Thank you for here, for being here on day three of BofA's Global Tech Conference. I'm Wamsi Mohan, part of the tech team here at BofA. Delighted to welcome DigitalOcean to our conference. We have the CFO, Matt Steinfort. He's been the CFO since the beginning of the year.
Yeah.
Matt, thank you so much for being here today. Since you recently joined DigitalOcean, can you maybe just give a quick introduction of your background and what brought you to DigitalOcean?
Sure, Wamsi. Thanks for having me. I've been, as Wamsi said, I've been the CFO at DigitalOcean for just five months now. I joined in January. Prior to that, I was the CFO at Zayo Group, which is a large, at the time, private, communications infrastructure company. We're about $2.5 billion in revenue. I was the CFO for six years, three of which were public. We took the company private in 2020 right before the pandemic hit. At the time, it was the largest telecom take private. It was $14.5 billion, and I stayed the CFO for three years with that group before joining DigitalOcean.
Prior to that, I was the CEO and founder of a SaaS software company for 10 years. Prior to that, I was at a variety of corporate strategy and corporate development roles at Level 3 Communications, ICG Communications. I started my career at Bain & Company and Cambridge Technology Partners.
Excellent.
In terms of what attracted me to DigitalOcean, it's a phenomenal market opportunity. I mean, the cloud infrastructure is a, you know, kind of a core part of, I think, the internet going forward. It'll be relevant for, you know, decades to come. That was exciting. DigitalOcean has carved out a very interesting niche in that, you know, below the hyperscalers and above some of the lower, kind of lower-end, cloud providers. That was compelling to me. The economic model and the customer kind of acquisition model were very compelling to me.
I mean, a company, particularly coming out of telecom, where we were trying to grow 3%, 4%, 5% and spending $1 billion a year in capital and burning cash every year. Seeing a company that was able to grow in the 30s and even now the low 20s, but generate meaningful cash flow, was very attractive to me. Yancey and I had actually met at Zayo. He was the lead director working on the go private, so that's where I got exposure to Yancey. I have a tremendous amount of respect for him and for his team. It was a, to me, a no-brainer.
It's an exciting company and exciting space, and I feel really good about our ability to make an impact and improve the operations.
Excellent. Yeah, well, so glad that to have you over here, in this seat. What do you see as the TAM for DigitalOcean? When you think about the growth of the company, obviously, organic growth had been, you know, 30%-40%. you know, it's kind of decelerated some from there. How are you thinking about, A, the TAM, and B, the growth, as we think about the next few years?
Sure. The TAM is very large. It's a very fragmented market. IDC suggests that the SMB portion of cloud infrastructure, which SMB is defined, I think by them, by companies under 500 employees, it's like a $100 billion market, and it's projected to grow 20%-26%. If you spin that in terms of the quantity, there's 100 million SMBs in the world today, and that's growing at about 14 million a year. The market is enormous. But if you think about it, the challenge for a bigger company that's focused on enterprise, trying to serve that space and why DigitalOcean is so well positioned, is there's no one customer in that makes a difference, right?
Right
It's law of large numbers. You have to have a very low cost of acquisition, customer cost of acquisition. You have to provide the service in a much simpler way. We think that the market is enormous. Again, if you take our midpoint of our guide, we're less than a percent of that addressable market. We think there's a tremendous amount to be able to grow into, to capture share. Again, the market itself is projected to grow at around 25%-26%. We feel really good about that. I think our growth is a function of being able to take more of a wallet share of our customers as they grow and scale on our platform.
You know, the bulk of our revenue and our growth comes from our larger customers, comes from customers greater than $50 in spend. Capturing more of their wallet and being able to, you know, as they go multi-cloud, you know, over time, for diversity reasons and to be able to address more kind of deeper needs in specific technologies, we can pick some of that off, then, you know, we can definitely supercharge the growth from where it is today, which is clearly lower than it was, along with the rest of the industry, than it was last year.
