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

Good afternoon, everyone. Tyler Radke here. Thanks for sticking with us for the software track, been a busy day. We're excited to close out the day with Amplitude CFO, Andrew Casey. Andrew, thank you for joining the tech conference. Nice to see you again.

Nice to see you as well.

For folks less familiar with Amplitude, how about you just give a quick overview of the business?

Sure. So Amplitude got its start by trying to solve the problem that every customer who is trying to digitally engage with their clients, whether that be through a mobile application, a website or classic desktop applications, was trying to figure out how they improve that application, how they improve that level of service. And they were typically using screen captures or sessions to try to understand that interaction. And Amplitude came along with a product that thought we could do better at a very detailed level. We call events. And those events are -- think of it in terms of how you're scrolling across things, how you're engaging with, how much time you spend on content. And that led to an application that gives that feedback to developers and product creators to constantly improve their product. I mean people ask me all the time, like once the fill in your blank application has used you, what are they going to use you for next? And so, they're constantly changing in how they're trying to better provide those services to their clients, better provide more curated personalization. And so Amplitude is used constantly by application developers to try to improve that product or improve that level of service. We've got everything from classic digital natives like a DoorDash to -- and B2C companies to a company we even announced last quarter, First American Title, who deals with title and mortgage processes, and they're instrumenting in modernizing how that process works. So Amplitude is started with this unique position of trying to create better products, better services. But as we've -- as the space evolved, there were all these other areas that popped up, different capabilities around experimentation, session replay, guides and surveys, but they were all connected back into an analytics core. And the postulate for us about 3 years ago was why don't we start incorporating some of these capabilities into more of a platform approach. And over the last couple of years, I'd say the innovation pace is when we've created more and more capabilities in the platform and started to ingest that into a better together story for our clients. And so now today, Amplitude is not just core analytics, it's much more than that. It's reaching out to more personas like in marketing analytics and customer analytics. And we certainly view that there's an increasing effort by enterprise clients to bring together various data silos so they can get a better picture of how they're best serving the client.

Right. Got it. And I think the original product analytics story was more about at least from an ROI perspective of doing things more efficiently versus an in-house or homegrown solution. Is it now more kind of a consolidation play in terms of knocking off a bunch of these other features that you might have to buy separately or...

I still think there's a group of, call it, the Fortune 500 or the Global 2000 that are moving for the first time to digitally engage with their clients. I said that First American Tile is a good example of that. But we see those all the time. Like they're moving from paper-based systems and trying to engage digitally. And so that -- the story there is we're talking about the value proposition of instrumenting that engagement for the first time. But I acknowledge that as we've added more capabilities, a very typical conversation I had one yesterday with a client in our New York office was about how they could bring together the different groups and the different silos and efforts through one integrated platform such that web analytics and the mobile analytics and their data silos could all be brought together and they could break down some of the workflows in a more efficient way. And so the opportunity there for us is replacing a session replay vendor, replacing a web analytics vendor and establishing a core taxonomy across all those environments for them to be more efficient.

Got it. Got it. Okay. And the customer base has grown quite a bit since -- well before your time, but the time a few years ago when the company went public. But how does that look like just in terms of the types of customers? Is it still primarily digital native? You talked about the North American title company. Obviously, that's a more traditional enterprise bias. But where does that kind of composition look today?

I think that's -- it's a big misnomer. We talk about that the -- if you look at from a classic NIC code basis on our customer base, it's pretty broad. We still have about 30% plus that is, I would call, software digital native, and there's some classic customers in that space. Some of the largest software companies in the world use Amplitude as their core instrumentation. But increasingly, we're seeing more nontraditional classic enterprises that are trying to figure out how they reach out to customers in a more effective way. Now we talked about the retail environment is increasingly expanding companies like Walmart and others that are digitally engaging. But two, it's funny this last quarter, we had this big real estate conference like a number of customers had come forward like Irvine Company. So real estate, health care, telecommunications and media, publishing. We talked about the economists on our last call, and they're moving from print to a digital subscription business. So it's really quite broad-based. And financial services, media, telecom, the software business is certainly the biggest areas, but increasing, we're seeing more and more.

Yes. And I'd be remiss not to talk about AI.

How long did it take us?

Only 7 minutes, but how -- like high level, there's obviously a lot of different ways we can go with that. But like high level, how do you think about the way apps and product development is changing and how kind of Amplitude can be tethered to kind of those changes?

