DigitalOcean Holdings, Inc. (DOCN) Earnings Call Transcript & Summary

All right. Excellent. Thank you all for joining. My name is Josh Baer, software analyst at Morgan Stanley. And we have CFO of DigitalOcean, Matt Steinfort. I have disclosures to start, for important disclosures, please see the Morgan Stanley research disclosure website at www.morganstanley.com/researchdisclosures. If you have any questions, please reach out to your Morgan Stanley sales representative. Matt, thank you so much for joining us today. Happy to have you.

Thank you.

Wanted to start with a little bit of a -- an intro. It's been about a year since you've been on board. What are you most excited about looking ahead and specifically want to pull in, you have a new CEO. So with sort of the new energy and framework for thinking about this next year?

Yes. I'd say, I'm excited about the same things I was excited about when I joined a year ago. And first, it's like this is a phenomenal market that we play in, right? It's -- IDC's most recent report was like $113 billion, growing 25%. That's a phenomenal market to be participating in. And we've got a really nice niche that we've carved out, being the developer cloud, right, and focused on those smaller customers and developers that are trying to build businesses and experiment and learn below kind of the hyperscalers where those kind of customers aren't going to get the attention. They're not going to get the love. It's a lot simpler solution. It's simpler priced. It's easier to use. And we've got a really loyal and large customer base. I'm really excited by that. But I'm even more excited now having been here a year about the opportunity that we have to grow. And I think Paddy, as you mentioned, is going to be a key catalyst to that. The -- as you know, we're growing slower than the market today. And it was -- last year was certainly challenging for a lot of software companies. But we have the ability to accelerate. And what Paddy is bringing is a focus on the developer again, like being obsessive about the developer experience and innovating and driving kind of capabilities that build on that great relationship we have with our customers. So I'd say my first year, all of that has been reinforced and now with Paddy adding to the team, I'm very encouraged by our ability to execute in that market.

Perfect. Want to ask a follow-up on that focus on the developer. I think that was 1 of the 2 elements of Paddy's strategic priorities, the second being go-to-market. But around focus on the developer and product innovation, to me, DigitalOcean's reputation has always been focus on developers really easy to use. Like so what does that actually mean as far as like where the investments are going and what changes we might see from a product that I think developers already love?

That's a great point, Josh. We are -- we have been known for being that kind of favorite of the developers. We have tons of content and tutorials and we've done a lot to the developer to give to the developer community, and that's created the loyal customer base that we have, like north of 650,000 customers. The difference, as you said, so what's different? The difference is those customers as they come onto our platform, they grow into real businesses. So they started as developers, maybe hobbyists and they started messing around with learning code and maybe they have a website or an app that's running on us, but they spend $15, $20 a month, right? That's a lot of our customers. It's like 450,000 or so of our customers just do that. But -- and that's the way the company was started, right? That's what -- that was their core market. But as the customers grow on our platform and they build big businesses, we have customers that spend hundreds of thousands a month with us. And we have lots of customers that spend thousands of dollars a month. What they need, as they scale is different than what they needed when they were a mom-and-pop development shop. They need global load balancing, for example, where we manage their workloads across the -- our different data centers. They need more data resiliency. They need more flexibility in their backups and they become slightly more sophisticated. Now they're not anywhere near as sophisticated as an enterprise customer that's going to go to the hyperscalers, but they have basic needs as they scale their business that we believe if we can address, we can get -- we can provide them better experience on our platform. We can get a higher ARPU, and that will drive growth. And so it's really just maturing a little bit with our developer base. That's the difference. It's the same developers. But as they get bigger, they just -- they have better, more expansive needs that we can take care of and that's where we're focused.

Is any element of that moving upmarket? Or is it really about serving the SMB and providing all the tools that they need to have their business as SMBs?