You guys have given some pretty interesting disclosure around, how you think about your, you know, 600,000 customers, and you've segmented it across different spending buckets. Can you talk to the audience about, you know, A, like, how does that pipeline change over time? How does that cohort how have been the graduation rates of these cohorts and the spending trends? What are you doing proactively to drive that increased spend or graduation rates? Or how much of that is organic, versus how much of that is effort that you guys are putting behind it?
Yeah, that's a great question 'cause the... When I first started, I'd had, you know, a number of sessions with Yancey, talking about the business and talking about what the market's reacting to and what they're not reacting to. One of the issues was, you know, people were expressing agitation around the customer count, total customer count. "Hey, your customer count is up by 1,000, it's down by 1,000." That seems concerning. When we unpacked it, what we saw was, and I think we knew this, but we weren't doing a good enough job of explaining it, is, the lifecycle model at DigitalOcean is: People come to our platform to test an idea, to try an application, to try to build a business.
Sometimes they only spend $15 a month when they start. Cloudways, which we recently acquired last year, was our biggest customer, was doing $1 million a month when we acquired them. They started as a $12 customer. Customers come on, and they try, and they experiment, and they stay for a really long time. The average age is four years for a learner, which is someone who spends less than $50. Every month, every quarter, every year, some portion of those get success in their own business. Their app takes off, or their website gets successful, they build a business. They scale on our platform, and they become the builders and the learners.
The cutoffs are, you know, statistically important because we look at the data and the trends, but it's a little bit of a notional in that once you get to $50, you're running a business. You're not just like a hobbyist who's, you know, got a website and is, you know, paying like a Netflix kind of equivalent fee every month. You're a business. Once you hit that point. We typically, a graduation rate from learner to builder has been around, call it 2%-3%, or sorry, 1%-2% over the last five years. That's a, you know, when you're talking about a base of 460,000 of them, that's a decent amount that move to make, you know, a real business.
Sure.
When you get to the 500 level, that's when your product attach is really up, your churn goes way down. Those customers are incredibly sticky. They grow faster. That's where we get, you know, a tremendous amount of the value. What we're doing to try to play into and accelerate that, we're doing things at multiple ends of the funnel. If you think about, you know, we have 10 million customers, or I'm sorry, prospects a month that hit our website. Tens of thousands of them actually try the application and play around, and then thousands kind of convert to be customers.
In that process, if we can better identify who's a candidate to be a builder or a scaler faster, and we can devote more resource to them early on to guide them, you know, one part of the guide is just, you know, with the Cloudways acquisition, are you more likely to want a managed solution, or are you gonna wanna do it yourself? I mean, we weren't asking questions before. Literally, customers would come in and sign up, and then we'd look backwards and try to figure out what they were doing and based on their usage. Now we're asking questions and trying to funnel them into the right channels so we can accelerate that.
The other thing that we're doing is new motions for us from an outbound standpoint, using partners and direct sales to try to find builders and scalers that are already at that scale. They're sitting on Amazon or Microsoft or, you know, one of the hyperscalers, and see if we can attract their, you know, their business. The third thing that we're doing is more of an inside sales motion, which is we know which of our customers, particularly the large ones, are using other clouds. We can proactively reach out to them and say, "Well, what workloads did you move to them? What are we not getting? Therefore, what would we need to do in order to get that back or to have you move that workload to us?" That's informing our product development roadmap.
Very kind of holistic approach to, you know, kind of lean into that graduation, life cycle and see what we can do to accelerate it.
From a trend perspective, how many of your scalers or your highest-paying customers outgrow the platform, so to speak?
Very few, cause our churn overall for the company, churn for us means a customer leaves us, goes to zero. The churn has been relatively stable, you know, all of last year, it was a little bit elevated end of last year. It's come down a little bit in the last couple of months. That's a good positive sign. The churn for the scalers, the largest customers, is tiny. If you think the overall business is in the 10-ish, 11-ish percent range, the churn for the scalers is in the low single digits. That's a testament to customers grow on our platform, and they don't leave. What we do see is they go multi-cloud. We might lose a workload. It's not a workload we had, it's a workload they need that we don't get.
Mm.
