good care of him. You know, we constantly built stuff for him. We'd be very responsive to everything he wanted. You know, we talked to come by regularly and helped him and his team out. And so it worked out. It worked out great for both sides. But it was, you know, and then I realized I needed to—

hey, I need to go become a salesperson and start just selling software all the time. Got it. So then what did you find that worked then? Like just like getting in a, like what's a good system for thinking of, you know, you just described getting the first customer. It sounds like you didn't really have a system.

You didn't have a process. Yeah. Well, I'll say a few things about, about turn. So one of my mistakes was spending too much time building the product because Curtis and I, and then Jeffrey were all engineers and you know, we know how to solve problems that, you know, people have a problem, we can solve it. But the going out and selling, that was foreign to us. And so, you know, we were very reluctant to spend a lot of time there, and that was a mistake. Um, finally, after Jeffrey joined us, I switched full-time. I switched from building the product to full-time selling it because like, and so one of my recommendations for early-stage teams is spend half your collective founding team's time on selling the product. Like build half the time, talk to users, sell half the time. Like, you know, it's like, that's where you should be allocating your time. Too many engineers are like all in on this bucket and very little on customer development or sales or what have you. I got a sales coach, this guy named Mitch Morando, who would give me tons of coaching and advice. You know, he, he, he— I remember he'd meet with me every week and he asked me this one question. He's like, we go through the list of customers talking to me. He's like, what's the business pain? And be like, well, they want to make some charts or some SQL dashboards. And he's like, Spenser, that's not a business pain. What's the pain? And so he'd constantly be beating me up, asking me what's the pain. And so finally, you know, in these meetings, I would ask these ask customers about like, well, how painful is this? Um, you know, like, why, why do you need this? Like, what's the pain to you as a, as a company? You know, how would you rate this on a scale of 1 to 10? Um, and then I started to learn like, okay, hey, the sales process is not about demoing your product. The sales process is about finding people who have the pain, who have enough pain and care about, care about, care about the pain. Um, so, um, you know, anyway, we then started selling, you know, next customer was $2,000 a month, next customer was $3,000, then it was $4,000, then it was $6,000, then it was $8,000, then it was $10,000 a month. You know, we were eventually sold to these guys, sold the product to these guys QuizUp in Iceland and had the team visit Iceland to, there's actually a small kind of tech community there. Um, and so yeah, we're kind of off to the races on, on, uh, uh, you know, grew from $0 to $1 million in ARR in the, in months. Nice. That's, I guess, compared to today. Yeah, but not, not that nice compared to even, even today. Zero to a million in 9 months is good. Oh yeah. The, the, the, you know, so yeah, you go from zero over the course of like, you know, it was a good 2 years, I think, or a year and a half at that point.

Um, to, you know, all of a sudden now you're on the trajectory of, of, of building a business. Yeah. Cause what was it? It was basically like, I think the rules back then, or like the general mantra was You get to a million in revenue in like a year or something. A year. Yeah.

24 months.

If you can go 0 to a million in 18, less than 18 months, that's good. And then it was like triple, triple, double, double. And then you eventually, and then you can IPO roughly. That was the back 10 years ago. Totally. Totally. Totally. So it sounds like you kind of did that.

I think if I remember, you did like 0 to 1, I think it was 4.5. Yeah. So 0 to 1 in the first year.

Wow.

You've done your homework. 1 to 4.5. In the second year, 4.5 to 13 in the third year, 13 to 31 in the fourth year, 31 to 53 in the fifth year.

And then it was like 80-something and then we passed 100. So nice. Okay. And then you're still growing like, is it like 40-ish percent today?

Yeah. What are you growing at today? Oh, it's brutally low. We're at 10%. We got to be growing a lot faster.

We're 60% year-on-year growth in 2021.

Now we're 10% year on year. But we, we, you know, we need to set ourselves up to grow at 20% and then 30%.

To grow a lot faster than we are.

So, but you're like profitability-wise, you have a pretty good profitability.

