How Amplitude Became the #1 Product Analytics Tool | Spenser Skates, CEO & Founder

How big is Amplitude today? How many product managers are using it?

So we have hundreds of thousands of users on Amplitude, 3,000 companies on us, and we're doing about $300 million in revenue. So I'm going to show you something I've never shown publicly before, which is some of our real-life data and dashboards. He had the craziest response to me. He said, and I'm like, what? That's why, that's why this was very controversial for me to get out, by the way. Some of our product leaders hated this. We worked on lots of different things that didn't work out before Amplitude.

My mind was blown because it's easy to use, non-technical way here.

That's where Amplitude shines. So the best designers, the best product managers, the best engineers I know out there, they're all data-driven because they want to understand what their users value.

Really quickly, I think a crazy stat is that more than 50% of you listening are not subscribed. If you can subscribe on YouTube, follow on Apple or Spotify podcasts, my commitment to you is that we'll continue to make this content better and better. And now onto today's episode.

Spenser Skates, you have founded one of the most significant companies for product managers out there. It's a tool that most of the A-tier PMs I'm talking to, they're using for their product analytics. How big is Amplitude today? How many product managers are using it?

I appreciate that. So we have hundreds of thousands of users on Amplitude and about 3,000 companies on us, and we're doing about $300 million in revenue. By the way, Akash, I actually really, really liked how you asked that question of how big Amplitude is today, because when most people ask me that, they're always like, how many employees do you have? And that's like the stupidest way to think about the question because like anyone can hire hundreds of people, but it's much, much harder to build a real business and create real value and get people to use it. So I appreciate it.

Of course, winning the enterprise like that and winning all of the best accounts is totally mind-blowing. But I want to start at the very beginning of your journey when you were at MIT. What are the significant things that happened while you were in college that led to Amplitude being founded?

The path to get to Amplitude, believe it or not, was, was actually a long one. We worked on lots of different things that didn't work out before Amplitude. And so, so just, just as an example, the story I tell most often is the story of Sonolight, which was the company before Amplitude. But even before that, there was tons of different things. We worked on a website for our class alumni to help them understand where different people were going after college. Like everyone kind of disperses. You want to keep in touch while people are in different cities. So we worked on a website, like a kind of social networking website that did that. We worked on a website for photographers. We worked on an outsourcing website. We worked on all sorts of stuff. And we— the first real thing that got traction was this company called Sonolight, which was a voice recognition company. That allowed you to send and receive text messages by talking to your phone. This is actually before Siri came out. So we had developed this technology that allowed the phone to be listening in the background all the time and allow you to talk to it and the phone to talk back to you. And it was a magical experience. It was like, whoa, like we're living in the future. This is how, you know, this is a totally new interface for technology. The interesting part about that was um, actually the transition from, uh, going from being a college student to, uh, starting a company. So just, just for context, I graduated from MIT in 2010, and, um, at that time I was lucky enough that I could do a lot of different things in my career. I could, uh, go join a big tech company, I could go into finance, I could take consulting, I could do a lot of different paths, and what I was trying to figure out was, okay, well, what's the best way for me to channel what I think is important in life? Because your career is a huge part— it's not the only part of that, you know, there's other parts. There's your family, there's your community, there's other pieces to life too. But your work is, is in a lot of ways the biggest way that you make an impact in the world. And because I had a lot of different options, I wanted to, uh, figure out what the best way to, to to channel that energy was. And at the time, a, a few years before me, the founders of Dropbox, Drew and Arash, had come out of MIT and basically started that right out of school. And I was blown away by the success that they had. And I ended up doing this competition called Battlecode that, that got me into that, that got me into the world of top programmers at MIT., and through that, you know, I met Dropbox. I met the, the founders of Dropbox. I met Drew and I met Arash and I met Albert Ni and I met a bunch of the early team there. Um, and I met some other companies that had been started around outta that. Like there was this another one called Etherpad that was started by Aaron Iber, who was actually one of the, uh, the original developers of, of the Battlecode competition. And it was like, wow, look, if these guys can do this and figure that out, a college student, maybe I can too. Um, And so as I started studying this world of startups and technology, I was like, wow, like it's actually possible to create something massive out of this just by showing up and trying it. Like, you don't need any formal training. You can— if you're willing to just show up and figure it out for long enough, you can find success. One of the really consistent themes I saw was that any startup that had been successful normally had a period that was about a year or a year and a half in where all things were— looked like they were going to be terrible and they probably should have given up by all rational means. Like if you look at the founders of Airbnb, if you look at, you know, most of even the very successful YC companies, they had gone through that. But for whatever reason, they didn't. And so I looked at that and I said, man, if I just commit to this and I try at least for a few years, I'll probably make it worthwhile, uh, my time. Um, and it was hard because, um, you know, especially as an engineer, engineers love the safe path in life. They, they absolutely love the safe path. So, you know, a lot of friends of mine that year, maybe they just thought I was lame, who knows? But a lot of them were like, yeah, I'm interested., but you know, I wanna work first and you know, I, I don't know about the startup thing, Spenser. And so I had a lot of friends that went off to Google and, and, uh, some that went into finance and high-frequency trading and, and some that had, uh, gone to, to other companies. Um, but the one person who, uh, who was willing, uh, to work with me and said, okay, hey, actually, I actually really do wanna learn this, is, is my now co-founder Curtis Liu. And he, we were the same year. Uh, he was like this really outstanding algorithms programmer. He was like probably the top algorithms engineer in our graduating class. And, uh, it was funny, he was actually, uh, a teacher's assistant on a graduate level algorithms class, uh, his junior year of college. So it was like, you know, really outstanding. This guy is like, you know, orders of magnitude more genius than I am. Um, and so Curtis was like, look, I want to do this because I'm really interested in learning what it takes to build a company, but let me, I wanna work for a year first. So he goes off and he goes to work at Google. And because no one wanted to start with me, I'm like, all right, I guess I gotta go get a real job. So I, I went into, I took a job in this place called DRW where I did high-frequency trading, except on nights and weekends we would hack on stuff together. So we'd work on the bunch of the ideas I mentioned. So a website for our alumni of our school or, a, you know, a website for photographers or an outsourcing website, all sorts of, all sorts of projects. And other people would join us occasionally, come in and out of the group. But Curtis was the— Curtis and I were the two who were very consistent that entire time. And then after about a year of working in finance, I called up Curtis and I was like, hey, I'm leaving my job. You know, it's been a year. You were actually going to do this, right? And he's like, yeah, yeah, yeah. Okay. Okay. So I come out to California and I end up going to live on his apartment floor on a mattress for— yeah. And it was such a transition because you go from, you go from this job where you have a boss, you're getting paid a lot of money. It's recognized. It has high social status. Like my mom would brag to her cousins in, you know, I'm Chinese. So, you know, it's a big thing to compare, you know, between your relatives. Okay, how much is my son making? How much is my daughter making? You know, do they own a house? Do they own a car? So, you know, I go from that to like now I'm literally have nothing. And so it was so, so that, that part was hard. But the harder part actually for me personally was, you go from an environment where it's like you're kind of on rails and you just kind of show up and follow the path and you have a boss that tells you like, you know, hey, if you do this, you'll be successful to like literally nothing. Like you're, you're, you're, you're setting your own path. And it was so bad. So at that time, Curtis was still working at Google because he hadn't fully left yet.

