Title

Presentation

Operator: Thank you for standing by and welcome to the DigitalOcean First Quarter 2024 earnings conference call. All lines have been placed on mute to prevent any background noise. After the speakers remarks, there will be a question-and-answer session. [Operator Instructions] I would now like to turn the call over to Rob Bradley, Vice President of Investor Relations. Please go ahead.

Rob Bradley: Thank you, Rochelle. And good morning. Thank you all for joining us to review DigitalOcean’s first quarter 2024 financial results. Joining me today are Paddy Srinivasan, our Chief Executive Officer, and Matt Steinfort, our Chief Financial Officer. After our prepared remarks, we will open the call to a question-and-answer session. Before we begin, let me remind you that certain statements made on the call today may be considered forward-looking statements which reflect management’s best judgment based on currently available information. I refer specifically to the discussion of our expectations and beliefs regarding our financial outlook for the second quarter and full year 2024. Our actual results may differ materially from those projected in these forward-looking statements and direct your attention to the risk factors contained in the company’s 10-Q filed with the Securities and Exchange Commission and those referenced in today’s press release that is posted on our website.

DigitalOcean expressly disclaims any obligation or undertaking to release publicly any updates or revisions to any forward-looking statements made today. Additionally, non-GAAP financial measures will be discussed on this conference call and reconciliations to the most directly comparable GAAP financial measures are also available in today’s press release as well as in our updated investor presentation that outlines the financial discussion in today’s call. A webcast of today’s call is also available on our website in the IR section. With that, I’d like to turn the call over to our CEO, Paddy Srinivasan. Paddy?

Paddy Srinivasan: Thank you, Rob. Good morning, everyone. And thank you for joining us today as we review our first quarter results. After my first three months in the role, I’m pleased with both the solid execution and durable growth we delivered in Q1 and the early progress we are making to position the company to take advantage of the material growth opportunities that are in front of us. In my remarks today, I will briefly highlight our first quarter results, share some initial observations from my first 90 days, provide several examples of our increasing product velocity, and discuss our progress pursuing the tremendous AI growth opportunity. Before I get too deep into my remarks, I want to highlight our Q1 performance, which was solid across the board.

Revenue growth accelerated quarter-over-quarter, and we continue to deliver strong adjusted EBITDA and free cash flow margins while increasing investments in our higher growth businesses, demonstrating the strength of our business model. We are also encouraged by the improving growth fundamentals. Net dollar retention continued to slowly rise from our low last summer, increasing as expected to 97%. Our core product usage grew faster in Q1 than it did in Q4 of 2023. And we’re also seeing strong uptake of our still early stage AI and machine learning platforms. While we still have a lot of work ahead, our Q1 results are very encouraging. Matt will walk you through more of the details later in this call. I will start my deeper commentary with some initial observations after my first quarter as CEO.

I’ve spent a meaningful portion of my first 90 days with customers, partners, and employees. The feedback and insights I have received have only increased the excitement and optimism I have for DigitalOcean and our growth potential. More than anything, I’ve been thrilled with the positive and constructive feedback that I’ve gotten from our customers. We have an incredibly loyal customer base that relies on DigitalOcean to run their businesses, that want to do more with us as they grow, and they also have very clear feedback on how we can help them accelerate. Most of the builders and scalers on our platform run revenue generating software products on DigitalOcean and have come to know and love us for our simplicity, our valuable technical content, and our compelling price to value.

I’ve been actively engaging and listening to them, and they have helped validate some of the hypothesis that I have long held about this solution. Number one, that the market opportunity for cloud platforms is large and growing, and is only increasing with the advances of AI and machine learning technologies. Our customers are optimistic about their own long term growth prospects and are telling us that they see opportunities to expand their business with DigitalOcean. Our platform matches what growing technology businesses require, a scalable and performant platform, a platform which is simple to get started on and scales with them, a platform that is cost effective and more importantly provides transparency of ROI with robust technical support both directly from us and also from our passionate community of developers.