Oh yeah, we're profitable today. Yeah.

Okay.

Today. Yes.

Yeah. If you can stay profitable, inflect the growth rate up. Oh yeah. Rule of 40. That's the goal. That's the goal. You know, and we just brought in a CFO, Andrew, who, you know, we laid out a plan and when we're going to get to different stage gates on that. And so I want to be clear, we're still— this early days for this category in this business. Um, and so we're primarily focused on growth. The great thing is at the scale you get all, all these economies of scale, so you can actually do it quite profitably. Nice.

Yeah. And I know you, one thing you mentioned was talking to your board members. I know you were like pretty, I don't know, like deliberate about like bringing the right people on. Like, how did you think about and prioritize who you added to your board?

We, we got lucky, frankly. Um, but through that luck, we have gotten a very, very incredible group of individuals, very, very special group of individuals that have— are aligned towards the long-term vision of Amplitude and have provided an incredibly stable foundation for our success. And so I'm very, very lucky to work with the group. I mean, it started with Eric Vischeria from Benchmark, who I frankly think is the best Series A investor of this day and age by far. He's a modern version of Bill Campbell, I have learned more from him, 10 times more from him than I have from anyone else about what it means to be a great CEO at this stage. He's interviewed hundreds of executives for us and helped me think about all the different transitions we've made as a company from 1 million to all the way to 300 and as we go towards a billion. And he's very long-term oriented, and so we've been working together for over 10 years at this point. So I was just incredibly lucky to work with him. We've had Pat Grady from Sequoia who's very long-term oriented as well, impressively so, impressively so. A lot of investors say they're long-term oriented, but not many are. Most are tied to their fund cycles, but they actually changed their fund structure specifically so they could be long-term oriented. And not have to sell in the continuous 5 to 10 years. And not have to sell for 7 to 10 years. And that's starting to become a problem, especially as the life cycle companies, it's getting longer and companies stay private longer. And so, you know, they've been, Pat's been incredible. Sequoia has been incredible. Um, Jim Whitehurst, uh, former CEO of Red Hat has been incredible. I've learned a ton from him. He's a living legend. Um, Erica Schultz from president at Confluent. She's been, you know, as we think of what does great go-to-market look like at different stages, she's been an incredible, uh, help and supporter of of me and of Amplitude and kind of guiding or learning, you know, as I've been learning, hey, what does great go-to-market look like as you continue to grow? Ron Gill, like Catherine Wong, like I, you know, and we just added Tim Swell to the board who is the founder CEO of Zora. I actually saw that on Crunchbase. That was one of the things they showed. Good.

They got it right. So yeah, they got it right. Interesting. So, and then one of the things you did was you went public in 2021. We did. So you could have, I mean, you could have like sent an email to Tiger Global and they could have just given you like all the money in like 5 minutes. Yeah, they did give us an offer on a round at some point. Yes. So why did you do an IPO in 2021?

Like what was the thinking around doing that? I think once a company gets beyond $100 million in ARR with good growth, that you So I think the first thing is there's a bunch of opportunities that you can take advantage of as a public company that you cannot as a private.

So you have a liquid currency for your stock.

It's not like fake monopoly money too. It's basically priced in real time by the market. It's worth specific amount. Yeah. So that allows you to recruit, that allows you to do acquisitions, that allows you to do other things. Also, acquisition debt or raising funds that you just much more difficult to do when you're private., and then secondly, it's a, it, it sets a higher bar for what great looks like and people recognize that. So companies, customers, partners, employees, other people in the ecosystem recognize that. Um, and, you know, it's like that little edge that you can get over your competition. So we're in a crowded space. And so it's like, yeah, we'll take every advantage that we can get., to get amplitude to help stand out. So I'm a big fan of taking your company to the public markets. I think one of the criticisms is that the bureaucrats take over, you know, and it's like, oh, you're driven by the quarter. And so, but this is where you have to be very intentional as a CEO and leader about where you're taking your business and why, and not let the short-term pressures of the quarterly reporting and the discipline around it, uh, oversteer your company. And the great leaders, you know, you look at, you look at, you look at, you know, Larry Ellison, you look at Bezos, or you look at Gates, you know, Microsoft and Bill Gates, like, you know, they've all, they all did that incredibly well. Yeah.