This episode is brought to you by WorkOS. If you're building a SaaS app, at some point your customers will start asking for enterprise features like SAML authentication and SCIM provisioning. That's where WorkOS comes in, making it fast and painless to add enterprise features to your app. Their APIs are easy to understand so that you can ship quickly and get back to building other features. Today, hundreds of companies are already powered by WorkOS, including ones you probably know, like Vercel, Webflow, and Loom. WorkOS also recently acquired Warrant, the fine-grained authorization service. Warrant's product is based on a groundbreaking authorization system called Zanzibar, which was originally designed for Google to power Google Docs and YouTube. This enables fast authorization checks at enormous scale while maintaining a flexible model that can be adapted to even the most complex use cases. If you're currently looking to build role-based access control or other enterprise features like single sign-on, SCIM, or user management, you should consider WorkOS. It's a drop-in replacement for Auth0 and supports up to 1 million monthly active users for free. Check it out at workos.com/akash to learn more.

That's workos.com/akash.

Today's episode is brought to you by Amplitude. Building great digital products is hard. You know that better than anyone. Getting teams aligned, measuring what matters, and scaling your product strategy isn't easy. But what if you had a clear framework to guide your next steps? That's exactly what Amplitude built. They studied the best product teams to understand what really drives impact and turn those insights into the Digital Experience Maturity Assessment. In 2 minutes, you'll be able to see where your team stands and what you can improve to build better products faster. Click the link in the caption to take the free assessment and get a clear path to product growth.

And so, uh, what would happen when I was living with him is I, uh, would wake up really late in the day. So I'd wake up at like 11 or noon because I had stayed up, uh, late the previous night. And then I'd play like StarCraft for like a bunch of hours and it was so bad. And then at like 4 o'clock, I knew Curtis was coming home around 5:30. So I'm like, man, I better actually get some work done on this company so that like, you know, he thinks like we're actually gonna make something of this and, and it's legit. And then he'd come home at 5:30 and I, and I'd be like, Curtis, look at all the amazing things I built today. And like, I did this and I did this. And it was like only like an hour of work. But I don't know. I don't know if you could tell or not or what, but, uh, you know, and then we'd work on stuff together for a few hours and then he'd go to sleep around like 10 and then I'd stay up till like 2:00 AM playing video games again. And that was just on repeat for like months, months at an end. And so it caused me a little bit of like, man, am I really cut out to do this thing? Now, uh, fast forward a few months and, and he actually, to his credit, he actually left his job at Google. Um, and we started, uh, working on stuff full-time together and we scrapped the previous stuff we had worked on and worked on Sonalight, and that was, that was a dramatically different, uh, setup because now it's like, you know, I had someone else who I was working on who, you know, I could feed off their energy and was depending on me and we collaborate on stuff together and it was exciting. And so then we got a lot, a lot of stuff really done and, uh, yeah, we, um, you know, got into Y Combinator, we launched Sonalight, we got a few hundred thousand people to download. We're like, holy shit, someone's using this. That's not us. Okay, this, there's something here. This is cool., and, and we're, we're off to the races.

There are so many lessons there, but 3 or 4 of them that I would highlight, and I'm curious if you would agree with starting with 2 really talented engineers who can build stuff fast and put it out into the wild, getting commitment for both of you guys to go full-time and leave pretty prestigious, cushy jobs. And then the third one, it seems like, is a really high willingness to pivot. And that's where I want to double-click on is how did you guys pivot into the idea for Amplitude and where did you find this core problem to solve?

Yeah, we, I mean, we were looking to work on something that had traction and after working on lots and lots of different things that didn't, it was so, so the how we came on to analytics in particular was it was obvious to us building Sonalight that the way you would understand how to build a great product was by looking at what your users did, because that would tell you where they were successful. That would tell you where they're frustrated. That would tell you how to keep them coming back. That would tell you where their issues in the product were. And it was a much— it was, it was just obvious to us that that's how you should do it. Right. You know, there's only, so many users you can directly talk to or see. And so the best way to understand that at scale is to look at what they're doing through analytics. And the crazy part was while there were a lot of free analytics on the market at the time, there was Google Analytics, there was Flurry, there was Appsalar, there's Localytics, there's Kissmetrics, there's Mixpanel. There was, you know, there's tons and tons of companies out there that were doing similar sorts of things. But the only one there was, sorry, there wasn't anyone that was capable of telling you why people were doing what they were doing. Like we had these questions like, okay, how does whether you have a first successful voice recognition event predict long-term retention? Or what are the 3 things someone does right before they churn? Or What do my best users have in common? What are the patterns that my best users go through? And nothing off the shelf could do that at the time. Um, and so, you know, like a lot of engineers with hubris, we had, you know, just two of us building something ourselves. Um, and so I kind of fast forward to a year into Sonalight when we realized Sonalight wasn't going to be that successful. You know, it was okay. Had a few hundred thousand downloads, about 10,000 weekly active users. So somebody was using it, like it's great, you know, it's not just us and our friends. Um, but we knew that the technology wasn't there for this to break out as an interface. Voice was still a little early at that point. It's still early as a, as an interface if you ask me. Um, and so we knew we had to go work on something else. And what was interesting when we talked to a lot of other companies in YC was they all had the same problem we did. They all wanted to understand how customers were using the product, where they were getting stuck, what kept them coming back. They all wanted to understand, um, what was, what, yeah, what was going on. And they weren't able to with anything off the shelf. And so that was like, that was a signal to us that there was an opportunity there.

And did you guys try out some of the existing tools like Kissmetrics, Google Analytics? And I think the other elephant in the room at that time probably was Adobe, but it might not have been a fit for an SMB that you guys were at that time.

Oh, totally. Yeah. We, we tried a lot of them out. I mean, Flurry was the big one. So we were a mobile app. And Flurry was, you know, this, this, this will date myself, but for people who were building mobile apps in 2012, Flurry was the thing to use because they had done a really good event-based structure and had great analytics. And I could tell you about it. They ended up selling a bunch of years later to Yahoo for $300 million through their ad networks. Their analytics wasn't even the valuable product, but that was kind of the gold standard., but we tried that out. We tried Google Analytics, of course, which is kind of the default on the web. We tried, um, Appsalar, we tried Kissmetrics, we tried Localytics, we tried Keen. We tried a lot of different things. And what was interesting is I think a lot of them were just like, hey, we're going to give you base analytics that can count stuff. So it can, you know, tell you how many active users you have, tell me how many users are doing a particular event. Maybe if it's really advanced, it can do a funnel, but probably not well. Uh, definitely can't tell you retention. It definitely can't cohort your users by behavior. And so that was a clear signal to us that like, okay, we gotta go build this ourselves. That's awesome.

And when you built it internally for you guys, what did it look like? Was it for those key questions you mentioned, like if they had a successful recording, what's their retention? What are the things they did before churn? Or how did you build the very first kernel of the product?

Yeah, so the first question I wanted to answer was how does a successful first recognition event predict long-term retention? And so you kind of need to have two different data points. One is did they have a, did a user have a first successful recognition event? And the second is did they come back two weeks later? So we accumulated this dataset over a few weeks and then we ran the numbers and it turns out enormously predictive. A first successful recognition event doubles the chance of long-term retention. So that was massive. So that meant, okay, yeah, the accuracy of the voice recognition, you know, matters a lot. And it might sound obvious being a voice recognition product, but knowing how much it mattered so that we could understand what the marginal effort, the marginal impact of effort in terms of long-term engagement in the product was. And so that, that was kind of a big, a big breakthrough. Then we started to ask all sorts of other questions. Okay. You know, where do people get stuck in the onboarding funnel? What are the most likely things someone is going to do right before they churn, and so on.

