Amplitude, Inc. (AMPL) Earnings Call Transcript & Summary

Okay. Hi, everyone. My name is Taylor McGinnis and I head up the software application SaaS equity research here at UBS. And in this session, we have Amplitude's CFO, Andrew. So Andrew, thank you so much for joining us.

Thank you for having us. It's great to be here.

Perfect. So maybe we'll just jump right into the current momentum that Amplitude's business has been seeing. So you guys have had an impressive reacceleration to 18% growth. So maybe you could just give a little bit of insight in terms of the current demand environment you're seeing? What Amplitude has been doing over the last several months, plus the last couple of years in order to position the company to really capture some of these emerging new opportunities?

So I'd start by saying there's a couple of big things we've been working on for a while. Before I even joined, the founders were looking at the analytics ecosystem and all these other companies that have been created in experimentation and session replay and guides and surveys and activation and said, look, some of these companies, they really are not whole platform. They are by themselves creating one part of an ecosystem, which really should be consolidated. And so there was an effort to start building those products around analytics and started with experimentation and then to activation, and then on to web analytics and guides and surveys and session replay and on and on. So we created this environment where we're setting ourselves up for a true strong platform play. Even though we weren't calling it at that point in time, we're saying like the value proposition for clients is one that all these things are together. Now about the same time, Thomas Hansen, our President and COO, who runs all go-to-market joined and said, if we're positioning this as a broader sale and really going after Google and Adobe, especially in marketing and analytics, we should be focusing more on an enterprise sales process. And he started then that advancement of our go-to-market team, moving from very much a transactional model over to an enterprise selling model. And they were going along. And I joined in August of 2024 and said, "Hey, guys, there's a lot of glue and gum on this process. Let's make sure this is actually going to result in the financial outcomes that we want". So we started building and instrumenting our processes better. We made a lot of changes that led up to the beginning of 2025. The products were coming along at that point in time, and it was a very strong value proposition for clients, looking to consolidate and reduce vendor spend and drive greater outcomes. So I would tell you, the growth that we're seeing now, this is a discussion I have with the board just yesterday. The growth we're seeing now are the results of some of that core instrumentation we put in place late in '24 that's now coming to pass. And I think we're not done. We're doing a number of really good things this next year in order to really drive sales productivity higher, reach out to customers more effectively, drive better process optimization and drive greater profitability in our business model. The -- you know this, but for others who don't, I've been talking about our planning is one that focuses on driving that growth on those 2 major areas growing more enterprise and driving our product portfolio, but also doing it in a way that we're doing -- showing leverage. We're showing increasing profitability associated with that operation.

Yes. I want to touch a little bit on the former and what you were talking about earlier, which is you've made a lot of changes in the business to get to the growth levels that you are today. You've seen significant improvement in NRR. I think last quarter, it was around 104%. So when you think of what's driven the improvement to date, how much of that was coming from the cycling through of a lot of these optimization efforts that we saw post COVID and the rightsizing? And I think that's an important question to ask because then it will tell you, okay, now with all these expansion opportunities and everything you've been doing on the execution side, like what's actually left to come, right? So can you impact how much of what we've seen so far is just cycling through those efforts? And actually, what does that mean for the durability and improvement of NRR and top line as we look ahead?