What would you say is the right and wrong way to do that? Because I feel like, like, you guys have a good example of like, you went public, the stock went up, then it went down. Like, how do you, how do you just continue to focus on what actually matters versus like, oh, the stock price is down, employees are thinking this, investors are thinking this, keep people focused on what matters.

Um, it's, it's a test of leadership by far. And people will look at it all the time. Um, so I spend a lot of time explaining the, what the stock is. The stock is an out, is kind of an output of business, not the other way around, not an end in itself. If that makes sense. And so it's kind of like, you know, I think of it as exhaust coming from the business. And, you know, if the business does well, it'll reflect in the stock. But if you set the thing— if you set your thing up to drive the stock price, you're going to get all sorts of weird and bad results in terms of where you're taking the business.

It's very easy to run it into the ground as a result.

A lot of it's market-driven too. It's just like, is the market going to value it? Oh, yeah. Any short-term thing, it's like, you know, I was walking the team through change in stock price. It's like, oh, why'd this change? Change in interest rate. Why'd this change? Election results. Why'd this change? Hey, there's a sell-off in this category. It's like, okay, literally none of this stuff in our control. We build a great business and achieve what we set out to do, then we'll be successful and that'll reflect in a lot of things, including in the stock. But in the short term, I think it was Buffett that said, in the short term, the market is a voting machine and long-term it's a weighing machine. And so are you, if you're living, it's like a high school popularity contest. If you're living by, did you get invited to that party? And you know, are you in the know? It's like, fucking, what are we, what are we doing here? Like, it's no way, no way to run a public company. So this is no way to run any sort of organization, particularly a public company. One of the exercises we did actually before we went out that I thought was very useful was, to pretend like we had a stock price and show the team it going up and down every week. So at all hands, I'd be like, oh, it went up this much. Oh, it went down this much. Oh, it went down. Oh, now we're really worthless. How do you feel now? And the point was the business hadn't changed, you know? Yeah. Um, and we were still focused on the long term of, of what we were doing. And if you stayed focused on that, then you get, you get incredible results. There's also this thing. Because of, you know, one of the ways you can model the change in price of a stock is with a random walk. And so if you just— there's this thing in psychology where a loss is twice as devastating as the equivalent gain. Oh yeah, I've heard that before. And so if you just go down one day and you go up the next, you will feel two times as bad about the going down than you will about the going up. And even if you end up at the same spot, you are psychologically worse So it's like, why would you do that to yourself?

It just makes no sense. Yeah.

Yeah.

Cause there's like looking back, it's like the last X amount of time is like, I've just felt like shit and it feels like I haven't gone anywhere cause I'm flat.

Totally. Totally. Totally. Well, you know, you feel even worse than if you just hadn't looked at the thing at all. So I just, it's just a, it's just a, yeah, it's, it's just a, anyway, so, so there's, there's all sorts of like having a, uh, you know, I just want the price to be stable, but because that allows to defocus on it, allows you to plan around it better and all of that. But, you know, the good part is I've been really clear with the company about that. And I think people understand that we're in this to build something extraordinary over the long term and that whatever volatility you may see in the short term is not reflective of, of the opportunity and our execution against that.

Yeah. Plus, like, if you think about, you know, public versus private stock or whatever, like if you if you join a company when it's undervalued and you're getting compensated when something is like, you know, quote unquote undervalued and you're going to create value, I think like it's going to like you want to be getting paid from a stock price in equity that is, that is actually worth way more than what you're really getting. So you're getting a good deal then. So like in, I mean, this is like, this is probably like not something that it's like that exciting for people to hear when they're like, oh, my stock's worth less.