Think of it as, they've got a really complex storage requirement. Our storage doesn't scale enough for them, they go to Amazon S3. That's what I was talking about before, is if we can identify those friction points, what is it about our platform that we need to advance or, you know, repackage the pricing in a way that makes it economic to you so that we win more of those workloads? Very few customers graduate off of our platform to a hyperscaler, but customers certainly, you know, they spread their workloads, and that's where we see a big opportunity.
You also mentioned sort of looking at customers who are currently at hyperscalers and attracting them to DigitalOcean. Like, how do you go about doing that? What is the value proposition that you are pitching to them? They already have an existing, sort of functional, you know, platform. What is it that you are approaching them with that's causing them to come? We've done survey work that actually has shown that customers are coming to DigitalOcean from hyperscalers.
Yeah.
What is it that you're doing to insight that?
Yeah. It's, I'd say the general value proposition between us and the hyperscalers, you know, regardless of whether they start on one side or the other, is we're a lot simpler to use, right? It's a, there's a complexity issue that they don't typically have. We're very simple to understand. The pricing model is very transparent. We have a higher level of direct support. If you're a small customer on an Amazon, that might be a big customer for us, you probably don't ever get to talk to anyone directly from them, and they're not giving you a lot of support. We provide support to all of our customers.
When you say, okay, that's maybe what wins you a customer who's starting from zero. W hat wins you a customer that's already there, it's economics in a lot of cases. What we've found is, when customers migrate from a hyperscaler to us, they typically save 50% on their bills. You know, we've seen that the other direction with Cloudways when, before we had acquired them. They were contemplating moving a workload from us to Amazon, and they did, and Aaqib, who's now our chief revenue officer, was talking to Yancey. He's like, o h, I did that, and then I got my first bill, and I, you know, it was way bigger than I thought it was. It was, you know, it was a mess, so they undid it.
Economics is a big factor, and that's where, again, we think that the niche we have below the, you know, the enterprise level is a very compelling value proposition for us, very differentiated.
We've seen actually at hyperscalers like Amazon and Lightsail before, and other, you know, products that have been out in the market, your growth rate was kind of unchanged through all that. There is something more about the value proposition than just giving a lower price point-
Yeah
to your point.
Absolutely. I mean, think about the customer acquisition model we discussed. We take customers at, like, $10-$15 a month and take 18-24 months to grow them to $50 or above, or. I mean, you gotta be really patient for that. You gotta provide a lot of service. It's gotta be super easy. Think of the market motions, the mechanics, the cost structure of those bigger companies. They've got a bigger addressable market. I get it. They've got, you know, bigger, you know, bigger potential with some of the enterprise customers. That's a different motion. It's very, it's unusual to find companies who can do both.
You know, be like , I'm gonna be a hardcore enterprise-oriented, i t's all about customization and complexity and getting embedded, or I'm gonna be one that's like, you know, deals with 10 million visitors, you know, a month, and trying to convert thousands and hope that you just put them in your pool of learners, and then that learner pool will grow over time. That's a, it's a very different motion, very different economics.
It's not intuitive to a lot of people who think about hyperscalers who kind of have this notion of, you know, unlimited CapEx in some ways, right? This is a capital-intensive business in some ways. Can you talk about the returns on the business? I think that when you look at your business model through a lens of returns, it's actually very attractive.
Yeah. No, that was the other thing that I talked about, was one of the big draws for me in coming to the company. I mean, a company growing in the, you know, 20%, you know, plus or minus range, that's generating 30% free cash flows, which, you know, we said we'll approach that by the end of this, by the end of this year. That's a huge growth in free cash flow, right? That's a free cash flow CAGR over a couple of years of, like, 40%-50%. You combine that with the share repurchases we're doing to bring down the, you know, denominator. It's a very compelling free cash flow per share story.
You know, if you think of the margin kind of stack, gross margins, we have a lot of ability to drive improvement there. We've, you know, as we grow into some of the recent investments we've made in the Sydney data center, some additional kind of networking capacity. On the EBITDA front, we've got the ability to drive pretty good margins there. We did a cost reduction initiative earlier this year, where we announced about $60 million of run rate, kind of, expense coming out. $25 million was headcount, $35 million was non-headcount.