Yes. It's not really moving upmarket. It's really staying with the same size customers that we target today and kind of expanding our fulfillment of the things that they could do on our platform to make them more effective. And so -- and a lot of times, again, this has been new for me. I have been here a little bit over a year. When people talk about SMB, I don't really think about this as an SMB-focused customer -- company. We focus on smaller customers, sure, but SMBs like pizza shops and retailers and small law firms. That's not us. All of our customers -- the majority of our customers are digital in some nature. They're using technology, they're developing technology. It's part of their business model, and it's smaller growing digital companies versus thinking about it, somebody on the corner that's selling milk shakes. I mean, that's not our customer base.

Similar question on the second priority around go-to-market. There's always I guess, to highlight what has changed or where the investments will be going from a go-to-market perspective?

Yes. The vast majority of our go-to-market investments, which again, if you look at our cost structure, we're highly, highly efficient go-to-market engine with our self-serve funnel is focused on harvesting the customers that come to us every month and join our platform to try to experiment and see if they're going to like it or not. So we spend a ton of money on customer success, a ton of our investment on customer success and making our customers successful as they grow, making sure that we're understanding their challenges and helping them kind of scale in the business. We spend a lot of our resource on trying to identify through both AI and through some of our own kind of initiatives. When a customer hits our platform and signs up, how do we predict that they're going to be one of the customers that grows into a builder or a scale or early in the process so we can give them a very focused onboarding experience. And so most of our sales and marketing is aimed at both the attraction of the customers to our platform. And then it's on, once they get there, engaging them in as quickly as we can and as appropriately as we can to help them accelerate on our platform. We do very little that's like -- that's what I would consider like dialing for dollars or reaching out to customers that we don't know. I'd say most of our investment in that regard is through partners, where we try to get partners to come on and they bring their own audience and we try to drive new logos through that. Most of our sales and marketing is really just trying to mine the value of the self-serve funnel that we've built.

But as far as the actual team, like what's the current size? And like is that growing? That's been a relatively new initiative not from the history of the company where it was 100% self-serviced.

Right. I'd say it's still -- again, we have hundreds of people in customer success and in some of those, what I would consider inside hunting. So they're hunting, but within -- you're hunting in your own pond kind of thing, where we've got all the customers that are coming on. We have still a modest team around the partnership channels. But we don't have any -- at this point, any material dedicated resources on the kind of cold outbound. It's mostly just focused on the growing and enhancing the self-serve.

In the intro, you talked about growing below the growth of the overall market. I mean, the overall market and certainly your customer base faced a lot of macro headwinds in 2023. So wanted to ask you for an update there, if you could characterize the current spending environment, consumption trends, what are you seeing out there?

I'd say "cautiously optimistic" is the word I would use, both for our customers and what we hear from them and what we're seeing. So as you know, the kind of the decline in our growth rate bottomed in kind of second quarter or third quarter time frame. And as we said, we've seen a steady improvement in the [ core DO ] usage since then. You didn't see it in our NDR from third quarter to fourth quarter, it was still rounded to the same number at 96%. But if you unpack that a little and you look at the core DO business, we've seen kind of steady improvement. And we saw increased usage in January that continued into February. So I'm, again, cautiously optimistic that the core DO business is kind of returning to better growth. The Cloudways business has continued to do quite well. And clearly, Paperspace is pretty much all net new for us. So that's going to be a positive growth contributor as well.

And when you think about guidance for this year, is it that cautiously optimistic? Like what's embedded in guidance from a macro perspective?

Sure. So what we've said and what's included in the guidance is, we believe, very appropriate. Let's not assume any market improvement. Let's just assume that we continue to execute and we continue to see improvements in the expansion rate, mainly because of the new products that we're launching. So we'll launch new products. We launched new products last year, new capabilities will come out this year. That should increase our expansion. And so if NDR only improves on the back of our -- the product launches and things we control, we should be able to get NDR back to, say, 100 by the end of the year and towards the latter part of the year. And so we'll reduce that headwind. And then the other contributors, the growth from new customers, again, primarily through the self-serve funnel, that's a very strong contributor, and then we'll get a couple of points from Cloudways and then 3 points from Paperspace. And I think it's solid. We feel very comfortable about low-double digit growth for this year.