My conversations with dozens of customers offered key insights into the gaps in our platform and highlighted the emerging needs of our core target customer. This reinforces our mission, and we are going to continue to focus on product innovation and ensuring we delight our customers and the developer ecosystem. Their endorsement of DigitalOcean is powerful, and we will work tirelessly to earn and retain their business while attracting net new customers. With this backdrop, let me give you an update on what we’ve been working on recently. Over the past few weeks, we’ve made demonstrable progress accelerating our product innovation and the velocity of our new releases. Let me share a few highlights with you. First, we recently introduced turnkey data protection for our customers by launching daily Droplet backups.

In true DigitalOcean fashion, to make it super simple for developers, this capability enables one click Droplet data protection, providing peace of mind from accidental deletion through automatic retention of the seven most recent copies. In parallel, we also improved the speed of our snapshots capability by up to 6x, enabling customers to back-up even larger droplets much faster than before. While it is still early days, we are seeing robust adoption from both existing and new customers, enabling daily backups. We have already seen more than 1,300 new customers enabling daily backups and 150% month over month increase in overall droplets being backed up daily just between March and April of this year. We also rolled out new additions to our premium Droplet offerings, expanding these premium options to memory and storage optimized families.

With high-performance, non-volatile memory express, SSDs, and 5x increased networking throughput over regular droplets alongside our flexible egress bandwidth allowance, these new memory and storage optimized droplets are ideal for memory, data, network, and bandwidth intensive workloads. We are very excited for the potential they will offer to new and existing customers across a variety of use cases like caching, databases and many others as they get ramped up. Also in Q1, we introduced horizontal scaling for Managed Kafka, continuing our focus on making the complex simple for our customers. Horizontal scaling for Kafka is particularly critical for our customers who manage large volumes of streaming data and want to prioritize scaling bandwidth and highly performant end user experiences.

This facilitates right provisioning of nodes in support of fluctuating workload requirements, enabling customers to handle spiky data volumes and traffic, improve the reliability of their clusters and optimize their resources. A few other notable releases we are excited to highlight since our last call include a series of app platform improvements such as CPU-based auto-scaling and dedicated egress IP support from our platform. Turning into our managed hosting cloud-based offering, in January, we were proud to launch cloud-based Autonomous. With a few months in market now, the initial feedback from customers and community has been very positive with over 650 customers adopting the new capability today. Alongside Autonomous, we shipped a number of other crucial items to simplify and secure our managed hosting offering for our customers.

These include DNS Made Easy to simplify DNS management for users, an integration with Patchstack to provide an extra layer of vulnerability detection and alerting, and most recently in April, client billing, the first tool launched for our planned Agency Suite that will automate and streamline various agency workflows to enable simpler, more efficient, and more agile operational support to help our agency customers grow on the DigitalOcean platform. More to come on this throughout the rest of this year. These examples are just a few highlights from the list of capabilities we continue adding in our mission to simplify the cloud for our customers. We will continue to listen closely to our customers and strive to accelerate our delivery velocity to ensure our customers are positioned for success as they grow on our platform.

We know that when our customers win, we all win. And our focus remains squarely on delivering a rapid cadence of new releases aimed at delighting our customers, enabling them to scale their businesses, and ultimately increasing our net dollar retention. Before turning it over to Matt, I would like to share some updates on what we’re seeing across the AI landscape. It is an exciting time in AI, and you can see that every day in the headlines that you read. Companies across virtually every vertical that you can imagine are eager to incorporate AI into their value proposition. While large language models or LLMs get most of the headlines, we’re learning that our target customers, many of which are software vendors, are looking to consume a variety of different AI models into their offerings, like fraud detection, sentiment analysis, natural language processing, live translation, demand forecasting, and of course, LLM based models like image and video generation, coding assistance, Q&A bots, and many more.

We’re seeing strong revenue growth for our early stage AI solutions as we continue to ramp up our initial GPU capacity through the first part of this year. And our expectations are that demand will continue to outstrip supply for the foreseeable future. As of March, our ARR grew to $19 million, most of which is our platform-as-a-service offering, a 128% annualized increase from December 2023, driven by demand for both AI model training and consumption of models, also known as inferencing. In addition to our existing AI platform-as-a-service that helps AI and machine learning developers consume a variety of open source models, we also launched our GPU based infrastructure as a service offering in January of 2024, and are seeing strong traction with GPU hours sold and consumed increasing 67% just from March to April of this year.