Yes, because some of those things are actually in different ways instrumented as well as and durability. So I'll come back to that one. But the first 2 things I would say is you really have to understand it that Amplitude is not a traditional seat-based SaaS platform. We have 2 major mechanisms through which we monetize. First is our classic, which is our product analytics level, we ingest event data into the platform. And event data is you can think of it simply as any type of user system that interacts with a product or a website or other technology that's creating some type of cursor click or an engagement or a movement from one place, that creates an event at a very detailed level. And that's one way in which you monetize. The more event information that is ingested within the platform, the more we will charge higher and higher ARR. And since we've done our pricing packaging changes, which is another question we have, I would tell you that's at a decreasing rate for customers now. The other mechanism is that we charge for additional modules. So you have both an upsell and a cross-sell mechanism in which we monetize, and the reality is, as you mentioned, there was a lot of overbuying during COVID. But I also say when you're going from a transactional sales model to one that's more value-oriented, there were a number of things that we created headwinds for ourselves, like some large contracts that had planned downsells upon renewal. Same capacity, but a reduction in the ARR. Like, why would you ever -- why would you ever put that in the ARR? Well, that's what happened. And we had to go remediate some of those things. So if you look at over the last year, what's been driving our growth, we have had an increase in the level of data that's being ingested in the platform. But we also had these headwinds around overselling capacity that wasn't being used, and we had to rightsize. And in some cases, you lose contracts as well. That side of the upsell business has been relatively meager. So the growth has really been driven by the success of our new product offerings that we've been surrounding analytics. Now here's the good news. As we progressed and gotten better in remediating some of these issues, we start to open up the ability to grow via upsell as well. So now we have both vectors that are starting to work for us. And that's what's the exciting part because the innovation pace for us has increased. And as we add our new Agentic capabilities that are coming in the January time frame, we're going to have a really great product launch around our new agent capabilities, our MCP server and a number of others that we haven't talked about much yet, but it's very exciting. We got a demo of it at the Board the other day. We think that, that not only opens up more data ingestion in the platform with more use cases especially when you think about like MCP, which opens you up to whole different data sources that you can have behavioral heuristics exposed to, as well as it really emphasizes the power of our platform and one large set of applications working well together.

Yes. And I believe Amplitude has ambitions ultimately to get to 20%-plus revenue growth. We're not that far off from those levels today and getting NRR back to 110 plus. So in terms of what's still needed to get there? Do you feel like all of the changes that you've made over the last year plus position you well to go capture that, and it's really just a function of execution? Or are there still things that you guys need to do on the product side, go-to-market side, whether it be pricing and packaging that you mentioned earlier, like anything left to really go capture 20% plus more durable growth?

So when I joined Amplitude, I told Thomas, look, there are 15 things that I've identified we need to go change in go-to-market. And I would say we're about half of those things we've done at some level. And it takes time to drive change management. It's the hardest thing. It's not about the what. It's about how you do it and how fast you can do it without creating major disruptions. I mean, the first major thing we did was we moved the quota basis for our salespeople from gross ARR to net. And I always tell people show me a company that has a gross ARR quota basis, and I'll show you a company that has a churn problem. So we changed that. We started teaching our sales reps how to sell the platform. We implemented a lot of changes in the comp plan design and emphasized more durability in our business and sales process, leaning in more towards multiyear contracts. And I can tell you when a customer looks at our platform and they see they want to replace multiple applications, they can't do it all at once. And many times, they have contracts that are with other providers that are staged over a period of time. And so they are looking for a multiyear engagement as well and cost predictability. Well, you do all those things in the instrumentation side and you fit it well within the sales process and the value proposition to clients, and you see some of the progress we've made in our RPO. It's growing quite rapidly, especially in the long term. And why is that good? Because we're not having to overly discount for that. It gives us great revenue visibility. And here's the secret that a lot of people don't understand about the instrumentation part of this, which is when I first joined, we had to renew 89% of our installed base, okay? In the back half of '24, we started doing more multiyear contracts and deal constructs that leaned into the multiyear agreements, but we only had 6 months to go after it. But that led us in '25, to only have to renew 74% of our installed base. And the progress you've seen in Q1 through Q3 would lead you to believe that that percent is going drop even further as we go into next year. Well, what happens when you have an installed base that is much lower as you have to renew, you have one, less risk on churn dollars, just basic math, my gross retention is going up, that's great. It's applied to a lower balance. Guess what? Churn dollars are lower. But the other thing is you get a lot of capacity back in the sales team. They don't have to spend as much time working on the renewal. They can spend much more time going after new logos, and they can spend much more time in driving the value proposition of expansion with our clients. Only 39% of our customers are actually on multiproducts, there's a huge opportunity just going into our installed base and just doing expansion. But we know too that some of these new products we're bringing out are opening up the aperture of where we can have use cases, especially in marketing and analytics. So as we move into next year, focusing on driving many, many more enterprise engagements is where we now have the capacity, even though we're not having to bring on a lot of new reps. That helps with sales and marketing as a percentage of revenue dropping. So all these things I was explaining to the Board yesterday that we've done, it's helping us to instrument the financial expectations we have on our multiyear plan. And so it's not a question of what anymore. It's how fast you can do it to achieve the results you expect. A great example of that is on the sales side is like when you do territory redesign, you can't rip apart all of your account and rep relationships at once. It would create chaos. So you've got to be very methodical about how you do it, and making sure that you're setting up new reps in new territories with existing customers to talk to. Well, if you do that, you have to take existing customers away from the reps. So that level of churn in chaos is one that you want to -- you know you have to do, but you want to kind of keep it like 30%, 35%, not have a tremendous amount of churn. And if you do that well, then the system starts to expect it. And as the system expects it, then you're ramping your reps faster, they're reaching attainment faster, your attrition falls down, you're instrumenting the outcomes you're expecting.