Yes. So I think the first thing, when we look at AI as simply another piece of software. And frankly, some of the best implementations of AI have been around developer productivity tools. And for those of us who have been in the industry a long time, there's been just a continuous trend around how do you drive software creations more effectively. When you think about in terms of moving from instruction set architectures to middleware, from middleware into using open source and open source libraries. And this is just the next implementation of it. So the reality is more software is good for us. More software means more need for instrumentation and understanding how that software is interacting with users. And there is a large group of AI companies that are using Amplitude as their core instrumentation. Some of the larger ones, too, that we can't necessarily talk about. But the reality is we think that this is actually a buoy for us that we're going to see more and more usage of Amplitude in nontraditional cases because of the adoption of AI. Now one of the things we are implementing Agentic capabilities within our platform. And it's its core focus is going to be about driving optimizations and better utilizing Amplitude. So where before customers may have had 5 data scientists, the business analysts and IT working on an experimentation platform, we think you can create an agent that can do double, triple, quadruple the amount of work that the team used to do. So increasing the level of productivity, even if you're not getting rid of them, you can increase the level of productivity associated because an agent can run thousands and thousands of experiments and give you recommendations based upon the insights that's driving from those and take action if you want them to.

Right, right, right. Got it. And on that digital native customer base, I know there's been some talks about like the cursors of the world, and there's various other companies out there that have sort of sized AI -- sorry, AI natives. Is that a significant like end market or customer for you, like more than 1%? Or is it still pretty low in terms of revenue today?

I would say in the context of like all the rest of the customers are still relatively low. But I would say we've had some really good positive trends on the size of those contracts, like there -- they started out small. And there's -- John and I were talking earlier with investors that there isn't a week that goes by that we don't see a win where with some company that's related to AI technologies, and you don't really know them, but at some point, they may become very, very large.

Right, right. And you alluded to kind of the future of product development as you have agents both on the coding agents. Many of those are your customers, the curses of the world and then you have your own agents in terms of driving those recommendations. Like how do you see that evolution long term? Do you think this is something that eventually gets embedded by some of these coding agents? I think recently this week, OpenAI made an acquisition in the space. I know Datadog has made some acquisitions around the product analytics space. So obviously, validating maybe the of the market, but how do you kind of see that playing out long term?

Well, first, I would say it's a good validation of the importance of experimentation for any software developer, because I think that the notion that you don't need that capability means that you're ignoring some of the feedback that comes back from how users are interacting. And so experimentation is a big field for us. It's certainly one of our biggest areas of investment and that has been growing rapidly. And I don't deny that there will be some cases where software companies will decide they want to do it themselves. I would say that some of the largest software companies in the world have chosen not to specify that, and they've chosen to use Amplitude for that, because they want to spend their time on developing the application and using the tools to actually enhance their capabilities on that development process. So there's always that push and pull. I think for us, it's constantly driving value for our clients such that they don't look at it as a cost that they could do better internally that we can always provide greater value to them. And the more that we expand beyond just product analytics into more capabilities and drive workflows and drive better relevance to a broader set of use cases and users, I think that really establishes a broader value proposition for us. As we move into more marketing analytics use cases, core customer analytics use cases to workflow optimization, I think that, that enables us to increasingly drive value.

Got it. One of the big focuses for the company even before you got there was sort of managing down the churn, which kind of became elevated. And I know you and I have talked, there were some pretty unusual contract terms for some of these customers that had maybe some unusual terms around downsells built in. But just as you think about like that composition of ARR like between kind of event volumes and now you're in a lot of different markets, like how diversified is that in terms of multiproduct or any way to kind of cut the data just to help us understand like the platform or multiproduct nature of the business?

Yes. So I'd say, remember, too, that some of the products are not -- haven't been in market that long. And so they each have their own maturity curve and adoption curve. But the ones that we've introduced that have been in the market longest like experimentation, activation, web analytics and then we went into session replay and then guides and surveys and recently voice of the customer. All those are increasingly driving greater and greater adoption within the platform. And the metrics we share with investors are now of our ARR base, 67% are multiproduct beyond just product analytics. Now most of that is customers who have 2 products, not like 5 or 6. But that to tell you that there's great opportunity for us to continue to drive cross-sell just into our installed base. In fact, at our Investor Day, we sized that back in Q1, that was a $160 million opportunity just with the existing installed base. not to mention what we would do with new customers. Now it's also true that we're -- of the number of customers we have, only 34% of them are considered multiproduct. So there's still a large portion of our customer number that is on product analytics. But if you think about it, a minute, 67% of your ARR, but they're only 30% of the customers. That should tell you that the average ARR for customers who are multiproduct is much, much larger. And many of the customers who have adopted the full platform, 5 products are multimillion dollar implementation. So the ability for us to drive value is there. And customers recognize it. Like I said, the customer I was speaking with yesterday had a multimillion dollar investment into this space across 6 different vendors. And we can constantly go in and say, well, we can replace that all. And not only we can give you better value for your dollar spend, we can show you how you can optimize the operations within your business and just more data so you're getting more and more relevant as how you derive customer sentiment increasingly expands beyond just the digital footprints in which they're interacting.