And, and that opened up a much deeper understanding of how people used our product. So really deeply feeling the problem yourself seems to be the critical takeaway here, and then validating it with small little solutions that aren't like building the entirety of what Amplitude is today, but very specific questions How did you make the transition to building for yourself to building for others?

Um, so one of the mistakes we made with Sonalight was not spending enough time talking to our users. 'Cause we are engineers and the easy thing as an engineer to do is to solve problems, to build a software that solves, solves problems, 'cause that's what you know how to do. And that's what you know how to solve. Um, but we spent very little time, maybe less than 10% of our time talking to users, and we should have spent a lot more on it. Uh, because that would have given us a much deeper appreciation for who they were, how they used our product, what would be valuable to them. Um, in fact, my recommendation now is founding team, you should always spend half of your collective founding team's time talking to customers, always half of them. Of time. And that's a lot, that's a lot of time. And it's like, well, you couldn't, you get further if you're building something. It's like, well, no, like you don't know if you're building the right thing unless you're talking to people. And so that, that, that is the kind of constant signal of, are you going in the right direction and building the right thing? So anyway, so, so with Amplitude, one of the things we did before even building anything is we went to talk to 30 different prospective customers, companies that, potentially had this problem before building anything. We're like, you know, we're not gonna make the same mistake of building a bunch and like, we're just gonna ask people about their problems and how painful this is for them. And so we, we talked to 30 different companies and out of those, maybe 10 of those had this problem in a way that was acute enough that we felt if we solved it, they would use us. So we're like, all right, that's good enough for us.

Let's, let's, uh, let's go ahead. And did you start with the enterprise, which is where I think you really have a strong foothold now, or did you start with SMB?

No, we started with SMB. We talked with other technology companies like us. So we started with, in the first place, with other YC companies in our batch. So they had just honestly, we had begged a lot of them to use our analytics. We didn't even ask for money for any of them.

And then, then, then it kind of grew to other companies that those companies knew from there. So that awesome YC distribution, even though YC, you were there for Sonalight, but the connections you, you, you had built, you use those to get the initial users for Amplitude.

Totally, totally. Yeah, we had, we had, yeah, gone through YC, done Demo Day as Sonalight, and then right after the batch ended, uh, changed into Amplitude and said, okay, let's, uh, you know, let, let's, uh, reconnect with everyone again and see if they're, they can use these analytics.

So there was this phenomenal rise you guys had from that point through to eventually IPOing. What were some of the hardest failures or roadblocks you faced along that journey?

Oh, so many, so many. Um, hiring out, um, the initial team and kind of going from a founder to a team leader, that was a, that was a big change. Um, hiring the first set of executives, that was actually probably the hardest because you go from people having a bunch of frankly 20-year-olds who are all inexperienced and don't know what they're doing to a, um, now you're hiring people with experience who are frankly better leaders than you in a bunch of ways. And all of a sudden it caused me an existential crisis of, uh, uh, of what my role was. Uh, then, you know, another really hard part was the first time I had to change out the executive team, and that was, that was incredibly tough. Um, yeah, now to being a large company executive, it's very, very different from being a founder. So I'm happy to, happy to talk about any of those.

I think the one that people will really want to hear about is changing out the executive team because Hiring people, you know, it's always fun. Like, okay, I'm bringing you along now, I can afford to pay you.

But firing them or making changes is a lot harder. It's so hard, Akash. I mean, that is, yeah, that's the hardest. I, you know, it's for, before I tell you about my experience, actually, I remember coming across this clip of Larry Ellison from Oracle and, you know, if you know Larry Ellison and you've seen him, he's not, he's not a sentimental guy.

Incredible.

Very ruthless, very, very aggressive. Um, and he was asked in an interview once though, what the hardest part of the Oracle journey for him was. And he, he said the same thing, which is, you know, he had gotten the company to a certain point, but he realized the set of people he needed to get it to further were a different set. And it was so hard for him to say goodbye to those people because he had so much loyalty to, to them and, and the work they did to get the company there. So it was, that was, that was hard. That was incredibly hard. Uh, that was incredibly hard for him. And that was the hardest for me too. Um, I, I think the, the intellectual part of, hey, you know, we're not scaling and this thing is broken, you know, that's not hard. You, you often recognize that in the business, but the emotional part of, you know, one, you know, you wouldn't have been able to make the company successful without a bunch of these folks. Two, it's like you're barely struggling to keep up with the growth of the company as the founder CEO and feel out of your depth. And so who are you to judge someone else? That was actually really— that was the hardest part. It was like your job as a CEO is to judge someone for something you can't do. And that's like, uh,, you know, we have this value at Amplitude, humility, uh, which is all about you're looking to learn and you know other people have more, have perspectives and skills and insight that you don't. And so you want to come at with that interaction with humility, um, because you could be wrong. And, and so that was a really core Amplitude value from the start. But what made it, but it, it made the, the changing of executives very, very hard because it was, it was, really difficult to, you know, for me as CEO to say, hey, look, you're not the right person for this next stage of growth and I need to bring in someone else because, you know, we're not scaling. And that was, you know, I was too slow to it. Like all founder CEOs, those changes to that, like all founder CEOs are actually one of my executives kind of realized it themselves and they actually came to me and was like, oh, you know, let me, let me go. I, you know, why don't you go find someone else? I'll transition out. This isn't the right thing for me. But I was like, I doubled down and tried to save him, which was totally a mistake in retrospect. I should have been like, no, no, you're right. You know, I, that's very mature of you to recognize. Um, and so that was so incredibly hard, you know, and then there's even, how do you even communicate it to these people? Um, you know, cause like, you know, you don't feel confident in, like, you feel confident intellectually, but like you don't feel confident emotionally with being the, the person who makes those judgments. So that, that was incredibly, incredibly hard. And that, I was lucky, you know, I was lucky I had some great board members then, um, you know, who, who had done, gone through that and seen that before and, and were big helps to, uh, coaching me, uh, through that process. And, and, you know, we ended up, we did, we did end up transitioning the executive team to the, to the next group and ended up working out great. And we continue to accelerate growth. And now, you know, today at Amplitude, we're, we're on our, our third executive team. You know, if I look at the history of the company, so we've, we managed that to, to, to, to change that group multiple times now.

I feel like running a company is almost constant firefighting. It's always, there's some problem that you wake up and get punched in the face with, but there are some brief moments that are truly euphoric. I imagine going public was one of those.