Perfect. Let's hit on some of those growth areas. Maybe starting with Paperspace and more broadly, AI and machine learning offerings. How are they differentiated versus some of the alternatives out there? And a question on kind of what attracts a customer to your platform. If it's more around the simplicity, like are those AI products, does that fit the theme?

Yes. I think you nailed it. That's -- the differentiator for us in this space is we're very bullish on the AI/ML market. We think it's really early -- really early stage in terms of how people are going to monetize that and how that will manifest itself in different business models, but developers want to be able to use that kind of capability to build their own businesses either to make themselves more efficient or to embed that as part of their value proposition. And just like the company did, DigitalOcean did when it was founded, the strategy and the differentiation is, okay, how can we take that technology, which the hyperscaler is going to all over, people are going to spend tons of money on, how do we make it simple and easy to use so that they can embed it in their development processes, and they can leverage it. And so to do that, it has to again be -- it can -- the economics have to be different. It has to be simple to understand. It has to be simple to use. And if we can do that, we think we have a very differentiated offer. It's a software layer differentiation, it's not a -- we're not -- I mean, if you look at the bar charts that you'll see out there on who's buying the most GPUs and it starts with hyperscalers and it's got Tesla and it's got Facebook and Meta and everybody else. Like we wouldn't even be on the same stage if you grew that chart up here. If we're just not buying at that scale, that's not where we're going to compete. Where we're going to compete is making it simple and easy for the large and growing developer community. The thing that's really exciting about AI/ML is that should drag more developers. I believe the developer market, the people doing development will go up because it will simplify the development experience, you'll have more people that can be developers. And so I just see it as a phenomenal tailwind over the long term for the market. And again, we're going to stick to our strategy around. We're all about making developers successful so they can build businesses in a way that's -- with very low barriers to entry.

Got it. You mentioned economics. So from a pricing perspective with Paperspace or AI in mind, is there also -- is lower pricing consistent with those products as it is with the rest of your portfolio?

Yes. It has to be a similar again, easy to use, less expensive than the hyperscalers for what you're getting, the price to value needs to be different. The market -- we're talking about this earlier today, the market for AI capabilities right now and if you just go to look for a GPU, and you want to lease a GPU, it's actually a pretty price competitive market, meaning that the suppliers are willing to go to, I'd say, pretty aggressive prices despite the fact that we're in probably one of the more supply-constrained environments that we've seen. It's a little bit like being in COVID when you couldn't get chips. But instead of the end price to the consumer being higher, it's actually -- there's a whole lot of competition. So again, that's why we believe that where we'll play in that market is where customers value the software and the wrapper that we put around those capabilities to make it easier to ingest into their development processes.

Another growth area you talked about was rolling out more products, more product adoption, essentially ARPU growth. What are some of the key products that are driving that growth? And then how should we track the success of adoption of more products like the overall, the ARPU metric captures more consumption broadly?

Yes. I agree with you. The ARPU metric is low. It's just odd. It depends how you divide it because you've got all these customers that spend $15. And if you get a lot more of those, your ARPU goes down. But are you worse off? No, you're not, you actually have a bigger pool of people that will eventually grow into the higher spenders. So the way I would think about it is coming back to NDR growth, we believe that we can get NDR back to 100 and above. And we're saying, we're not banking on the market to do that. We're banking on our new product capabilities and releases. And so goes NDR, so goes our reverse, so goes our product release, so goes NDR. So if you're seeing NDR improving, it's likely that it's coming from the increased adoption of these new products. And you asked it first about what kinds of products and we talked about this a little bit earlier, it's -- think of it as a scale-oriented products. So not scale like enterprise scale, but like, okay, if you're growing on our platform, you may need a little bit more diversity in terms of where your workloads are being run and you want that to happen automatically, okay? Global load balancing, that's capability that's in process. You may need tighter controls over who has access to what privileges on your developer team. Okay. Those are capabilities that we can add that enable people to just kind of mature a little bit on our platform. Those should enable us not only to grow with our existing customers, but also should make us more effective at poaching customers from larger providers where they're -- they've already got scaled teams and you say you want to move them from a more scaled platform like a hyperscaler to ours, there's some table stakes you have to have, otherwise, they're just not going to move, even if it's a better experience, it's more economic, you just have to have some of these table stakes. That's where the majority of our product development resources are going this year.