Yes. And to your point, you have to be very methodical, especially on these older cohorts of customers. But maybe a way to bring better clarity to this is what are you seeing in terms of NRRs with the newer customer cohorts? So the post COVID, right, where you've had some of these changes potentially implemented faster, like what are the expansion rates look like there? And could that give us insight into what the older customer cohorts could potentially ultimately become?

Yes. So I think that what you'd find is a gross retention rate you think is nearing best-in-class in some of those, especially given the -- many of them are multiproduct when they're adopting and the best indication of a return customer is one that already adopt multiproduct. So we don't even have a long history of that. So you can't tell, "Hey, well, it's a small cohort". But the reality is we're seeing expansions happen within the year to the tune of 5 to 20x your initial land. That happened with Empower, where we started very small and it went over to $1 million deal. I can tell you there are multiple customers I've talked with this quarter where we're starting off with a very small and the expansions are showing up or we're even landing with much larger deals. So go back -- that goes back to the value proposition of selling the full platform and the customer journey and what they can expect as they adopt more and what outcomes we're instrumenting in the sales process. Those were things we put through the instrumentation I'm talking about and how we're changing our go-to-market. It's dovetailed very well with the products themselves. You can't have one without the other. But -- but yes, that's -- it's showing up very well. And I would tell you the dollar retention rates on customers in the enterprise, whereas I'd say that as we've shifted our focus on enterprise, they are 3 and 4 points higher than the average. So the more our business becomes enterprise, and it's creeped up every quarter, the less and less you'll be asking me that question because the reality is enterprise customers are so much better than in the SMB mid-market with respect to gross retention and net dollar retention.

Yes. Let's focus in on the cross-sell opportunity. So -- when you -- so a couple of years ago, Amplitude was primarily product analytics. And over the last several years, it's really been building out to become a suite of a number of solutions. I don't know if you can give us exact numbers, but maybe just even providing color to help us quantify it, how much of the business is product analytics versus nonproduct analytics today? And when you have a product analytics customer and you upsell or cross-sell all of these other solutions, like what impact does that have to average deal sizes?