The growing customer base for our AI infrastructure-as-a-service offering are both venture backed startups as well as established businesses. Over the last four weeks, we have onboarded several customers that came to us for our availability, simplicity, and support. Customers are using our AI platform-as-a-service and infrastructure-as-a-service platform for a variety of use cases, including text and video generation, AI coding co-pilots, recommendation algorithms, model hosting services, and many more. Let me give you a couple of concrete examples. First is a venture-backed customer that is a leading AI-driven storytelling platform that helps build marketing storyboards, visual manuals for complex products, and comics, all from textual prompts.

Another customer example is an AI assistant tool that developers leverage to code with greater speed, flexibility, and accuracy. These are just two new examples of customers that we have seen this year as the AI market continues to quickly grow and evolve. I’ve spent a significant portion of my time and attention in my first 90 days working directly with these customers to understand their needs deeply and translate that into our longer term AI strategy. Our customer needs, while similar, are quite distinct from the needs of large enterprise customers using hyperscalers or large language model builders who use the raw infrastructure from GPU form providers. To put this in perspective, over a decade ago, DigitalOcean identified and delivered an innovative compute solution with a clear product market set that was not being effectively addressed by the larger cloud providers, creating an easy on-ramp for developers, startups, and entrepreneurs to learn, test, and scale their businesses by simplifying cloud computing.

We see a similar opportunity emerging now to democratize the access to AI and machine learning capabilities, not only by providing simple access to GPU capacity via infrastructure as a service, but also by integrating AI and machine learning into the developer experience itself to transform how developers build and run their workloads on our platform. Like many cloud platform providers, we are simultaneously turning up incremental capacity to meet near-term demand, while also rapidly learning and evolving our AI strategy. While we are certainly in the very early innings of this transformative growth opportunity, we continue to believe that software, more than hardware, will be DigitalOcean’s long-term differentiator and competitive advantage, especially for our target customers.

We are confident in the strategic direction we are taking and believe that AI will be a meaningful growth contributor in 2024 and in the years ahead. We will make the right choices on investment this year as we continue to see positive results. We will share more on our plans on this front over the course of this year. To close my comments, I’m pleased with our performance in the early months of 2024, and I’m optimistic on our near and long-term growth potential. We have a very solid performance and growing core business. Our revenue growth accelerated quarter-over-quarter, NDR improved, and profitability and cash flow margins were all very healthy. We are accelerating our pace of innovation and delivering new capabilities in rapid cadence which will help our customers grow on our platform, thereby increasing our net dollar attention.

Our AIML solutions are resonating very strongly with our customers and we are working to turn up incremental capacity over the balance of the year to keep up with robust demand. There’s a lot of work to do to take full advantage of our opportunity, but we are moving in the right direction and continue to make steady, rapid, and respectable progress each quarter. I will now turn the call over to Matt to provide details on our financial results and on our outlook for Q2 and for the balance of the year. Matt?

Matt Steinfort: Thanks, Paddy. Good morning, everyone, and thanks for joining us today to review our first quarter results. Revenue growth continued to improve quarter over quarter. We are seeing positive signals from our key growth drivers and we continue to deliver attractive adjusted EBITDA and free cash flow margins while we increased investment in our AI platform to pursue this material growth opportunity. This morning, I will provide a deeper look into the first quarter results before providing our financial outlook for Q2 and for the full year. Revenue in the first quarter was $184.7 million, which was up 12% year-over-year and up sequentially from the fourth quarter of 2023. We added $19 million of annual recurring revenue, or ARR, in the quarter, which was the largest organic quarterly increase we generated since the second quarter of 2022 and was 82% higher than the incremental ARR we generated in Q1 2023.

Contributing to this growth was steady incremental revenue from new customers, improving net dollar retention from our existing installed base, and healthy contributions from our managed hosting and AI platforms. Revenue from new customers in their first 12 months remained both steady and a key element of our solid growth foundation in Q1. As we discussed in February, improving our net dollar retention or NDR is a major focus area and a key driver for accelerating our overall growth rate as we move through this year. As anticipated, NDR improved to 97% in Q1 and has continued to steadily increase since we reached our low point in July of [2023] (ph). We continue to see steady and historically consistent levels of churn and we are seeing modest month-over-month improvements in our net expansion which is expansion net of contraction as our customers are slowly returning to growth.