One other area that sticks out in regard to scale-oriented products would be security. Could you talk a little bit about some of your security offerings? What the opportunity is for enhancing what you're offering to your customers there?

Yes. So security is a big aspect of our platform, not just in what we provide to the customers, but what we do to protect our customers and the use of our platform from kind of bad actors. So we've got a very robust and kind of advanced security capability where we're making sure that we're monitoring and protecting the use of our platform for anything that's kind of outbound like DDoS or malware or spamming, all that kind of stuff. But we also embed a lot of those same capabilities into our products to protect our customers and their customers. So that's -- clearly, we've got DDoS capabilities and other capabilities that we provide. But it's a big aspect of running a global platform that just has to be part of the infrastructure.

A few other growth areas, Cloudways. It seems like the performance not only as far as what's expected this year, breaking out that contribution to growth, but also 2023 performed very well, maybe better than expectations, I would say. What's driving this overall? And maybe if you could talk a little bit about the -- like cross-sell and what's been driving that performance?

Yes. No, you're absolutely right, Josh. The growth in the Cloudways business was 43% year-over-year in 2023, call it, 10 points of that was a price increase that Cloudways did in the kind of the April time frame that they hadn't done, I think, in 5 years, hadn't increased the price. But even without that, you're talking 30 -- mid-30s growth still. We were able to, I'd say, boost the Cloudways business after the acquisition by pretty materially increasing the traffic volume that goes to them by basically siphoning off customers that hit our website that self-select based on some questions into a more managed experience. Prior to us owning Cloudways, we would have just lost those customers. They would have either gone away because they concluded themselves that our solution was more do-it-yourself than what they were looking for or they would have tried us and then realized that they were looking for a bit more managed experience. So we've been able to direct that traffic to Cloudways and we've been very happy with the increase in their traffic as a result. And they just -- again, they're incredibly steady, right? It's just -- they grow very predictably month-over-month. Their core metrics are pretty consistent and we feel very good about that business. You'll get some wonkiness in their NDR year-over-year the same reason we did, because they'll lap their price increase in April. But other than that, it's pretty predictable business.

Matt, one more area of growth, want to talk about customer growth. In Q4, I think all the customer accounts like across scalers, builders, learners all increased sequentially. Can you talk about some of the initiatives that might be reducing churn there? And then also just more broadly wondering like where these new customers are coming? Are they still -- are they landing primarily with droplets? Are there other products that are common entry points for new customers?

Sure. So just on the churn front, churn has been very stable, like churn has been pretty stable at historical levels, most of last year and through this year. So it's not that we were losing customers at a clip. It was that we weren't adding customers at a clip that we wanted. But the other thing to remember is, based on our model, we bring customers in. They typically spend $10, $15 a month or they try us and they're spending $50 a month, and then they grow. So when you look at builders and scalers, the majority of those are graduating on our platform. They're not coming from outside. So they're moving from the learners bucket to the builders and to the scalers bucket. So it's the ramp of our spend. And that was the -- when I first got here, I'm like, this is an amazing model. You bring in all these customers and then some portion of them just start to take off. And so that's the way to think about that. What we're doing to drive the top of the funnel, which is the learners is we're investing in, in kind of our search engine optimization capabilities and our other -- the traditional things that you do with the self-serve funnel. But we're also -- and this is, again, one of Paddy's big kind of commitments is, we're doubling down on our commitment to the developer community to provide interesting content and relevant content and kind of refreshing this great 40,000 set of documents, and tutorials and training videos that we have. So we're really doubling our efforts around engaging with the community. And the advent of AI/ML into that conversation gives us a great new channel to do that where we can do that same -- take that same model for AI/ML and provide that kind of content for our -- for the developer community. And we believe that will drive more people to our website, we'll drive more people to drive the platform. To get people to move up the curve from learners to builders to scalers, that's a big investment that as we deploy our customer success resources and our customer support is like making sure that we're intimately engaged with the larger customers and identifying what's keeping you from growing faster. And that then feeds back to why the product road map is what I described it to be is the feedback we're getting is, hey, we could grow with you, but we need a couple of things. And those couple of things are what's driving our product road map.