How did that feel? Um, it was good. Going public was good. Honestly, my biggest regret, um, on the going public process, you know, that was definitely a moment. I think to your point, Akash, it's important to celebrate the milestones because, you know, to take a little bit of time because otherwise you just don't. Um, and, and, um, you know, enjoying the journey is, is an important part of it. Um, and so one of the points I actually made early on was to not celebrate any fundraising milestones and to only focus on revenue milestones and customer milestones and product milestones. Um, and so, you know, I remember when we celebrated when we first got our, our few to, I think it was like $300K. Um, and then we, we took the team out when we got to $1 million, uh, on a trip to Hawaii. And then when we got to $10 million, I ran a marathon for the team., and then yeah, we took the team out to Vegas at one point, which ended up, ended up working out, uh, it could have gone really bad, but ended up working out just fine. Um, you know, and then at $100 million in ARR, we launched 100 rockets to, to commemorate that milestone. So I, I've always been very, very focused on, on, you know, the, the, okay, hey, are we delivering value milestones and less so on the, the public listing milestones. In retrospect, I think I let the team and myself frankly get a little bit too caught up in the listing process and not enough about the actions that would make us a successful public company. Like I look at the team that is here today and every single executive since the public listing has changed. Every person who's a VP or more senior at Amplitude has changed. So it's like, you know, 25 people. Leaders who were there then that are no longer here now that are totally different. And I think in retrospect, I'd wish I had made some of those changes earlier. But, you know, and thought about that more. But, you know, it was, it was good to, it was good to appreciate the moment at the time.

Certainly. Today's episode is brought to you by the AIPM certification on Maven run by Mictad Jaffer, who is a product leader at OpenAI. This is not your typical course. It's 8 weeks of live cohort-based learning with a leader at one of the top companies in tech. OpenAI just doesn't stop shipping, and this is your chance to learn how. Run along with product faculty and Mo Ali, the course has a 4.9 rating with 133 reviews. Former students come from companies like OpenAI, Shopify, Stripe, Google, and Meta. The best part? Your company can probably cover the cost. So if you want to get $500 off, use my code AAKASH25 and head to maven.com/product-faculty. That's maven.com/product-faculty.

I think a recurring lesson we heard in both phases then is once you're intellectually ready or you're recognizing that an executive needs to be replaced, you don't want to go slowly or second-guess yourself.

Absolutely not. Absolutely not. Yeah, you're always too slow. I've, I've rarely ever I, I don't think I've— sorry, that's not true. Almost 99% of the founder CEOs I meet go too slow on it. Um, I remember there's one exception who I won't name, but who I asked, I, I remember asking one of my Series B, my Series B VC, uh, board member, uh, Neeraj Agarwal from Battery. I'm like, hey, do you know if anyone— everyone's always saying they go too slow on this. I'm trying to figure out, is there a way for me to go faster? How can I go faster on these changes? Do you know anyone like that? And he actually did name a CEO., and I met him a few years later and I was like, hey, I'm, I've heard this by reputation. How is it possible you, you are able to change out the team faster than everyone thinks is possible? Because like almost all of us go too slow. And, and he, he had the craziest response to me. He said, he's like, hey, I'm insecure on how incompetent I am. And so I always feel like I have to do things early., and I'm like, what? That's why, that's why, that's why you're ahead of the curve. Jeez, man. Okay. And I'm like, all right, well that's, that's one way to, to, to mentally do that. Maybe I should try that out myself. So, um, but yeah, it's, it's, you're, you're always too slow on, on making changes in the team.

That, that's almost universal. And from the early days of really knowing your customer, knowing every in and out of the product to now leading a $1.5 billion company, your role completely changes. Especially as it relates to building the product.

What have been the big approach changes for you about how you're involved in that product building process? Um, that's actually a very hard one. Um, yeah, that's a hard one to describe cuz like in the early days, let me, let me talk the philosophy and I'll give an example of it. You go from being in all the details yourself to you can no longer do that. But if you completely abdicate control, that's almost even worse because then your company completely goes off the rails. So the change in your role goes from leading from the front to really thinking about do you have the right people working on the right problem?, and then reviewing and inspecting and giving feedback on their work regularly. Um, and that's a hard change to go through because you kind of go from this inspiring founder CEO who's leading from the front and is not going to ask anyone to do something that they can't to frankly almost an asshole. Like, like you, you, your job is now to just judge other people's work and tell them in all the ways is not good enough all the time. And, and it feels bad in a lot of ways, but that is a key function at the company, which is how do you set the bar high, uh, for what needs, uh, yeah, for what you need to, to drive success. And so, um, that's a, yeah, that's a hard change. I was lucky actually, you know, people are talking about this founder, founder mode, um, and all the changes that come with that. Um, and I was lucky enough to work with the right chief product officer who actually recognized part of this change and then helped me set up product reviews and operational reviews with the team so I could give my input on them regularly in a way that they were ready for and expecting. And so setting up the right forums for you to do that is, is, is really— yeah, that, that, that's the part, you know, where it's like you can get leverage from, from great executives to help you be involved in the details and give, You know, very, you know, yeah, very, very aggressive feedback and, and, and, um, yeah, really push people to, to be great.

Yeah. And only the CEO can really keep raising the bar, keep pushing for more quality, keep pushing for more speed. If the CEO isn't doing it, the whole company is not gonna be able to function. Yes, absolutely. Absolutely. So I wanna see what you guys have built now. I, the last time I used Amplitude actually was when I was at Apollo. So it's been almost a whole year since I have used Amplitude.

Can we take a look at the latest and greatest? Absolutely. Let's get into it. There is, yeah, any, any founder who is not capable of demoing their product is not a real founder in my book.

So I, I love it.

I love that we're doing this. Awesome. Where should we start? So this is Amplimoney, which is our demo product. Um, and if you sign up for the demo, this is one of the demos that, that you have access to and can see. Um, and so I just want to start you off by orienting you on this dashboard. One of the things we were very deliberate on at Amplitude is getting you to your data and your metrics right away. So we didn't want you to have to do a lot of work to properly get set up on Amplitude. Um, and so right out of the box, you don't have to do any configuration, nothing. You're just dropped into a whole set of dashboards. And if you're from Google Analytics or another product, this will look very familiar. So, um, here we have your overall active users, and, you know, you can see, okay, we got 5,000 users per day, and on weekends it dips down to 2,800, and then it comes back up to 5,000. You can see how many users of those are new. So of those 12,000 active users the last week, 2,000 are new. We can look at how long their sessions are, 13 minutes, their retention after 1 day, how many I have active in a week, how many are onboarding, how many are using our guides and surveys product, and then, you know, all sorts of other stuff. So we have a speedometer. This just shows you in real time. This was very controversial for me to get out, by the way. A lot. One, some of our product leaders hated this because they're like, oh, this isn't useful at all. What action can you— are you— and I'm like, are you kidding me? If I'm on my website, I want to know how many people are using my product right now. And I get really excited to, to see this thing move around.

This thing is addictive, you know. It's like, it's like a drug. That was the original aha moment, I feel like, for people with Google Analytics, that free tool.

So I love how you guys have brought that in here. Absolutely. Google Analytics, not real time. And another big issue with it. So if you're doing a launch, you know, you won't be able to see that. Then you can see, okay, what pages are people hitting? So most people are hitting the homepage and then, you know, looking at their accounts, a bunch of people are signing up. You can see what geos people are coming from. You can see on a map, you can see the sessions people are going, which I'll go into a little bit. You can see a funnel. You can see this little, this wonderful retention triangle. And how many users are coming back day to day. And you can see people using guides on the product too. Let me take you through a little bit more of some of the other— so, so anyway, so, so this is any analytics product worth their salt these days will have a bunch of these basics, you know, so we're not particularly special here. The thing we focus on is, okay, let's show it to you right out of the box.

So that you're not scrambling around trying to figure out how to set up a dashboard. And let's talk about that for just a quick second. So let's say I'm implementing Amplitude for the first time. I drop in, I guess, your SDK into my app and it's gonna automatically collect all this event data.