Right. Right. So obviously, getting into larger contracts, part of that is products, having more products to sell. But go-to-market, I know there's been a lot of improvements there, just kind of maturing the sales force and bring in some new leadership. So maybe just talk to us how that's going? What are some of the things you're doing from a go-to-market perspective to drive those larger deals?

So I'd say when I first arrived, we certainly had the aspiration to go focus on more enterprise, but there's a lot of practices and procedures -- as you noted, they were resulting in poor contract structures and that we really didn't have the incentive aligned or the processes aligned. And so some of the first things we did was just, as you suggest, bringing in more enterprise sellers who know how to do value-based selling as opposed to transactional selling. And I know that's a simplified thing, but it really is a lot of nuance about understanding from the customer perspective, how you're going to drive that outcome and driving constructs and the processes that follow from that, how you orient your opportunity funnel, how you drive your demand creation at the top of the funnel, how you're incenting your sales reps. One of the first things we did is we moved from a gross ARR quota basis to a net ARR quota basis. So I love to tell people, show me a company that compensates their sales reps on gross ARR, and I'll show you one that has a churn problem. So certainly, we had that issue and switching to that basis drove accountability down to -- all the way down to the sales rep about being the primary interface for all commercial transactions associated with the customer in your territory. So that was a big change. The other change was, as you were suggesting, is teaching them how to sell the platform effectively. A lot of our sellers got very used to selling only product analytics, not selling all the different aspects of the platform itself and being able to go and have conversations about why we're better than full story, why we're better than Optimizely, how we can show the benefit in a use case that is based upon the customers' workflow, not just our technology capabilities. So that was a big change we did at the beginning of the year. And then the other thing that we did that I think was really going to be positive, helping us to reduce churn longer term is incenting customers -- the sellers to actually drive multiyear contracts with their customers. Now in many cases, customers who are adopting the platform, this dovetails well with the platform is going. Many times, they're not thinking of a big bang, I'm going to replace all 5 products all at once. They have a vision of how they're going to replace those implementations with amplitude over a period of time. And they'll want to have predictability on what the costs are going to be associated with that. So they're more willing to engage in a multiyear contract discussion than maybe somebody who's just single product. That's the first thing. And teaching our sales reps how to actually do those constructs such that the customer has the perception that they're getting value in events of what they're actually paying for is the art of it, right? Now when you get that right, you start to see that customers are leaning in and they're willing to think more longer-term contracts. And that's now showing up in our RPO growth. Our RPO growth grew 31% last quarter. Our long-term RPO grew 63% last quarter, and that's continuing to give us better and better visibility into our revenue stream that's going to happen over the next year and beyond. So it gives us greater predictability. Now here's the thing that people don't understand with respect to churn. When I first joined Amplitude, we had to renew 89% of our installed base, 89%. That's a lot. And that's a lot of sales annually. That's a lot of sales time spent on just the renewal process. Now through our efforts in the second half of '24, we came into '25 with having to renew 72% of our installed base. And if we execute well the remainder of this year, that should drop closer to 60%. Now think about that. Our ARR has increased during this period, and yet I reduced the amount I actually have to renew. If I reduce the amount, I actually have to renew, even if I had gross retention rates at the same rate, which is not, it's been improving. But even if I had the same rate, the actual dollars in churn should go down. And sales productivity should go up because they have more time to spend on new business and on expansions with existing clients. So structurally, there's no way you can say that that's going to be the panacea for driving churn down. You have to deliver value. But structurally, if you're running your SaaS business the right way in an enterprise sense, you're building the capabilities to see improvements.

And you're not having to give up like crazy discounting on those multiyear deals, just kind of standard 3-year terms.

And that's because enterprise customers, they value cost predictability over this notion of being fearful of paying for something they're not using. And that gets back to the construct.

Right. And as the product portfolio has evolved and maybe with the rise of AI, like what -- how would you sort of articulate like what budgets and lines of businesses you're tapping into? Because I think in the go-go days of 2021, it was anything developer related, just throw money at it. Is it this coming out of an AI budget? Is it still kind of being scrutinized with the same scrutiny as you saw during the wave of cloud optimization? I'd just be curious how that funding dynamic and sales cycle has evolved?