You know, that's gonna drive EBITDA margins close to 40% by the end of this year, and we'll take that 16% reported free cash flow margin in the first quarter to, you know, close to 30% by the end of the year. It's very, very compelling from that standpoint. CapEx, we think we can keep around the 15% of revenue, which is where it was in the first quarter. We think that's a sustainable level and still enable us to invest in maintenance, to invest in the growth capital, to match the scaling of our business, and also invest in new kind of capabilities and new footprint.
How do you go about going from 16% free cash flow margins to over 30% exiting the year?
Yeah, that's the, you know, take the EBITDA from the, was it 34%, I think, in the first quarter to close to 40%. That's, you know, primarily on the back of two things. One, the gross margin will increase from the 56% to 60% as we grow into those investments we made. The cost reduction initiatives that I just talked about, that we did that mid-February, and some of it was immediate. Like, about half of the headcount reduction we did right away. The other half is a move and a migration of resources and roles from the U.S. to lower-cost global capability centers.
We had about 120 roles that we shifted from the U.S. to Pakistan, India and Mexico, and that's a long-term structural change. That's when we add developer resources, you know, we're looking at where do we add them? U.S.? Do we add them in some of these other locations where we're getting really strong talent, very good productivity, and a fraction of the cost. You know, you take all of those together, you keep CapEx at kind of around 15%, and you get to that 30% exit, kind of, run rate on the free cash flow margin.
What about on the gross margin side? You mentioned, obviously, you opened the Sydney data center. You can sort of scale into that, leverage that. What are some of the other factors that get you there, and is there room to improve from procurement, and what are some of the initiatives there that you're doing?
Yeah. If you go back to when Yancey joined the company, the company was spending 44% of revenue on CapEx, and they reduced it to in the 20s really quickly. You know, how'd they do that? Well, they didn't have the company was founded by developers. It was really successful. They were growing like crazy. Didn't have a lot of gray hair, you know, financial discipline. Like, "Hey, you might want to get contracts with some of your larger vendors. You might want to do basic procurement." They were able to just, through some basic kind of discipline functions, we were able to get that down. We're still doing that, though. I mean, we just started signing up longer term contracts with the colo providers last year, so that's kind of a new thing.
We have opportunities, tremendous opportunities on the bandwidth side. That's something, again, near and dear to my heart from my background. Just kind of in an environment where you know that the market is not growing as fast as it was, and where you have a commitment to a free cash flow margin target, which we accelerated our long-term target, you know, to this year, you have to make, I'd say, more disciplined and you have a higher bar on investment. You say, "Okay, well, you know, we're not gonna spend as much capital on this if it's not gonna earn the return we want," or search engine, paid marketing, paid third-party placement. We weren't getting a great return on that, we dialed the spend back.
I think there's a tremendous amount of opportunity. I think you're seeing this. It's perversely, it's part of the reason there's a slowdown in the overall market, is a lot of companies are doing this. Like, "Uh-oh, if we're not growing as fast, we've got to get to free cash flow faster. How do we do that?" We look at all our spends and we see, man, we're spending, we have 10 different HR applications, and we have 12 different marketing automation things. These were great ideas, and when everything was growing, super fast, it was okay.
Now in a different environment, everybody's got to be kind of, I think, more prudent in dialing in their spend, which is causing near-term headwinds, but, you know, will eventually, run its course, and we'll be back to growth in that, in that area.
Yeah. Well, that's helpful. Matt, your aggregate growth loads have slowed down, despite the Cloudways deal. Do you think revenue growth in the next few years will be more driven by ARPU versus increased customer growth? Or how do you think about the balance between those?
Yeah. I think both will certainly contribute. I do think ARPU growth is a bigger lever for us to pull, 'cause the self-serve funnel, the engine of the company, it's super predictable, it's reliable, and it's been, you know, relatively steady. You're always adding customers into the beginning of the funnel that are learners, and then, you know, again, that pool, it can fluctuate. As long as it's big enough to generate enough kind of output, you'll see growth through the customers. We'll also add customers through those direct motions, you know, trying to bring in customers later in their life cycle. That'll all, I think, be very positive. If that's all we did, I think we're, you know, we're a high single digits kind of grower.