And you guys are basically storing all the event data that's coming outta the site on your end. Yes, that's, that's exactly right. So Um, let me actually show you. Um, so our SDKs, um, all you have to do is say, okay, I want, you know, let's say we're implementing it on our browser on JavaScript. All you have to do is go to our browser SDK, drop in this line, uh, drop in this snippet of code, and then we'll be tracking a whole bunch automatically for you. Now, if you want to do more than that, you can, you know, configure the SDK and, and, uh, yeah, like mess with auto-capture marketing attribution and then start putting in events and stuff like that. Um, but, um, Yeah, you just put in this snippet of code and then you're off and running with the dashboard that you see here. So you guys have both the auto-capture method and the more defined event tracking method. You handle both? That's right, yeah. So, so we have what's called auto-capture and we'll put it in by default, but if you want to turn it off, you can just set this flag to false. What AutoCapture does, it'll capture all page views, clicks, mouse movements, actions, all of that stuff so that you don't— you get a whole bunch out of the box. Um, uh, so all of these events here, right? Page view, start session, end session, form, a whole bunch of form stuff, file download, element clicked or changed. So you get all of this with AutoCapture out of the box. The other thing that you can do is, um, is, uh, put in sending events, right?

So you can put in, uh, an event type here. Hold on, let's see. Because I think sending events is where a lot of people get tripped up when switching analytics providers, is they have to think about, okay, are we gonna send these all so that they match our other data that we'd been showing in the past?

So I think that tends to be an area a lot of people get tripped up. Totally, totally, totally. So here we go, track an event, right? So you amplitude.track, you put an event name in, you can put in a dictionary of properties if you want, and then you can do, there's a whole bunch of other stuff you can do like groups and group properties. And this with typical analytics implementations, we recommend using these, we call these precision events. So in addition to the auto-captured events, you can precisely track a whole bunch of events on top. And so as you get more sophisticated in your analytics tracking, you will typically want to track a whole bunch of these so that you can be much more precise with how stuff is captured for sure. And I'm sure as you guys moved into the enterprise, you probably saw that that's all enterprise cares about. Yeah, yeah, absolutely. You know, enterprises are going to use this way. Actually, there's a good number, believe it or not, using the, the, that, that code snippet is the actually most common way to get started with Amplitude, even if you're an enterprise. So it's been helpful for them too. But invariably, as your analytics needs get more complex, you'll, you'll want to use these precision events. Okay, so if we return to Amplitude, what can you do next? Okay, let me show you a few other dashboards really quickly. So you can take a look and start dive into our bread and butter product, Product Analytics, which gives you a lot more detail beyond the, the home dashboard that I showed you there. So here you can start to group, you can look at different metrics and then you can segment them by different parameters so you can see, okay, what are the most popular device families that use us? Okay, so look, I'm looking, you know, iPhone, iPad, Samsung Galaxy Nexus, HTC, and so on., and I can see their breakdown across all of these different metrics. You know, how they, how many active users are coming in, new users from these platforms, session duration, how many are in the onboarding process, and so on. I'm going to show you one other part of the platform I really like, which is this feature engagement matrix. And what this does is it shows you So what's called adoption of different features versus frequency. So adoption, this, this on this axis, you see how, what percent of your users are doing a particular event. So, you know, a third of your, 35% of your users are using this, are doing this alert event, whereas only 6% of users are using this provide ID event. And then relative to that, how frequently they're used.. And so, you know, here only a user who does this event will only do it 1.1 times on average, whereas a login event, users are going to do 3.3 times. And so this gives you a good feel for the distribution of usage of different features in your application. We also have a retention chart here so you can see how users are dropping out over time. So at the start we have 100% of users on day 0, then you go to day 1 and you get 78% of your users come back, and day 2, 73%, and so on. As you continue to go down this curve so that you can see how you're stacking up in terms of keeping your users engaged. In general, for this, getting— you want to get, you know, over 10% of your users to retain by day 7. That's kind of a good— that's kind of like the minimum baseline for any product. You know, if you're getting less than that, that means you have major issues where, you know, more than 90% of people are not seeing enough value to stick around a week. And so, you know, it's really good information too. Last thing I want to show you, and there's a ton to show here, but is this traffic by campaign. And so here you can see this is a web analytics view for those of you who are used to Google Analytics, and you can see similar things, how many visitors you have, what's your bounce rate, what your session totals are, your average session duration, how many sessions per user. And then you can see this split out by different campaigns to see where people are coming from. Okay, so this $500 bonus for new customers, we have a lot of visitors coming from this, which is great. You know, we have less visitors coming from this Ampley Money for Business Banking campaign. So you can see anyway, so, so, you know, a lot of different dashboards all right out of the box. You know, you get all this stuff with just a single line of code.

You can get more advanced from here, but these are kind of like the basic metrics you're going to get. So people might be asking at this point, I'm paying for a BI solution, whether it's Looker or Tableau or just making people write SQL code.

And I have, you know, my free Google Analytics implementation. Why are they paying for Amplitude? So Amplitude is all about being the system of record for user behavior. We want to be the deepest, most powerful product that helps you understand what exactly is going on and why people are doing what they're doing. So if you want to understand why people are churning, why people— what— why people are coming back, what your best users have in common, what features they're using and why, that is what we are best in class at. It's all about understanding the user journey. We're not unlike BI. We're not for all data ever, right? So your BI system can hook up to your CRM, can hook up to your financial data, can hook up to other data sources. And be this layer that kind of sits on top of a lot of different pieces of data, but it's very shallow. So when you start to ask these questions about why are people dropping off, like I'm going to show you one a little bit, you're not going to be able to get that out of BI. Whereas for Amplitude, we can go as deep there as you want and without having to write SQL, without having to be a technical user. One of the things we found is that if you're a product manager, if you're a marketer, if you're on the growth team, you really want to understand what is going on and why it is, why certain things are happening in the user journey. And that's where BI is going to fall short and you're not going to be able to even construct— it's going to be hard to construct even basic things. Like one of my challenges that I give people who have this question is try constructing a basic funnel in BI and you'll really quickly see where that breaks down because it's very hard to think about, okay, well, How do you get sequential steps to happen one after another? How do you join on the list of users that hit one step and then the second step and then the third step?

And that's— those are not questions BI is set up to do well. Yeah, I worked at Epic Games where we only had a BI tool and I kid you not, probably 25% of my time as a PM was just playing around with those funnels, trying to match device IDs, totally different things. And then when I came to Apollo, I was like, my mind was blown because it's all in a very sort of easy to use, non-technical way here. And more often than not, you're not going to make like a big mistake in calculation where even two really competent PMs who had spent years at Epic Games, I remember we came to different numbers on what our drop-off was when we were having the battle against Apple on how many people are sideloading.

And so this, I think it, what it's doing is it's giving you a source of truth for all of those product people. That's right, that's right. For everything a user does in your digital product, that's where Amplitude shines. Um, and so, you know, to your point, yeah, like what I, I call— I talk about the funnel challenge where it's like just try to even construct a basic funnel as a non-technical user and you'll be bashing your head with BI or other tools in this space. All right, great. All right, let me show you, um, uh, some of the great AI capabilities that we have. So I'm going to go here. And one of the things that we've invested a lot in over the last year is building out a chatbot to help you construct charts. So we want to be able to help you do very— to help you do basic and advanced things without having to understand how to use any of the controls you just saw. And so what we've created is Ask Amplitude. So the way you get it, so if you know, if you're, you know, at home, you can just go to Ask Amplitude here. And then what you'll do is you can type in a question or use any of the suggested questions and then it'll post the results here. So I'm going to ask you right now to show the onboarding flow and we'll see what it comes up with. Okay, cool. And then I can get this chart that shows the conversion rates from start sign up to complete sign up to start money transfer to complete transfer. And we can see the full funnel here. Wow, that's so funny. It says if you put show the onboarding flow versus show me the onboarding flow, it gives a totally different result.