In the past, there's been some context for how to think about customer growth. Anything that you could add as far as like thinking about customer growth, whether it's overall or for any of those specific cohorts?

Yes. I think that -- and this is why we -- when I first arrived, we used to publish total customer counts and we -- we had a group that we call testers, which are people within their first 30 days or 3 months on the platform. some of them stick some of them don't -- the number of -- that number is de minimis. It has almost no revenue impact on the business, and it's volatile. And so -- and then you said, well, what about learners. I'm like, well, there's 450,000 of them. If there was another -- and they represent, and we do the math in my head, I think it's like 10% or 12% of total revenue, like not much. So you got another 1,000 of them or you got 10,000 of them. It doesn't move the needle on the revenue. So it's not a -- it's a super leading indicator, which is well, I'd like that number of learners to be stable. I'd like it to be growing. If it grows a little bit or it grows a lot, it's -- what's more important is how many are graduating out of that into the builders and the scalers because that's -- 86% of the revenue comes from those customers. And so to me, it's those 2 accounts are a more clear indication of near-term revenue progress than the learner count. Learner account is -- you want that to be healthy and growing, but it's not a quarterly metric, it's more of what are the trends and how are you building it slowly and steadily over time.

Let's talk about operational efficiency. You -- company was very quick to rapidly expand margins when some of the macro was turning. And now looking ahead into this year guidance is calling for, I think, 3 points of margin contraction. A lot of that is from the Paperspace acquisition and CapEx investments. But could you talk a little bit about that dynamic, like where we saw so much operating leverage and then where these investments are going that are kind of weighing on margins this year?

Yes. So as you said, we -- when we saw the market was -- growth was going to be challenged last year. We accelerated our longer-term margin targets and we're able to achieve that through a combination of kind of non-headcount just efficiencies and spend and then through some headcount reductions that we did in February. And we also shifted our cost model. So we have more than half the employees in the company now are international. We have a lot of resources in Pakistan, in Mexico and India and other places. So we've changed the long-term cost structure. We believe that there's still more efficiency that we can drive and particularly as we focus, we want to increase the amount of focus we put on our R&D, want to increase the throughput and the pace at which we're deploying new capabilities. We think we can do that by just shifting resources internally. But by virtue of getting to the higher margins, we were able to invest in Paperspace and grow that and only give up a couple of points of free cash flow while we're investing fairly heavily in both CapEx and OpEx around our AI/ML capability. So I think there's more room for us to drive leverage in the core business, but we're sitting in a pretty good spot where we've got the ability to invest in the AI/ML capabilities and see how that goes this year, and we can adjust our spend accordingly.

Wanted to ask about AI, but from the context of how you're using it internally, maybe from an efficiency standpoint. If you could -- if you have anything that you could add around how your R&D teams are leveraging generative AI or how your marketing teams...

Yes. So we've been using AI for a while. One of the bigger areas is in the fraud and abuse, like identifying based on patterns and use cases, et cetera, whether somebody is likely to pay or likely to abuse on the platforms. So we've been doing that for a long time. The customer success and customer support organizations have been using AI/ML capabilities to divert a reasonable chunk of calls and e-mails to be able to be served through that. I'd say on the marketing front, we're still early stage. That's an area where we're exploring. And on the developer side, the developers, certainly, we have the copilot kind of capabilities internally. And that's also an area where we're trying to figure out is could we be doing more than what we're doing today. So I'd say we're like many companies experimenting and we've had some good successes, but there's a lot of potential, we think, to leverage it that's not been tapped.

Great. And as we think beyond '24, where can margins go? Where will there be big sources of leverage? Like how should we think about margins beyond this investment year?