But where you get the, where you get the lift, and you get into the more attractive growth rates, is you figure out how to increase the ARPU. Not just the ARPU, because people are kind of scaling their own business, but figuring out how do you get more share of their wallet as they become bigger businesses. Enhancing our storage capabilities to be more fulsome, adding some basic capabilities that, like role-based access control and other things that are necessary as you become a, more than a 5-person team, you know, that's using our platform, you need a bit more flexibility in that.
Those are the kind of things that we're looking at, and we're looking at other applications, you know, the, AI and other applications that customers might be demanding. We certainly have a long, kind of, thoughtful product roadmap about how to integrate some of those capabilities. I think, yep, growth will come from both, but I think ARPU is probably the bigger driver in the near term.
When you think about that ARPU, and I know that, you know, you've changed you've, there's different ways to think about what you're offering, but your Premium Droplets, for instance, products that can help you move up. You have richer CPU, GPU configurations, memory configurations. How do you look at your customer base and say, which are the workloads, and where is the probability of actually looking at the data, looking at the usage, and saying, "How do we recommend these to them?" How long has that process been underway, and how much more room is there to go on that?
I'd say there's a tremendous amount of room to grow there. I'll give you an example. The premium dedicated Droplet that we launched in the first quarter was a bandwidth-optimized Droplet. You know, just think if somebody uses our application, they're a gamer, they're a streamer, they need more bandwidth, and previously, they had to do two things. One, they had to buy a bunch of Droplets to get up the amount of bandwidth they wanted. They didn't need more compute, they didn't need more storage-
Mm-hmm.
Because of the way we packaged the product, they had to kind of scale those things linearly, okay? The second thing was they weren't optimized together, so that because they had to write software on top of our application to use these little building blocks instead of having something that was more easily consumable by them. We got enough feedback from customers that there was a big demand there, that we said, "Hey, well, we'll do this." What we were able to do was, we were able to launch a service or a SKU, that we're able to charge them two to three times what we would've charged them before. They were happy because it reduced enough of the cost that they were having to do on their side, that it offset it.
When I see things like that's a tremendous opportunity, 'cause there's so many different verticals or so many different use cases on our platform, you could think of a lot of different optimized SKUs or even just unbundle your components a bit, so people can configure them more easily themselves. That should drive a tremendous amount of growth potential. I think we're actually even though the company's been around for what? 10+ years, we're relatively early in that because the company was founded, and a lot of the original developers and a lot of the original team, the persona they were solving for was the developer. Like, don't do anything to mess up the developer experience. Make it super simple, you know, one product, one SKU, one price.
That's what the kind of development engine was geared around. We're you know, we're in the middle of a transformation. We're saying, "Look, that's not where the value is. That's interesting. We got 470,000 of them. We need to make it an attractive place for them to come and come into the business and take care of them so they can grow." All the growth is gonna come from the bigger guys, who have a different need set that's slightly more, you know, nuanced than the developer. I think we're very early in that game.
Can you talk a little bit, last year, middle of the year, you did a price increase. As you think about your guide for this year, is there contemplated in there another price increase or not?
We increased Cloudways' pricing by about 10%, started in April. That's the only, like, list price, kind of, price increase in there. The rest is more monetization, like we're talking about, re-bundling, repackaging the products.
Within your guide, we are not anticipating any other price increases.
No.
That would be incremental if they were...
Yeah, if we were to do.
Offset.
Yeah.
Maybe even that.
It would offset risk, or it would. The only price increase that was complicated or contemplated was the Cloudways price increase.
Can you talk a little bit about the churn after the price increase? I mean, our survey work showed that there was very minimal churn, but curious to get, like, what data you guys have seen around churn metrics after the price increase.
Yeah, churn increased a little bit in kind of late fourth quarter. I don't believe it was attributed to the price increase. It was, I think, more attributed to weakness in the overall you know, market, we've seen that come back down. We didn't see a meaningful increase in churn at all with Cloudways, so when we did that price increase. Churn has been, if you know, in terms of green shoots, the churn performance has been pretty good.
Yeah.