The work that we are doing on both accelerating our product roadmap and continuing to enhance our customer success motions should enable us to continue to increase our net dollar retention as we seek to remove it as a headwind and eventually return it to a tailwind to our overall growth. Our managed hosting product, Cloudways, another key growth driver, contributed revenue of $22 million in the first quarter and grew 34% year-over-year. While we will see a slower year-over-year growth rate from Cloudways as we lap the managed hosting price increase that we made in April of last year, we do anticipate managed hosting continuing to be one of our faster growing platforms for the foreseeable future. As Paddy described, we are still in the early innings with our AI and machine learning solutions.

But the rate of growth of both the leading indicators and our revenue on this new platform is already meaningful. Our AI and machine learning platform contributed $4 million in revenue in Q1, and we exited Q1 with AI ARR of more than $19 million, a 128% annualized increase. This strong growth came despite our being capacity constrained in the majority of the first quarter as we navigated supply chain challenges and turned up the first wave of servers that we had ordered in Q4 of 2023. We are consistently selling through our available capacity as it comes online and we saw our hours sold on each 100s increase 67% in April just over March in just a single month. We will continue to add the next waves of our planned incremental capacity over the balance of the year and anticipate that demand for our AI solutions will continue to be robust.

Turning to the P&L, gross margin was 61%, which was an increase from 56% in the first quarter of the prior year. The largest factors in this 500 basis point improvement were the success of our ongoing cost optimization efforts and/or having grown into infrastructure investments from prior periods. As is the nature of our business, incremental investments in equipment, space and power, and networking cause modest step function increases to cost of goods that are then smoothed as we fill the capacity with incremental revenue. Given our planned AI investments, we anticipate that gross margin will moderate somewhat in the coming quarters. Adjusted EBITDA margin was 40% in the first quarter, in line with the prior quarter, as we continue to diligently manage expenses.

Our healthy profitability in our core platform continues to provide us the flexibility to make additional investments in R&D to accelerate our product roadmap and to invest in our higher growth opportunities such as AI. Diluted net income per share was $0.15 and non-GAAP diluted net income per share was $0.43. GAAP and non-GAAP diluted earnings per share increased by $0.32 and $0.15 respectively on a year-over-year basis. While we have been cash flow positive since 2021, it is notable that we are now posting positive net income quarters on a GAAP basis, which is a further indication of the profitability of our core DigitalOcean business. Adjusted free cash flow margin was $34 million, or 19% of revenue, which was an improvement from 16% of revenue in Q1 of 2023.

As we have said previously, free cash flow margin is a more meaningful metric on an annual or trailing 12-month basis. And quarterly free cash flow margin will vary given the timing of capital spend and other working capital impacts. Turning to our customer metrics, average revenue per customer increased 8% year-over-year to $95.13. The number of builders and scalers on our platform, those that spend more than $50 per month, was 158,000, an increase of 8% year-over-year. Their revenue growth year-over-year was 13%, ahead of our overall growth rate of 12%. The number of builders and scalers on our platform, which represent 87% of our revenue, increased by 1,300 during the quarter. The increase in our higher spend and higher growth customers is a result of our focus and concentration of our marketing, product development, and customer success investments on these builders and scalers.

Along with the increase in our higher value customers, we did see total customer count decline by 7,400 quarter-over-quarter. This change was due to a reduction of 8,700 of our lowest spending customers, our learners, with that reduction collectively representing only around $100,000 a month of recurring revenue as the average spend for those customers was less than $10. Our balance sheet remains very strong as we ended the quarter with $419 million of cash and cash equivalents. During the first quarter, we leveraged our material cash balance and free cash flow to repurchase 200,000 shares of common stock for $8 million as part of our ongoing share buyback. Looking forward and building on our steady growth in Q1, we expect Q2 revenue to be in the range of $188 million to $189 million, representing 11% year-over-year growth at the midpoint of our guidance range.