But like, if you are getting more of it because it's like the market's driving it lower, like that's actually— it could be a good thing. Yeah, absolutely. I mean, yeah, yeah. I mean, it's funny, even the Command guys, I was talking to them and, you know, the stock's appreciated a whole bunch since that acquisition. They're like, all right, yeah, we're happy.

So yeah, there's all these, there's all these funny dynamics in the market that can happen because of things like that. Yeah. What in terms of like how you've evolved from being like the founder Spenser, when you were, you know, Sonolight, I'm pivoting, starting this like product analytics idea to now it's like a public company, like the CEO of like you have a lot of employees, big teams.

Like what have you learned and kind of evolved and changed over that period of time? Oh, so much. We could do a whole— Turner, we could do a whole podcast on this.

Yeah.

Part 2 on this. What are some of the biggest? I will say being a large company executive is very different from being a founder. And most founders do not successfully make that transition. As a founder, you wire your brain in a certain way and you have to redo and undo a lot of that and redo it in a different way to be as successful as a large company executive. And what's interesting, even founders who are very successful will normally leave 7 to 10 years in the business because this change is so dramatic. Now the funny part is, believe it or not, being a large company executive is actually easier than being a founder in a lot of ways. But just making that identity shift is very difficult. So I'll give you a really simple example. As a founder, it's your job to lead from the front all the time. Like you should be out there selling the product, you should be building the product, you should be talking to customers, dealing with the fire, you know, recruiting people, all of that stuff. As a large company executive, you cannot be leading by example on everything all the time. There's just too much stuff to do in the business and you can no longer have to do, "Hey, I got to talk to this customer. I got to meet this employee. I got to do this thing on the PR side." And so you have to be much more selective, frankly stingy with your time. Whereas a founder, you're just being generous with it everywhere and you're just showing, "Hey, by sheer force of will that you're the most committed and you want everyone to follow your example." Um, but you have to rewire that thinking, and it feels bad. Now you become the person you hate. Now you're the person who no longer, you know, who's just like crapping on other people's work and setting high expectations for them and, you know, managing them and expecting them to do something while you're not doing yourself. You're no— you know, not to say there are places where you do lead by example, like I'll still be on customers, you know, I'm going to do product reviews, all of that sort of stuff. But, you know, there's a lot of places where I just can't. Can't, because you don't have leverage, you don't have time. And if you do try to do it, you'll burn yourself out. And so that is an example of the mentality shift. That's, that's, that's really, really hard to do, going from founder to large company executive. And so I've been trying to deliberately learn that over the last 3 years. And it's been, it's been, it's been quite the shift. What's worked best? What has worked best? Or like, how have you done it? How have I learned that? It's hard. It was hard. I mean, frankly, you kind of learn it because you have to, otherwise burnout. Um, I mean, we had to, we went to the public markets. We, uh, we had the stock come off, you know, from its highs by almost 90%, or actually it was 90%, $84 was the high and, you know, low was like $8. So just under $8. Um, and then rebuilt the entire executive team to be stage appropriate. Oh, did you really? And that was brutal. Yeah. If, if you look over the last 3 years, entire executive team is different set of folks. So, you know, new Thomas, president, head of go-to-market, Brent Swatchy, product officer, KJ, head of HR, uh, Andrew, CFO, Wade, head of engineering, all new, um, in the last 2, 2 and a half years. Um, and so that's, that was like brutal. So it's like, that was just like, hey, you just have to do it. Like there was no choice to, to be successful at this. Um, so part of it is just so you have to meet the moment and just go do it. I would say another part, you just talk to other CEOs, other founders who have gone through that change. You understand how do they spend their time, you understand where they focus on, they understand what you do and what you don't do. You understand, okay, how do you hold executives who are more experienced at a function than you accountable for the results? How do you run a cohesive team? How do you prioritize? How do you, you know, you see a change that you need to make in the business, how do you communicate about that change and the expectations around it. And so I, there's no formula on it. Like it, the biggest mistake honestly, it was I should have gotten in front of that change a lot earlier as a CEO and looked at what founder CEOs spent their time on, you know, between the $100 million billion mark. But yeah, you know, now I'm here, you kind of learn through it. And now we're on the other side of it. And so now, you know, not to say I'm kind of— there's still a lot more for me to learn about this, this topic. But, you know, I think we're— I think I, you know, I'm really excited about the executive team we've got. I'm really excited about the path forward and the innovation that we have the ability to make. And so, you know, well, you know, I'm sure a year from now or 2 years from now, I'm going to look back today and still feel like I'm an idiot. But I feel like, you know, it's in terms of like, okay, what does it mean to be a large company executive? I feel like I've that, that, that mind shift has changed a lot. Yeah, you, you think a lot about people and resources and where to deploy them. Um, and, and, you know, hey, is this problem get solved? If not, it's not— I can't dive into myself.