So I'll answer the last question. It's much larger. So average deal sizes have gone up quite dramatically with multiproduct. And part of that is the way we price. It's -- you pay for your analytics and the data associated with your use case is downloaded into our system, and that becomes the cost to serve, if you will. Session Replay has a little bit of that. But the other modules have very low marginal incremental cost to serve, and yet we're able to -- because of the value that they deliver, we're able to add on to that value. So you can get a customer that is analytics, it's call it, 30% of value in the contract, but the rest of the modules make up 70% or more. So when we add on to an analytics customer, it's a 25% to 30% uplift typically by module. And what I'd tell you now is that already the way we disclosed it, and then I'll tell you what you're asking, which is the customers that have multiproduct now represents 71% of our ARR. Only 39% of our customers are multiproduct. So there's a big difference in ASP. And then the other thing is, 2 years ago, we had virtually 0 revenue from other products. Now it's upwards to 30%. So you're seeing us leverage the strong product analytics foundation and broaden it into these broader use cases and customers are seeing and paying for the value of the activation layer that we've added.

Yes. And if I put AI aside because we'll touch on AI in a moment, but it seems like there's 3 big opportunities. So we've been talking about cross-sell, right? So I'd love to hear you've made an amazing improvement today, but where could that momentum go? And how are you thinking about that? The second piece would be, I think, historically, Amplitude has been focused on product teams, and there have been efforts to push more into marketing and other departments. So two, can you talk about -- not to throw too many things, but two, talk about that opportunity. And then the last one, I think the big early adopters of Amplitude were truly the digital native companies, and there's a big opportunity to go after those traditional established conventional companies. So where are we on the tipping -- like the tipping point for those?

Yes. So the -- on the more traditional companies, I would tell you, every day we see health care, financial services, media and telecom -- that kind of First American titles when we shared it on an earnings call. That's a title company around transferring property and like their business process automation was digitizing that application environment, and they're using Amplitude as the underpinning for it. A number of other -- Sutter health. I don't know if you're -- I'm a Sutter Health. My Sutter Health, Amplitude is the underpinnings of my Sutter health. And just a number of them that in very traditional industries that are figuring out ways in which they want to digitally engage with their own clients and are instrumenting and using Amplitude to do that. I don't know what your favorite fast casual restaurant is, but we have a lot of them under contract, Jersey Mike's, Chick-fil-A, Burger King, Dairy Queen, you name it. A lot of them have figured out that if they want to have a loyalty program with their clients, they need to meet them where they are and often that is on their mobile application. And so they're delivering promotions and engagement through that environment, and they're using the Amplitude to do it. So I think there's been a great push, I think the progress that we will have in more, call it Global 2000 and less digital native clients will be the pace at which they're engaging digitally with their clients. And for us, we think that as we move into adding more and more agentic capabilities to our platform and add more products, the reality is that's going to spur that interest and engagement. In fact, one of the things we showed on our earnings call, we know we do demos. We do demos of the product in our earning call was the whole MCP server capabilities. And for us, it was about opening up the aperture around who can engage with Amplitude without having to have a data science degree, understanding the data taxonomy or be a marketing analyst or a product analyst. And you'll see in January that same prompt-related interface is both going to be within Amplitude and where you can easily generate charts or ask queries or you could easily engage it through your favorite LLM model.

Yes. Let's talk AI because Amplitude has been really innovative on that front. So you talked about the MCP servers, Amplitude's introduced AI agents, even more recently, you guys have entered the AEO/GEO conversation as well, too. So there's a lot going on at Amplitude. But just in terms of the areas that you as a CFO, right, and you think about the impact that could have on retention, right, or direct monetization, which one are -- maybe there's a couple are you most excited about? And what has early customer feedback then around adoption and readiness for these solutions?