So I'd still say we have a predominance of appealing to Chief Product Officers, Chief Digital Officers. I'd still say that we're very well known for our product analytics capabilities, and there's plenty of opportunities that come up through that. Our sales team is instructed to make sure they broaden and make sure that customers understand all the value they can get out of the investment in Amplitude, but there's still that. Increasingly, though, we're seeing more and more Chief Marketing Officers, Chief Revenue Officers. And I will tell you, the customer I was talking to before, we had all of them. We had 17 people on the call all looking at Amplitude, and they were looking from their various angles. Somebody who is responsible for the mobile application, somebody who's responsible for the web application, somebody responsible for the data taxonomy. And they were all trying to figure out how they could allocate portions of their budget so they could use one consolidated tool. And ultimately, at the end of that meeting, they asked me if I would go talk to their CFO because they've got a great business value proposition. They said, help me look at it and rationalize it to your peer in my company because this is going to be a large investment for us. So I still think that to your question, is there still scrutiny? Yes. Do I think that our value proposition has gotten better and that's easier to go through? Yes, but there's still the work that needs to be done.

Right. And I think the ROI was always a bit of a question that I think investors had on Amplitude. And obviously, with the expansion of the product portfolio, I think that's helped. But like in that use case, just like walk us through that conversation you had with the CFO, like what was the ROI? How is it measured? And I assume that was a customer success story. So I'm sure you did convince the CFO, but...

The first part was we walked in there with them having spending, let's say, multimillion dollars on the various applications they had. And what we were quoting was certainly below that. So we were already were showing, hey, we can save you money simply on the consolidation efforts. Now the thing that, frankly, the CFO was quite savvy and asked me, well, how can you show me that you're really confident on driving this? And I said, well, we already have through the proof of concept. We've shown here we optimize your workflows. And we've organized this such that as you're rolling out these new applications, you -- the amount that you're effectively paying for is increasing as you adopt. And if you move faster than that, then you get value in advance of what you're actually paying for because we no longer have -- a platform approach is one where you enable the customer to actually get access to the capabilities very easily in their journey. A point product approach is, no, I'm going to gate that. I'm going to make sure I'm nickeling and diming you every time you move into every application. And that switch was one that I think the CFO got confidence that, one, we would save them on licensing; and two, we could show workflow optimization based on the results we did with their own data and their own use cases.

Right, right. Got it. Okay. And I guess as you just think about margins, we've talked a lot about growth, and I know top line have this acceleration trend into the back end of the year. Like how do you think about kind of that long-term framework, Rule of 40, et cetera?

So it's funny as when we rolled out in March at Investor Day, we talked about how we're going to have a growth with leverage focus. And we're going to increasingly drive optimizations on how we operate from a sales and marketing perspective, G&A perspective, certainly going to drive greater optimizations of how our products and services run within our hosting providers' environments. But I actually did this at our leadership meeting recently as well, reminding all of our executives about how the budget is going to be rolled out over not just this year, but the years to follow. And that I'm expecting that we're going to grow our revenues faster than our expenses, which shows greater leverage across each of those major environments. And they will have their own targets to hit those levels. Now the interesting thing is most people look at, oh my God, you're going to cut our budgets. And I said, "No, no, you're not understanding. We have a plan to grow revenue. We're going to have a plan to grow expenses slower, which means that we're going to show operational leverage. We're going to show positive operating income over the next few years. And certainly, that's the way I'm setting up our budgets for even next year." Now how does that get instrumented? Well, there are actions. From an engineering perspective, it's how do we get the marginal incremental cost of data to go down, right? From how do we get our professional services business to be not a drag on gross margins as much as it is. How do we make sure that our sales team is getting more and more efficient, such that the sales and marketing as a percentage of revenue goes down from it's 44% down into the mid-30s over time. And how do we get G&A to be lower and lower. Now to be clear, we've made improvements there. We've seen improvements as a percentage of revenue, and we'll continue to do that.

Right, right. And then on the M&A front, I know part of the margin guide for this year contemplates the recent acquisition. But how are you thinking about M&A going forward, just the types of technologies and businesses that you're interested in?