Most of our pressure is on customers themselves are just not growing as fast, and so they're not growing as fast on our platform, so that's lower expansion. We are still seeing optimizations on our platform by bigger customers, so the contraction is still a headwind.
You briefly mentioned AI in your comments earlier. I'm curious, which customer cohort or like, you know, within your builders, scalers, learners, where do you see that the most, or where do you see the most opportunity?
Yeah, it's interesting because the needs of, for AI of smaller businesses is certainly different than some of the large kind of headline names you'll read about in the, in the press, and folks you're probably talking to at conferences like this. Our customers don't go to these kinds of conferences, and you would never probably know of them. We have gaming customers today that are using AI on our platform, you know, $250K kind of ARR customers that are using our existing capabilities. They're not using or even asking us for GPUs. They can do it with the high kind of performance compute we have. We're seeing the application, you know, and interest. We're not seeing a lot of people asking us for GPUs specifically within our customer base.
Mm.
We have GPUs in our network. We're testing it. We're trying to figure out what the value prop is and how do we productize that for a smaller kind of set of customers? That's something that's on our roadmap, and we don't have anything to say right now, but hopefully we'll be able to say something soon about what our plans are there.
Okay, that's great. I also want to talk to you about capital return. You guys have announced a fairly large capital, I mean, a buyback plan, and you've been executing against that. What are your thoughts around sort of how aggressive you want to be? How do you think about being programmatic versus opportunistic in terms of buybacks?
Yeah. We got an authorization for $500 million of buybacks this year. We broke that into two different tranches. We had a $270 million, what I would consider an opportunistic program, was still executed through a 10b, but where, you know, depending on where the price sat, we would be aggressive or more or less aggressive. We have another component, which is the $230 million, which is the we're gonna be in the market every day, and we're gonna buy some stock. We might buy a little bit more or a little bit less, but we're gonna be in the market every day. We've been executing on both of those.
We're probably about $350 million in, we have got $150 million left for the balance of the year. We think it's an important part of the overall capital return profile. Next year, we've said we'll do, you know, up to 125% of free cash flow. The intent is, whether it's 100% or it's 125%, is we intend, as part of our formula, to drive operating leverage to improve the free cash flow margins, we intend to reduce the share count. When you do that together, we think the outcome from a free cash flow per share perspective is pretty positive.
How do you think about the liquidity of the stock in light of that?
The, you know, from a float standpoint?
Yes.
Yeah. We're not concerned about the float at this point, although if, well, there's a short interest that might be concerned about the float, particularly. We feel that we are well positioned, given the flexibility we have with the cash flow generation we have, to manage, you know, to a reasonable leverage level and to be able to invest in M&A, to be able to invest in organic growth, and still return, you know, a fair amount back to our shareholders.
Yeah. Since we're almost out of time, Matt, I just would like to give you the opportunity to talk to the investor base and talk about why DigitalOcean is a compelling investment.
Yeah. I think it comes back to the two things. The free cash flow profile, I think, is unique in the space. We talk about Rule of 40 or Rule of 50, you're talking about a lot of companies that are saying, "Hey, I'm gonna cut costs now to try to accelerate free cash flow generation by a year or two," and we're already there. We're already generating very compelling returns. The other thing that I think that people has underappreciated is the durability of both the SMB market as a whole, which is 40%-50% of GDP, depending on kind of where you're looking. In some countries, it's even higher. It's very durable, right? That's a market that's been around.
It'll continue to be around, it'll continue to drive a lot of growth. If you look at the resilience of our customer acquisition, we can acquire customers at, I would say, effectively zero, but 10% of revenue on sales and marketing is incredibly low. It's incredibly efficient, and once we get them, again, we keep them for a very long time, and they grow. So it's like a free option, where customers are paying you to be prospects, and then, you know, some of them kick off and generate, you know, meaningful, attractive, long-term customer value. That, I don't know that is fully appreciated by the market at this point.
Yeah, absolutely. No, very compelling story with a solid free cash flow profile, and you guys seem to be doing a phenomenal job. Thank you so much for being here, Matt.
Thanks, Swati.
Thank you.