So I guess I'm probably most interested in the core platform areas where we've had agentic capabilities. Like if you have an agent that can actually do experimentations continuously. I've talked to customers that talk about how they run an experiment every 2 weeks, and it's one. And they have to involve all these different parties within the organization, IT and data analysts and the engineering team, and they have to have handoffs and sign-offs. And this discussion I have with I would call our marketing analyst, they have different names. We're talking about person, but she's the marketing house. And she leaned over to the IT person, let's call them Tom and said, "I don't want to talk to Tom. I just want to run the operation experiment knowing that he set up the framework directly in the data tech, they don't have to ask". And it's very funny because Tom, I'm on. But the funny thing is, she's right. She's right, that is what they need to do. They need to increasingly think about how they can run the experiments more rapidly and the agents can do that for them. You can set up the framework and run thousands of experiments continuously and make recommendations and propose a guidance survey engagement or a cohort targeting, different ways of looking at the analysis, deep dives into the reason and the causes of what you could do with it. So experimentation is a really, really interesting area. And I call it the death of the data scientists because you're going to need a lot less. Another one is a evaluation of your Session Replay, that's -- if you think about Session Replay's video. It's watching people's cursors and how they're interacting with your application environment. Well, One of the things that agents are really good at is taking lots of data and summarizing it for you. Now I think in terms of we have an agent that's constantly doing it and providing you the right actionable insights from that qualitative and then quickly engaging workflow, how you can make the changes to your website or colorization. Now it's fast forward. As you start to make that more autonomous, you enable personalization such that the application website that you would look at would be different from the one that I look at based upon the batch size of one. They're making the changes based upon the behavioral heist that you exhibit versus me. And that's something that's pretty powerful when you think about it in terms of targeting, marketing, advertising because you can then optimize your conversion rates associated with them. So I'm really bullish about some of those things, and I see the platform becoming more accessible, less about the UI and understanding Amplitude, but more about Amplitude being catered to your business needs. The other areas on opening up the data, MCP is very, very interesting because you get access to not just what's ingested into Amplitude, they get to support tickets to edit feedback to how you're -- how your CRM data is structured. And it becomes less of a task to ETL information from one place to the other because your data taxonomy becomes something that is -- think of it in terms of -- it's not a relational database required anymore. It's unstructured data and the agent understands that. So those things are exciting. And I also think some of our new products are pretty cool. I'm -- when we acquired Kraftful, they were in the voice of the customer space. And if you talk to Yana Welinder, who was the founder of that and is now part of our leader in our engineering team, to tell you that we're going to go disrupt Qualtrics and Medallia because they're doing basic survey stuff and that's one slice of customer sentiment, whereas that's important, you should have a broader sense with many more channels to establish that sentiment. And if you're not looking at how customers are interacting on blogs and websites and feedback areas, and you're missing some important feedback.

Yes. How are you thinking about monetization as it relates to all of these different solutions and initiatives if I recall, and correct me if I'm wrong, I think some of the past messaging has been you guys are really focused on just driving usage and adoption. But where -- even if that might be the goal today, when you look 2, 3 years from now, how do you see that playing out and impacting revenue potentially?

So there's a couple of different areas. I'd say the agents that we're very much in the platform people like to say it's free. I said they don't use the effort. The reality is that we have great monetization associated with usage and more of that increases usage, we'll monetize that and the more of that increases the adoption of the platform, we'll monetize that. But there are also new products that we will charge for. There are new modules. AI feedback is one we're charging for, and we will continue to charge for. You saw AI visibility, right? And that is in the agentic search optimization area, that's going to lead to some products too because what we did, we gave it away for free, but it was one that just kind of tested and poked at the world that is like the profound of the world saying that that's not a new product, that's a feature. And what you really need to do is not get the insight, you need to understand what you do with the insight, how you optimize from that insight. So we will definitely sell products in that space and be increasingly focused on developing our own capabilities to displace Google and Adobe.

Perfect. And on the topic of AI, I got to ask you about OpenAI's acquisition of Statsig, which has some overlap on the product analytics side, but they're more of an experimentation player from our understanding. But I think when OpenAI made that acquisition, there were some investor questions around, does that mean that OpenAI might further get into Amplitude's core market, right? So what are your guys' thoughts on OpenAI's acquisition, what that means for the direction they're potentially taking, do you view this as validating, right, the opportunity you guys are chasing? Do you see them as a potential rival in the future? How do you think about that?