So I should have stuck with the script on it versus trying to go fast. Stochasticity is probably the hardest part about building with AI. And for you guys as an analytics company, you can't have any hallucinations.

How did you guys manage that part of it? Totally, totally. We're still working on it. This is where we, what we'll do is we'll look at the prompts people submit and see, okay, well how can we unders— we kind of understand what they're trying to get at and how can we tweak the AI to better go, you know, to better get to the end result of what it is they're looking for.

So yeah, even this like show me the onboarding flow versus show the onboarding flow was different.

And did you guys build on top of OpenAI?

Do you have many models? Okay. Yeah, we use OpenAI for the backend here. Very cool. And what is the, what are the limitations of this Ask Amplitude? Like, let's say I wanted to, one of the things I remember I used to have to do is as a product leader especially is like make it really clear what version of a metric or what metric definition people should be looking at. Like one of the ones we had was active users. We had a funky definition on active users, so I needed to draw people's attention to certain one.

Can I do that here with AppDynamics, or is it mainly for chart generation? This is more for chart generation, but we will pull up charts or metrics that are relevant, uh, to your use case. So if you've— if someone has defined DAUs elsewhere, it can show up here as well. So like, you know, even on this one, like, we're, we're pulling a bunch of charts, where we've seen you create in the past, that are related to the onboarding process.

I don't know why they're not loading here, but anyway. And right now we're obviously seeing like a, a cloud, I guess, demo.

Does Amplitude also work like for people who want to be warehouse native? It does, yes. So we do work with warehouse, warehouses as well. Uh, so there's a number of ways we can get data to and from Amplitude. Warehouse native is one. The one we normally recommend going with, though, is what we call warehouse connected, where we'll continually pull in data from the warehouse and keep it in sync with Amplitude.

Got it. So I'm sure you look at a lot of charts day in, day out of where people are spending the most time in Amplitude.

Where are those areas? So the two places that people spend a lot of time on, so one is the segmentation dashboard here. Oops, hold on.

So like just in the chart constructor, building a segmentation chart.

That's where I have like all these users that are fine. And then the other most common part, believe it or not, is what we call user timelines. So seeing individual users and seeing the history of what they've done. So this allows you to get much deeper into, into the full history of an individual user.

And so this is the second most commonly used feature within Amplitude.

Can we show people live the segmentation chart, maybe answer a question using it? Oh, for sure. So let's see, let's go to create segmentation chart. So normally you're not actually creating this from scratch, normally you're building off of someone else's definition. Um, so what you do here is you start off with a metric. So let's say we want to look at, um, key funnel, so completing a money transfer and want to understand how that breaks down. So here we just have— okay, we're looking at this over time. So we see the number of unique users that are completing a money transfer every single day. So you can see, okay, on, on most days we'll have somewhere around 900 active users doing this and it dips on weekends. You can mess with it a whole bunch. You can, you know, change the, the form of the chart. You can change this. You can look at it hourly and see how many people you have hourly doing this. You can look at it on a monthly basis, you know, so we'll— I'll go ahead and set that back to the daily. You can change the time range to look at a 90-day window and so on. What you'll then want to do is you'll then want to cohort this chart by different segments. So let me give you an example of this. So this segment tool here, so you can say, okay, look, I want to see how the distribution of money transfers look like per country. And so I'm segmenting by country. Okay, so the vast majority of our users in the United States, and then we have some from UK, Germany, Japan, Egypt, Brazil, and, and so on. And this is really helpful. And then you can say, okay, well, let me look at this by what device people are using. Okay, so a lot of iPhone, iPad, Galaxy phones. I want to look at this by language so I understand how to localize my project. Okay, mostly English, but a little bit of English, Japanese, Arabic, and German.

So you can segment it and look at, you know, start to slice and dice the different data points that you have here. Yeah, I remember like one of the things we used to look at was like core versus non-core users, and there's so many different segmentations that are useful. So this is the core analytics product. This is what you guys made your name for, but you guys are actually expanding into a broader platform to help people know their users.

So what else can I do inside Amplify? I am so glad you asked, uh, because we have made the focus on how can you take the valuable data you have in Amplitude and apply that to different use cases outside of it. Um, so let me show you this example on heatmaps. Okay, so we have analytics and that's kind of the top things here. I'm going to show you a flow that goes through our session replay product, our experimentation product, and our guides product. So we have this thing called heatmaps where you can see where on a page people are clicking. So let's take a look. So this is a different product. This is AmpliStore instead of AmpliMoney. But you can see where on this page people are clicking around. So, and engaging with these products. All of this data is hidden because it's personal information and we don't want to reshare any of that. But the point is you can see, okay, hey, a lot of people are clicking up here on this banner at the top. Let's investigate these users more. So you can do a bunch of different things. You can view the events that are instrumented and attributed to clicks in this area. Okay, so we see a lot of element clicked events. You can use session replays or you can create a cohort. So one of the things I'm going to actually do— all right, so we'll create a cohort. So we'll call it Akash Podcast Special. Cohort, save that cohort, and then we're going to use that in a web experiment here. So we'll call it Kosh Podcast Special. Do that URL. You can set some parameters on it, but we're just going to do a simple A/B test right now and we're going to create a cohort. And now what we get is our product, except we have a visual experiment editor. So I just need to click this element changes. Where we can select different elements on the screen. And so this is really cool. So we have an editor that allows you to edit different parts of your website so you can mouse over different divs and have it highlight them and you can edit them and mess with them to create variations. So remember, I wanted to create a cohort and change this up here. So let's go ahead and do this. So first, let's call this Akash podcast special. We'll give you guys 20% off. How about that? How's that sound? Enticing. We can make the text a little bigger. So let's make that 30% and let's change this banner color to a different color. So let's change that to blue. So it kind of matches. Or what color do you like, Akash? Yeah, Amplitude blue, of course. Amplitude blue behind here. A little too, there we go. Something, let's see, how do we, there we go, there we go. That's an amplitude blue. We'll apply that. Okay, and then we can apply this. So, so this is my treatment and then if I control is here, so this is what most users see and then the users I expose to this experiment will see this, this different banner. So I'll apply that, it'll create those two variants and then it walks me through a wizard where I could set the goals, I can set pages, I can set targeting. And then I can save it and the experiment will run. A Kosh podcast special. Now that's actually going to take some time to run. We don't have time to— we don't have a few weeks to wait for to see how that does. But I can show you the results with another experiment that we just ran. So this is the winter sale banner and we can see how it did and we can see, okay, how did our control do in terms of in terms of engagement versus how did our treatment— and we see our treatment group had a significantly higher engagement than our control group, you know, almost 50% higher in this case or 30% higher in this case. And we can see, okay, we got more checkouts. So great. I can go ahead and say, all right, great. I want to roll out that treatment to everyone who's using the product. So anyway, very, very simple demo of Amplitude experimentation.