It's my job to think about how do you rally the right group around it. Yeah. And I know you mentioned there's kind of like these different phases of hiring executives. There's like, I don't know, early stage, mid-stage, later stage.

Like, is that kind of part of this? Like, how, how have you— Yeah, that's one of the more subtle things you learn over time. And Uh, so, so early stage, you want folks who are like founders who are going to make something happen from nothing. At early stage, you have no brand, no product, no customers, like not, not much, you know, talent, nothing you can do.

And they have to make something happen just like you as a founder through sheer force of will.

Build it, sell it, recruit someone on the team.

Exactly.

Just every function. Yeah. Make, make, how do you create an event where customers will come and show up even though nobody knows about who you are?

That's hard. Challenge yourself.

You can't just email somebody, just be like, hey, like event team, totally. There's no team to do it. Of it. So you get, you get executives who are very much like founders. Um, for the mid-stage, um, you want someone, then you need to start doing repeatable processes instead of just starting stuff all the place. So once you get to about $20 or $30 million ARR, you want repeatable. Uh, repeatable is the key. So it's like, okay, how do you create a sales process so that you can forecast accurately for the quarter? How do you get the right spec for the type of engineer you're trying to hire? How do you get your books in order so that You can reliably report your results. Um, for the late stage, then you get, honestly, people who their main job is actually not even running those processes. Their main job now is to hire other leaders who will run those processes. So now my job is to hire people who are good at hiring other leaders and rallying them.

And, you know, now it's like the people who are actually doing the work, a bunch of layers down.

How do you tell that someone's good at hiring? Like if you, if you, if I show up— Track record is, is, I mean, And first is, what are you going to work for them? That's like a good one. Track record of doing it before. Depth. One of the questions I like to ask executives is like, you know, tell me about the best hires you've made. And then I start testing them on how well they know these people, like what motivates them. It's like, oh, they want to make a lot of money. Like, cool. Why do they want to make a lot of money? More in depth you can go there, the better.

You see who wants to follow them from company to company.

Um, I've definitely heard that. That's a big, that's a big part of, of executives, you know, executives who have talent will have multiple people who will want to work with them again. Um, that's a strong sign, you know, it's kind of a flag if they don't. Um, so, but there's, there's not, but just knowing that that's what you're hiring for once you're past the $100 million ARR stage. That's the harder part. It's less about like, can you do all the work and scale the stuff yourself?

It's more about can you hire a great team around it?