So when we approach acquisitions, they usually come from a road map discussion where we've got a capability that we want and we determine is it best to build this organically? Or is there an inorganic opportunity for us to fill that gap quickly. And typically, I would say software businesses acquire others for 3 reasons. They want to acquire customers, they want to acquire technology or they want to acquire teams. And I think if you can get the last 2, it usually works out really well. And there's cultural mix and vision mix and all those are positive. And so we've looked at larger potential what people call transformative acquisitions. But the reality is most of the time, those don't work out so well. And you rarely get the value really expecting because if customers wanted to be acquired by you, then they'd simply switch, right? So those acquisitions, I'm a little less interested in. What I'm interested in is where we can supplement our technology, bring that to market quickly and have a great team that mixes with ours. I would say with Command AI, that's definitely been proven out. We've acquired them. They are the foundation for our guides and surveys. And a good portion of some of that technology is coming to build our Agentic capabilities. With Inari and with Kraftful, I would say each in their independent areas, those teams have really been additive to ours. And certainly, we think we can take the Kraftful product, which is relatively early in its maturity, but it's not a huge stretch to think that the evolution of voice of the customer can be one that goes after the installed bases for Qualtrics and Medallia.

Right, right. And I guess on just the talent environment. I mean, obviously, there's kind of this bifurcation where the expensive AI engineers are getting multiples of top NFL contracts or whatever. But how are you sort of managing that from your perspective? On the other hand, there seems to be maybe a surplus of just kind of broad-based folks in computer science. But is that something being predominantly in San Francisco that's been a challenge? Or have you kind of seen some improvement on the talent front?

I think that there's -- well, first of all, I think getting the right talent is always a challenge, right? You're always focused on that. And I think for us, in particular, where we brought in, call it, the core AI talent into the company, it's because we've had strong relationships with them. There's -- our founders have a really strong relationship with the start-up group within San Francisco. And certainly, they continue to work and associate with that group. And I can tell you, Yana, who came over from Kraftful, we've had long conversations with them and the others were all Y Combinator-based founders. And so there's that aspect. I think the other thing is a lot of them chose to come be a part of Amplitude because they aligned with our vision and focus on how they could bring their technology to a broader distribution. And really contribute to something meaningful. And I think that's allowed us to bring in great talent. The other thing I would tell you is it's very interesting that increasingly see great talent coming from basic universities where they taught them to use the tools around AI to create applications and the infrastructure. And we had a coding challenge recently within Amplitude with the interns, and they were pitted against our engineers throughout Amplitude. And they were out of the top 5, 3 of them were interns. So you've got talent now that's coming up and have lived and grown up in using those tool sets and they're not as expensive as you expect.

Yes.

But I do think that's always a challenge to get the right talent because you got to have a mix. It's like a good sports team, right? You got to have good veterans and you got to have good rookies as well.

Right, right. Got it. And then I guess just sort of turning to the recent results and as we close out here, thinking about into year-end, I think this last quarter was one of the strongest net new ARR quarters you had in quite a while. How are you seeing the pipeline evolve? Obviously, it sounds like you got some positivity in terms of the salespeople less focused on a larger renewal base because of the multiyear nature. But yes, how are you feeling just heading into year-end?

We feel good. We feel good that -- look, the conversations we've had with enterprise clients is positive. They're definitely seeing the value proposition. I think that our pipeline coverage has improved. Maturity could always be better. That comes with sales process and improvements there, but we feel good about it. We feel like this is going -- kind of going our direction.

Yes. Yes. No, that's great. Well, great. I wanted to close out. I know you've had a busy day of meetings, but if there's anything else you wanted to leave for the audience before we close out, I'll turn it back over to you.

So the one thing I would tell you that increasingly, we're talking to investors about is what distinguishes us from other SaaS-based companies. In the wake of our earnings, it was interesting. We went to a conference right afterwards. And the typical start was great quarter. Sorry you're not getting recognized for it. But that's just because you're front office SaaS. And I think...

You're not seat-based, right?

Exactly. Exactly. And that's the misnomer. I think increasingly, we're trying to educate that -- we'll lean into every optimization associated with AI there is. In fact, we're instrumenting into our platform because we certainly believe that as customers use more AI, it actually benefits us with greater data ingestion. And it showcases the value of the platform being well connected because workflows tend to be optimized as well. So that's something that I think we're trying to make sure people know more and more. And frankly, the growth that we've had over the last year has predominantly been associated with cross-sell. So once we get past some of these issues that we've created for ourselves, we'll actually start seeing benefits from upsell and greater data ingestion as well.

Yes. Yes. That's a great place to end. Andrew, thank you very much. Thanks, everyone, for joining us today. And we -- there might be a couple more sessions, but otherwise, we'll see you tomorrow for the last day.

Thanks, Tyler.

Thank you.