About how you can go from the data in Amplitude to creating a website variation, taking action, and then creating a better version of your product as a result. And what's kind of mind-blowing about this is there was no designer engineer involved, right? And so this is really empowering PMs who want to get closer to the bare metal builders, founders to be able to do these things without needing, okay, I'm filing a design ticket. The designer has created it. I'm filing an engineering ticket. The engineering has deployed it.

I'm filing a product analytics ticket to finally look at the results. It's end-to-end for you. Totally. Exactly right. Exactly right. So it's, it's a, it's a huge difference where we think of Amplitude as getting the power of data into the hands of non-technical users. And so whether you're doing analytics, whether you're writing an experiment, whether you're creating a guide or survey, whether you're viewing sessions, you can all do that directly here without needing to know SQL, without needing to know your data, because that data is powerful. So it's, it's a, it's a huge, huge deal. Let me show you one other quick demo for a second. I actually— so remember this drop-off chart that I had before here? You can actually see a bunch of people that are dropping out and you can view their sessions. So you can actually see what they're doing in the product and what they saw in terms of going through this funnel. So for example, we can load this replay. So we have a list of sessions here and we can load this replay from this one user who spent a while in the application and didn't get through the funnel. So we can start to gain some empathy for what they encountered and where they got stuck. And then if you don't even want to have time to watch this whole session replay, you can like, we'll create an AI-based summary of it over here that summarizes, uh, what happened and particular actions that you can do, um, that it recommends.

Awesome.

So we saw two implementations of AI so far, Ask Amplitude and session replay summaries. Where else are you guys building AI into the product? We're gonna be building an AI agent that comes out later this year that will automatically do a whole bunch of the actions you saw here for you. So it'll surface interesting insights for you. It can create variations of websites and experiments and create guides and surveys. It can create cohorts. It can pick out sessions to look at. It could do a whole bunch for you just by interfacing in the chat, in a chat interface. And so we're really excited about this. I think the impact of AI on the data space is very, very early. Extremely early. And so I don't think I've seen anyone drive anything significant in the AI data space yet. And so I'm really, really excited to, to get this out there. So we'll be coming out with something in the second half of this year. And so stay tuned for that for an Amplitude agent. And how does Amplitude use Amplitude itself? Oh, I love the question. Okay, so we are our best users of the product, as you would imagine. So I'm going to show you something I've never shown publicly before, which is some of our real-life data and dashboards. So this is— I'm going to show you one of Amplitude's dashboards. This is for Amplitude Easy. So one of the big launches we did last year was what we called Amplitude Easy, where we wanted to make it easier to sign up and get started. So we had created that, that snippet that you saw. We created a single line of code signup. We created auto-capture out of the box. We bundled session replay and a whole bunch else with it. And so these are the dashboards for Amplitude Easy. And one of the hardest parts on Amplitude Easy is we wanted to distinguish fluctuations in the number of people that were just hitting the site because of a launch or a marketing campaign from an underlying change in the conversion rate. And this is actually something that's really hard to do. It took us some time to tease out, but I want to walk you through how we did this and what we did here. So let me show you these dashboards first. So first, this is just, okay, how many daily visitors, visitors on amplitude.com excluding China, Indonesia, Brazil. One of the things we found is we'll get spikes occasionally from China, Indonesia, or Brazil, which are not representative of our normal traffic. So we exclude those. And then this is kind of without that excluded. So you can see, okay, so you know, in January we didn't, in, the holidays, we didn't have many people using us. It was a lot less. And then, you know, it spiked up and grown since the start of the year. Um, so we were looking at a few things. So the first is this metric that we call organization signups. So do you get through the signup process and create an organization once hitting the Amplitude homepage? And so we're tracking how many people do that each week. And then we track both, uh, what we call all signups, which could be anyone signing up, and then we distinguish those from business signups. So did you sign up with an ID, that was linked to, or an email that was linked to a business, because this is, this is kind of a better indication of interest. Like you might get a bunch of junk or random spam signups. Then the next, uh, thing we cared about is data setup. So of the people who sign up, how many of them get to the data setup process? So how many of them actually start sending us data? Because it's, it's a lot of work to send us data. You've got to put that line of code in, you got to be serious. You can't just put in your email and expect to get using the product. You actually have to send this data. And so this is a really important line that we looked at. And so, you know, here the same, same with this chart. We're looking at all organizations and business organizations because, you know, we business organizations more representative. Now, the interesting thing about this part, I'm going to dive into this for you, the data setup, because this is really where we wanted to drive a big change. So I'm going to actually extend this, to 340 days. So I'm going to show you the last year of how many people were sending data to Amplitude. By the way, I haven't actually previewed this chart before, so I'm doing this live, but it's all good. Uh, so one of the things that we saw is that, um, the baseline for number of organizations sending data, it's like just if you kind of look at this period from May to September of 2024, the baseline is just around 50 or just under 50. You see where that purple line, it kind of peaks around the 50 mark here on the axis, right? Yep. So we launched Amplitude Easy in, in September, in mid-September. And you can actually see the launch. You know, we go from a baseline of 100 or sorry, of 50 organizations sending us data a day. All the way close to the 75 mark. So a big, big spike. Now, the problem at the time was we weren't sure if this spike was due to the changes we made in making the signup and data ingestion process easier, or if it was due to the marketing launch, right? Because like maybe it was the marketing launch that caused it and that got a lot more users and they converted, but then it'll go away. Or if it was actually the product changes we made, we've made. Um, so the interesting thing in looking at it is like if you kind of go for past these first 2 weeks of launch, um, so we actually do look like we're at a new high. It's not this 75 mark, but we're starting to get past 50 to more like, uh, you know, high 50s, 60s in terms of total signups. And then if you look at that over time all the way to today, you know, that is sustained. So before we were doing a total number of signups per week or per day that was in the 40 to 50 range. Now we're doing a total number of signups per day that is more in the 50+ range. So from this we can say, okay, hey, this easier signup process and the easier data ingestion process has actually resulted in more people sending us Amplitude data. We're now at a new baseline high. We're happy. Things are good. We made a successful product change. So anyway, Very, very simple example.

Never shown this data to anyone externally before, but example of how we use Amplitude on Amplitude. Amazing. So I think that covers it for our in-depth demo of Amplitude. I don't think you'll get that anywhere else from the CEO himself. I want to talk a little bit about building product because this is a podcast for builders. And we talked a little bit about how you've gone from being the chief person who knows all the details about the product to somebody who's now plugging in through product reviews. Can you go through us in more detail over, let's say, like a week?

How are you plugging into the product development process? So what I do today is the main forum which I interface with the product development team is I have a 2-hour product review with me every week where the agenda is set in advance and it could be anything I'm interested in talking about that week. So actually, I'll pull it up for this week and I'll tell you live what it is. So I'm going to go through our Amplitude for Marketers demo flow this week. I'm going to go through a bunch of user timeline updates. I'm going to go through a review of what we call the metrics object, and I'm going to go through are an auto capture review. So I kind of chosen a bunch of different areas that I want to get a deeper review and inspection of the product. And then I collaborate with the team working on it to talk about what's important to work on. How can we improve the experience? Is this where we want it to be and how can we make— how can we build something great? And that agenda changes from week to week. So, you know, this week I'm obviously review metrics, object auto capture, user timeline, but. For example, last week I reviewed our guides and surveys launch, which we, you know, I wanted to see the impact and adoption of how our guides and surveys did. And I reviewed our session replay everywhere launch and I wanted to see how making session replay much more accessible was, was working in terms of people adopting that product. And so the agenda will change from week to week, but the idea is it's a 2-hour review that I will spend going deep with the team on a particular area of the product. And there I like to start from first principles where I say, okay, what problem are we solving?