So, well, speaking about like hiring and like team composition, how are you thinking about DEI today? So first, I just want to give my rant on this, which is tech industry fucking full of cowards as leaders. Okay. Like the fact that a few years ago, These leaders were all in on DEI and now, oh, new administration, all of a sudden we didn't actually believe all those things that we set out to do and we're going to get rid of them. And to really silly degrees, I mean, Facebook getting rid of tampons, men's bathrooms, it's just like, what the fuck are we doing here? And so it's like, okay, I'm clear, you just didn't have any principles in the first place. You only were interested in DEI At the time because your employees pressured you to do it, but you actually didn't believe in the power of diverse teams as a leader. And then now you're only doing, you're only flipping the other way because, you know, a different, you know, Trump got elected and Republicans are in power in Congress. It's like, all right, this is the exact fucking opposite of leadership. And so like, frankly, I think 90% of the tech industry and CEOs and my peers are fucking cowards on this. So that's, Yeah, put that on the podcast. Let's, let's make it, let's make it, let's clip that. Second is, we haven't changed our view on this at all at Amplitude. We've been a believer in diverse teams from very early days, we've been intentional about how do we create more inclusive environments, you know, particularly for, you know, underrepresented groups like women in tech, from from when Amplitude was very small to the last few years to today. And so we're still very, very intentional about that in terms of thinking about how we build out different teams at Amplitude, how do we build out our executive team, how do we build out our board. The second part of it is like, I frankly, I think it's just like Amplitude is a place where you belong and We want to include you whether you are trans, uh, whether you are from a minority group, whether you are a different gender or different national origin, or like whatever group you're from, you belong at Amplitude. And I think that is a very important message for individuals to hear from their leaders at this time. And so I want to make sure folks at Amplitude hear that. I want to make sure folks in the broader industry are thinking, considering joining Amplitude hear that as well. So I think it's frankly ridiculous that so many CEOs and companies have backed away from this stuff. It's complete lack of leadership on this issue. And it's like, okay, well, hey, let me at least give Zuckerberg himself a few kudos. It's like, all right, this guy just wants power. This guy's just pursuing raw power in whatever form it takes. The principles, the principles just shift to follow that. At least that guy's clear about what he's doing. The rest of them is just like, you know, sheep. Sheep, you know, there's a difference between— there's this great post online I love, um, that talks about sheep logic and it compares sheep to wolves. It makes the point that both sheep and wolves are very social creatures and very smart, very socially aware. And sheep. Wolves are very thinking about the success of the pack, willing to make sacrifices for their family members, for other wolves so that the pack can survive and do well. Whereas sheep are incredibly— they're very other-aware but incredibly self-centered. And so if they see one sheep panicking, they're gonna panic too. If they see another sheep has found an interesting patch of grass, they're gonna go try to eat it. If they see another sheep running off a cliff, they'll go join them.

And so anyway, I'm a strong believer in leadership as a wolf, not leadership as a sheep. Yeah, interesting, because you would not think in saying that sheep and wolf analogy, you'd not think of like the wolf is a leader.

I mean, you think of like the evil bad guy who's like eating the sheep, right?

Oh, Spenser, I love dogs. Who doesn't love dogs? Well, yeah, exactly. But to your point, it's like, yeah, the wolves are actually, well, I mean, if you're leading a company, do you want to be a sheep or do you want to be a wolf?

Like the leader of the company wants to be a wolf. Yeah, exactly. Hell yeah. Hell yeah. Yeah. Hell yeah. Too many sheep these days. Yeah. So, and that, and that, that metaphor applies to a lot of different, you know, there are a lot of sheep-like games like investing, you know, as they joke, it's very sheep-like. We're all sheep. It's like, hey, who else is in the round? Is it hot? Like, you know, I'm going to get on before someone else does. Oh shit. This company is in trouble.

Fucking get a better piece. Peace the fuck out. So yeah. Jesus Christ. Just delete it from my LinkedIn or my Twitter bio. I was like, yeah, Amplitude. I was an investor in that company. What are you talking about? Or I am proud early investor, like, you know, totally. You and Spenser are best friends and like, we love the company. Totally. Speaking of like games, I don't know if you just mentioned games. We're actually— we're in— we didn't even call this out.

We're actually in the Amplitude office right now. You guys have a game night literally right outside of us.

Sorry about the noise in the background. Yeah, hopefully people can hear it.