So our initial response was very much, oh, it's validation, much like when Datadog acquired Eppo, experimentation is an important aspect for anybody who's trying to figure out how they bank their digital engagement better for clients. What I can tell you is we saw Statsig once in a while in very scientific-related experimentation opportunities and data warehouse native opportunities, but it was never really that large. It was mainly in smaller implementations. But when it was acquired, that's what we thought. And then we got more and more feedback from people who are giving us insights, and they told us, well, half of that Statsig's ARR was from OpenAI. Oh, and 2 CEOs, they were at Meta at the same time. And they brought in and wanted him to be the CEO of Statsig, wanted him to be a Head of product. And suddenly it looks like an acquihire, not really a product continuation. And I would tell you, we don't really see them at all anymore. Now that doesn't mean that they won't do something, but I suspect that this was purely an in-sourcing and talent acquisition for them. And I would tell you that increasingly, we're seeing more and more AI native-related companies use Amplitude, you know that Cursor does. You know that, Granola does, you know that Character.AI does and might be another one of those major LLM model providers out there that's using Amplitude.

Yes, those definitely are great, great proof points. Let's look on the P&L now and shift gears a little bit. So net new ARR growth has been a blowout metric. So 2 quarters ago, that grew 200%. Last quarter, it was still up an amazing 45% year-over-year, tends to be a little bit lighter of a bookings quarter seasonally for you guys. But I guess just as we look ahead for that metric, we're going to start to come up against some tougher compares. So is that something we need to be mindful of in our models? Do you guys still feel like there's tons of momentum and runway left that some of the levels that we've seen can continue to be more durable? I guess how should we just think about the trajectory potentially of that metric as we look ahead?

Yes. I think there's a couple of things that I would mention. So first and foremost, I talked about that we are kind of instituting a better and better sales productivity framework for our teams. And so they've always been good on the growth side, but we've had that the headwind on the churn issue. Well, if we're doing better on the churn, variably you get better on the net aspect. So I think at the current capacity levels, we can continue to drive. And we've got -- even without the AI products, I think we still have a great one way on just driving consolidation in the marketplace. We're seeing great demand from both traditional and nontraditional customer bases. So I think we're in good shape there. But I think longer term, I think you're right, it gets harder from a compare perspective on a dollar basis. And that's when we start pulling the lever on capacity. We start adding more. If we're executing well on an efficiency perspective, that's when you want to invest in your go-to-market team and start building the capabilities to drive that growth harder.

Perfect. And maybe in the last 30 seconds or so that we have. When we think about the balancing of growth and profitability going forward, I'd love to get your thoughts there. Obviously, you guys have been executing really well on the improving growth story. When we think about profitability and those 2 in combination, do you see a path getting back to a metric like Rule of 40? How are you thinking about that? What are going to be the key unlocks? Is it a function of you've been doing all the right things. We just need to see revenue growth accelerate and that will unlock a lot on the margin side? Is there still areas of cost efficiency that you're looking at closely. Maybe you could just give the group -- what your high-level thoughts are there?

I would say it's never one of those binary things, and you're always trying to balance it out. I hear there's a big opportunity for technologies out there for us to bring in at a very low cost. And so occasionally, we'll look at making those growth investments. But I will tell you that the rest of the company just saw our approved budget at the Board level for next year, and it's very much in line with growth with leverage focus. And you're always making trade-offs, but I'll give you an example. The product team came up with a long list of things you wanted to go do. And the reality is I told them, here's the line that we can afford. So figure out with sales, what are the most important things we work on. And I think that's what investors are expecting. They're expecting us to make those hard decisions and drive towards a more and more profitable business. I will tell you that we think about growth and free cash flow margin that we're mapping out a path for Rule of 40.

Perfect. Awesome. Well, we'll end it there. Thank you, everyone, for joining. And Andrew, I appreciate all the time. I learned a lot. So thanks for joining us.

Thank you.