How are we thinking of solving it? Um, and then how fast can we solve it? Um, kind of in that order. So you're really trying to get back into that bare metal with those 2-hour product reviews. What are the broader ways you're organizing the team towards goals? Do you guys do a yearly planning with strategy that you'll give to them?

And then do you turn that into OKRs, or how do you kind of operationalize this so that they can pursue the vision you're setting? Yearly planning is way too far out. Like, it's, um, so the— what we focus on is the quarter ahead of us. That's where 70% of our calories are. Um, and so what we'll do every quarter is we'll set, okay, what are the things we want to achieve next quarter on two fronts? In terms of one, getting, uh, shipping the roadmap, so getting products out there, and two, in terms of metrics. So Who do we expect to adopt the product? How much adoption do we expect to drive? You know, what sort of impact on the end user, whether in terms of customers, dollars, engagement, do we expect to see? And both are important. One of the biggest mistakes I see teams doing is focusing on one to the exclusion of the other. So historically, product teams have been very focused on just ship the roadmap, and then I don't really think about what happens after that. But actually, just because you've shipped it doesn't mean anything. What you care about is are people adopting those parts of your product? And so one of the things I always push on the teams with is like, okay, what customers asked for this and what sort of adoption do we expect to see from those customers once this is out there? So we'll set goals for every part of, for every part of the surface area of Amplitude on those every quarter. Now, in addition to that, we have some annual themes, but they're more themes than they are specific goals like, oh, we need to be here by this point in the year. And so, for example, one of the themes this year, as you might expect, is AI. And so we're going big on AI and AI agents, but it's not like we have some goal by, you know, end of this year we need to get to X.

It's more that, okay, here's a bunch of places we're going to make significant investments in this. And where I'm focused on is what are we doing this quarter on that? So you only use the annual vision really for your headcount planning, your overall organization allocations of what types of things people are building.

But when it comes to actual features and roadmaps, you're sticking to more of a quarterly cycle. Yeah. So we're— so I want to be clear, it's not— we'll have plans for future quarters. So like right now we are just coming up in Q2 and we're focused on that. We have a little bit of a plan for Q3 and Q4, So we have some things that we've committed to deliver, but most of it we haven't decided. We're going to learn a lot through the course of what we go through in Q2, such that when we come to the end of it, we'll have a much better idea of what the highest ROI thing to work on for Q3 and Q4 are.

So I was reading some analyst reports about Amplitude, and if there was one metric I felt like they all were talking about, it was NRR, net revenue retention. And it's funny because that's kind of the journey of Amplitude we just went over in this podcast. You started with a question around that core usage retention. Now you're working on this big problem of revenue retention.

How are you breaking that problem down as a product team and really moving that metric? So one of the things we've been very focused on, like I just saw you in Amplitude, I showed you in Amplitude Easy, is how do we make it much faster for someone to get to value with an Amplitude. And so that way they're more likely to get there early in their buying process. And so they're less likely to overbuy and they're more likely to see that value and then, you know, grow into even more value when they come up for renewal in a year or 2 years or 3 years. And so that's been a huge, huge focus of mine on the product development side, which is how do you get someone to value faster? The other big thing we've been doing is building out the rest of the platform. Platform. So adding in Experiment, Session Replay, Guides and Surveys, Activation, and all of those, um, all of those— one of the things we've seen is that the more products you use, the stickier you are and the higher your NRR. It's about a 15-point increase if you're on multiple products than if you're on a single product in terms of NRR differential. And so that's a huge, huge deal because you go from just observing the data and getting insight on it to now you're taking action operationally in your product in different ways. So that's the other big focus. So part of it is win simple, which we've talked about.

The other is win the platform. So getting, getting people on the broader platform.

Okay.

And I think that that high-level framework will make a lot of sense and people will see that path for all SaaS companies.

It seems like where they're building out these platform plays, is that kind of the inevitable future for most SaaS companies? I think, I mean, almost every SaaS company who's been successful between $100 million and $1 billion in ARR will go from a single point solution to a category of products that are sold to the same buyer. I mean, you see this historically with Salesforce, with Workday, with ServiceNow, with Okta, with Datadog. You kind of name the company and they've— every application SaaS company has gone through, through that journey.

And so absolutely, I think it tends to converge on a dominant player selling to a buyer a suite of products that they need to be successful.

And you didn't mention amplitude marketing. So I think that that might be the same use case for a slightly different buyer also.

Yes.

So one of the things we've actually— this is subtle, but one of the things we've actually seen is the convergence between the marketing persona and the product persona. So marketing teams are looking more like product teams and product teams are looking more like marketing teams. So what do I mean by that? Well, if you're a product team, you're no longer just thinking about the in-app experience. You're thinking about how does someone find that experience? What are the different channels they hear about that? Do users coming through this channel retain better than this channel? And then vice versa. If you're a marketer, you're thinking more about product-like things. So you're thinking about, okay, not how do I just get new users, but how do I reengage my existing users? How do I get users that have a high LTV? How do I think about the users getting the right set of users so they'll have a great experience in my product? And so it's, it's like you kind of get, you're starting to get the merger of these two. It's early days, but you're starting to get the merger of these two functions because from a customer standpoint, we don't really think of ourselves, oh, am I in the marketing experience or am I in the, the product experience? It's all one digital experience. Like If I'm using Slack or Zoom, like, I'm not thinking, oh, I got an email from Slack that's from the marketing team. No, now I'm in the app. That's the product team. I just— it's just one customer experience. And so because the customers don't perceive the experience to be differently, the organizations are beginning to reflect that change to where these two functions are converging. And so because of that, we've had a lot of demand for marketing capabilities. So all of this stuff I just showed you,, you know, engagement, campaign analytics, reporting, LTV, ad spend tracking, you know, different landing page optimization sort of stuff like the web experiment. Marketing has been the biggest user of that by far, where it could create a new special for different types of users on your podcast like I just showed you. And so getting the marketing experience. So, so anyway, I say all that because The reason we're doing that is because we're seeing those two functions converge where product buyers are asking us more for marketing capabilities. And a lot of marketing buyers who would like to use us because of our deep product capabilities also want them to. And that's happening because the functions are converging. So that's why we're doing it. It's not so much we're going after a totally separate persona that's not related.

It's more like, hey, these guys actually have the same need and are evolving to be the same sort of function within a company. I feel like we could do another hour and a half just on that topic, but our time is up.

Spenser, if people want to try out Amplitude, what's the best way for them to do that? Go to amplitude.com. You can get started really simply. 3 steps to sign up. Just put in your email and password and put in a snippet of code on your website and you're off and running. If you don't even want to do that, we have a bunch of demos on the website. And if other way you can follow me is Spenser Skates on Twitter or X.

Check them out there.

We'll link all those below. Spenser, thank you. Akash, thank you so much.

Wonderful being here.

I really hope you guys enjoyed that episode. It would mean a ton to me and the team if you could please subscribe on YouTube, follow on Apple and Spotify Podcasts, and leave a rating and review. Those ratings and reviews really help grow the show and help other people discover the show, and they help fund the production so that we can do bigger and better productions.

Can't wait to share the next episode with you. Until then, see you later.