So like, do you believe that like games and gaming is like relevant to working or operating startup or building a startup? Games, gaming. Yeah. You know, there was a, there was a, there was a group of, of, uh, early employees. That's how Jeffrey Curtis and I bonded, frankly. Um, and so it can be, it can be a good mutual interest and hobby to have. Do you have a favorite game? Um, you know, I've been watching, uh, believe it or not, this will date me a little bit, but I've been watching a lot of StarCraft: Brood War recently. You've been watching? Yeah. Yeah. So they, they, so there's this tournament called the ASL, the Africa TV Star League, which is basically all these Korean pros that like are playing StarCraft against each other. They have it twice a year and the new season's about to start. And so I'm like getting pretty hyped up for that.

Interesting.

Okay. Which is like, I know, like, you know, it's a game from like 1999, but like, it's so in terms of quality of skill, it's like the most refined. Yeah. Um, yeah, it's, it's incredible. Like all of MOBAs and that whole like, you know, it was just like a seed to what StarCraft And it's such, it's just such an incredible game.

I don't think people who, who don't, people who don't play like competitive video games or like video games, like a, you know, you're like the best of the best.

It is like extremely intense. It is very, I mean, it's nuts. It's nutty how, how good some of these people are.

It's, it's great. I mean, they're, they're, they're real athletes. So yeah, like I, um, I played a lot of like Halo, Halo 3 back in the day. Do you remember Major League Gaming? Oh yeah.

Yeah.

Yeah. So like I, I actually went to a tournament once. I, and it's what made me realize I was not gonna be a pro gamer. And this is like 2010. Oh my God, I love that too. I got, we got destroyed. So we played, I played Halo online a lot. I got the MLG playlist in Halo 3.

I got a level 50 in it, which is like the hardest thing to do.

I was like, I felt like I beat the game. Oh, that's amazing. And then went to one of the tournaments where they had this like local tournament in Chicago. I was in Grand Rapids going to school. Yeah. And just with some friends from school, we like, we started playing Halo together and we went to this tournament. It's like a double elimination. You play 2 matches against each team and whatever. So we played a total of 4 matches or whatever. Yeah. 2, lost the first one, lost the second one.

Loser's bracket, lost the first one, lost the second one. Brutal. Like the skill level. Yeah.

Don't, don't go into a professional game. I'll tell you that right now. It's sort of, I described it to people as like, yeah, like speed chess because you're constantly, and it's like a team because you're all moving. Like if it's StarCraft, Halo, like you're positioning around the map, influencing spawns, like where they're going to be, where the weapons are. Where you're constantly like, you know, you're trying to force them to spawn in bad spots while having map control. And it's like, yeah, so it's like you're playing chess kind of, but it's also like hand-eye coordination at the same time. It's crazy.

Um, do you know what game you're gonna play when you get out there? Oh, I don't know what they're probably playing, like One Night Werewolf or Codenames or something like that. Oh yeah, Codenames is fun. Yeah, yeah, it's a good—

we do every Thursday night. Join Amplitude if you're interested in that sort of stuff.

Yeah, how can we We'll throw links in the description of the show, but like, how can people find you or like—

Spenser Skates on Twitter.

Yeah. Or amplitude.com. You can check us out there. Slash careers also? Yeah, yeah, yeah. There's, there's, there's, yeah, slash careers on it. Or, you know, find us on LinkedIn.

So a bunch of different places.

We'll post on Twitter and LinkedIn. Cool. Awesome. Well, it's a lot of fun. Thank you, Turner. I really appreciate it. Thank you so much. Sorry I was late and thank you Thank you for, thank you for doing this. It's great, dude. I appreciate it.

Dude, you did your background research on me is incredible. And thank you for listening and to Numeral for supporting this episode. If you missed it, check out last week's conversation with Dan Lawrence on scaling ChainGuard from $0 to $40 million ARR in 2 years. And if you don't want to miss next week's episode with Sam Ross at Numeral for a masterclass on all things startup growth, subscribe to my newsletter, The Split, in the episode description to get the episode plus a transcript emailed directly to your inbox.

Thanks again for listening. See you next time.