Yes. As I said, I think there's still the opportunity just to drive efficiency in the core business. There's data center footprint and kind of networking things that we can do that you do as a scaled company that will drive some gross margin improvement. So I feel good about that. I think there's still room on the OpEx front to drive efficiency, some through AI/ML as you're describing, some through just a very aggressive and high prioritization of what's getting the best return and being efficient with our spend. So I feel positive about that. The variable that, I'd say, least known at this point is, okay, we're spending -- we said $50 million in capital on AI/ML capacity in 2024. What's the right amount to spend in 2025? I don't know yet. And it depends on how we see the traction of the capabilities we're developing now. What I can tell you is, it would be an awesome problem to have to say, hey, this is growing so well and we've got such good economics that we want to spend some more to attract that growth, and we'll do that prudently and appropriately. As you've seen, we're not going out and build it and they will come and spending hundreds of million dollars on the AI space. We've made what we think is a prudent investment, consistent with our strategy, and you should expect that kind of decision-making from us as we go through this year as well.

Great. Want to shift to M&A just to hit out the overall strategy and philosophy there. You've definitely been acquisitive. And so when we think about build versus buy, is it fair to say that there's maybe a preference for buy to get to go to market faster? Like what's the philosophy? And does that change under Paddy's leadership?

I don't think the philosophy changes. I think there's not a preference to buy in an absolute sense that there's a -- there's a preference to accelerate the velocity in which we're advancing our platform and putting new capabilities out. If that means that if we can build it, there's a lot of advantages to building things and owning it and from economics and from a risk standpoint, you can partner. So you can partner and work with another technology provider to fill in a gap that you have, and there's certain different risk and economic profile there or if there's an opportunity to pull in M&A to -- like we did with Paperspace, where we didn't see a path where we could build it in a time frame that would be relevant, we'll take advantage of that. And the nice thing about DigitalOcean is we've got lots of flexibility. We had $412 million of cash at the end of last year. We said we're still very committed to driving free cash flow per share. So we'll have an ongoing share repurchase program, but it's modest relative to what it has been historically, $140 million over 2 years. That gives us a lot of flexibility to both invest in organic growth, should we want to, which we do. And also if there was an opportunity to accelerate our product road map, and we have some dry powder that can do that as well, all while we're still trending towards our long-term leverage targets, and I feel comfortable about that as well.

Got it. And you mentioned buybacks. You also have $1.5 billion of 0 coupon notes out there. I think due at the end of 2026 is still some time there before they're current, but any -- how should we think about your alternatives for addressing that debt maturity when down the road?

Yes. That's part of the reason why we -- as we communicated the buybacks being at the level that they are, is to give us flexibility around that and also be delevering to a point that we feel like we'll be in the 2.5 to 3x in a reasonable amount of time. If you just take the guidance that we provided, we'll be in the low-3s on a net leverage basis by the end of this year, just with the guidance that we provided. And so we're certainly paying attention to that. I'm highly confident we'll have a lot of different options of how we would refinance that. And again, when you're sitting on [ 400-and-something ] million of cash and you're earning 5% on it, and someone says, well, you could get a 7% yield if you bought back some of the convert. I'm like, okay, it's 2 points, and I lose all kinds of option value. I can do that, and then I don't have the cash. I can't invest it if our AI/ML capabilities take off or if there's an acquisition that we want to do. And so we're just -- we're being prudent and making sure that we have flexibility and -- but we're headed towards the right capital structure. So I feel very positive about. And again, the 0 coupon with maturity in just other 3 years, it's a pretty good instrument.

I'm going to ask one more and then see if anyone has any questions out there. I kind of want to lay out the bull case and just see your reaction or part of the bull case. I mean we -- like for some of the recent history, we've always been making adjustments around M&A and impact from pricing. And when reported growth rates were strong, like we were seeing decelerating trends. At this point, you do the same work and you see the complete opposite. Like that growth actually bottomed in Q2 by our calculations, around 3%, accelerated in Q3 to around 5%. And Q4 was like 9% growth. So pretty solid acceleration under backing everything out, making some assumptions. And then guidance for next year on the surface is, I think, 10%, 10.5%. But again, like doing some math, it's actually 9%. So like assuming no acceleration, even though we've been seeing this clear trend. It seems pretty conservative. I mean, would you agree or is there anything like fundamentally that gives you more caution about 2024?

I think it's -- I use the word appropriate. I don't think I would use the word conservative because you sit where we sit today. One, there's a lot of market, you don't know what's going to go on with the broader market. There's certainly supply chain and execution risk associated with the Paperspace business. I don't think there's demand risk. It's more -- can we get the gear in the time frames we think we're going to get it and then the quantities we think we're going to get it and can we turn it up in time. So because it's -- you're going from kind of effectively 0 to something meaningful, it's -- there's risk associated with that. So there's that. The other big kind of risk in the plan is still -- while we're seeing the things that you described and we're seeing improvements in the core DO business. In the fourth quarter, we're sitting at 96% NDR, which is a 4% growth headwind. And then if you said, well, you saw all those things that you described, and NDR rounded to 96% in the third quarter and rounded to 96% in the fourth quarter, this tells you it takes it a while for it to kind of get back to a level that's not a headwind. And so then the question is, is how quickly can you do it? And if it's all based on product releases and you're not assuming that just like the market gets better, it's appropriate to assume that it's going to take you to most of '24 to get there. So it's a headwind. And so am I very optimistic about '25 and how well we're going to be positioned and the growth rates will be very different than what we're talking about now, I am. But for 2024, I think it's appropriate.

Well said. Any questions in the audience? Okay. On competition, we haven't really talked about that today at all, want to get the update there. Maybe starting with the hyperscalers. Anything to note as far as their attention to your segment of the market? Or is it kind of the same story?

I think it's the same story. I mean, you'd say, okay, well, how often do you see them? And like, well, we see them as you would expect, if you do the pareto of our customer base, the largest customers are likely they're candidates, they could go to the hyperscalers. Conversely, hopefully, similar sized customers we could win over to our sites as well. That's where they become relevant. And then it's all about, well, what kind of relationship do we have and are they liking our service? Are there any gaps that we have? And we win some and we lose some in those spaces. But generally, them trying to come after our small space. I mean, think of all the money they're spending on AI/ML right now, like and think of like all the enterprise workloads that are still on-premise that could move to the cloud, like is the next best thing them to go after like little developers that will come on for $15 and maybe in 4 years might be $1,000 a month. I don't know. I'm not -- I don't know that I would sign up to take that job at that -- at those companies. So I don't -- we don't -- we haven't seen much of a change there.

And I mean, you mentioned win some and lose some, but when I talk to customers, often find that they're utilizing a multi-cloud strategy? Like is that still a case where you can have a really nice business with the customer who might use a hyperscaler for certain specific things?

That's a great point, Josh. When I say win some and lose some, I mean workloads, I don't mean customers. If you look at our -- and I think we shared this when we first disclosed the learners and builders and scalers model, churn goes way down as you get into the bigger bucket. What you risk when you get into the really big customers is you're chasing workloads. Like do you get all the new workloads? Do they keep all the workloads on you, you're playing for wallet share. That's the primary game when I'm talking about competition with hyperscalers.

And then just to round it out, hitting on the alternative cloud providers and maybe the more directly competitive. Anything -- any changes to note in that competitive landscape? And also just curious if Akamai's acquisition of Linode, do you think has made the environment easier for you?

That was the one thing I was going to bring up. I think the rest of the market is pretty consistent, and there's lots of folks trying to chase this space. It's still pretty fragmented. A couple of providers in Europe, a couple here. And -- but the one big change has been Akamai's acquisition of Linode, they focus more on the enterprise, and they say that, that's not me concluding that. We've observed it as well, but that's a pretty stated direction from them.

Okay, great. We hit quadruple zeros. Now, we're out of time. Thank you so much, Matt. Really appreciate it.

Thanks, Josh.