For the second quarter, we expect adjusted EBITDA margins to be in the range of 37% to 38% and non-GAAP diluted earnings per share to be $0.38 to $0.40 based on approximately 102 million to 103 million in weighted average fully diluted shares outstanding. With improving NDR and the strong demand for our AI platform that we saw in Q1, we are increasing the bottom end of our full year revenue guide by $5 million, projecting revenue to be in the range of $760 million to $775 million for the year, a $2.5 million increase in the midpoint of our guidance range and representing year-over-year growth of 10% to 12% for the range. On the profitability side, we continue to drive operating leverage in our core DigitalOcean platform, enabling us to increase investment in our faster growing managed hosting and AI and machine learning platforms, while maintaining attractive overall margins.

We continue to execute the plan we articulated in February and continue to project our adjusted EBITDA margin for the full year to be in the range of 36% to 38%. We also maintain our forecast range for full-year adjusted free cash flow margin at this point. As we continue to see positive signals from our AI solutions over the balance this year, we will continuously assess whether to deploy additional capital to further accelerate our AI growth, which may result in reductions to our free cash flow margins to support that. We are also maintaining our non-GAAP diluted earnings per share guidance, which we expect to be in the range of $1.60 to $1.67. That concludes our prepared remarks and we’ll now open the call to Q&A.

Operator: Thank you. We will now begin the question-and-answer session. [Operator Instructions] Your first question comes from the line of Raimo Lenschow of Barclays. Your line is open.

Raimo Lenschow: Thank you. Congrats from me for a great quarter. First question on AI, if you think about the state or the evolution of AI adoption, there’s like training and experience, inference, sorry, and the — what are you seeing at the moment on the platform side, and how do you think that will kind of evolve for you? And then another question on NRR, if you think about it, it’s obviously a lagging indicator. Like, how do you think the weaker periods coming out, kind of impacting kind of NRR going forward? Thank you.

Paddy Srinivasan: Okay, great. Thank you for the question, Raimo. I’ll start with the AIML question and then I’ll let Matt chime in with the NRR question. So from an AIML perspective, as I said, it is very exciting traction, but I have to remind that we are in very early innings, not just at DigitalOcean, but as a market as a whole. So I think for us, we are super focused on the needs of our customers. And our customers are a little different, as I described in my prepared remarks. A lot of our customers are tech businesses that build and run software applications on the DO platform. And they need a variety of different AI models, not just LLMs. And they are mostly AI extenders. So if you take the example of LLMs that a lot of people are familiar with these days, you can inject data and fine tune and extend current AI models.

So that’s a lot of what our customers are looking to do, not just in LLMs, but in other models as well. And typically our customers are AI extenders and AI consumers. And our AI value proposition is twofold, as I explained. We have recently introduced the infrastructure-as-a-service. And the dominant use case right now is model building and model extension and fine tuning, and also model inferencing across different parts of our data center infrastructure. We also have a very robust platform-as-a-service offering with Gradient, a platform we acquired through Paperspace, which is also undergoing considerable enhancements as we speak. So this platform-as-a-service has a much wider aperture in the sense that it deals with AI and machine learning throughout the life cycle of software development.

So we have both that as well as the raw infrastructure-as-a-service to fine tune and build and train and infer these AI models. So as I said, we are very happy with the progress in Q1. We will be focused on serving our customer segment primarily with our AI strategy. And we feel very confident that we are now starting to really understand the evolving needs of our customers and what they’re actually looking for, both in model training and inferencing. And as I said, these are very early innings. A lot of attention is now on model building and fine tuning and training. But the long term use case is going to be super heavy on inferencing and we have both those phases covered with our platform-as-a-service and infrastructure-as-a-service.

Matt Steinfort: And, Raimo, on the net dollar retention, again, we’re encouraged by the, I’d say the steady — slow but steady increase in net dollar retention that we’ve seen, basically on a monthly basis from July of last year. And as we’ve said historically, the churn hasn’t really been the challenge for us over that period and even in the prior year. The issue had been net expansion, which is expansion minus contraction. And we’ve seen slow and steady improvement there, which is driving the improvement in NDR. Contraction continues to kind of get better, a little bit better every month. Expansion, again, was the last of our kind of drivers of NDR to hit the bottom last year. And it is holding steady at the levels it’s been over the last several months and we see positive indications.

But it’s, I’d say, going to be slow and steady growth for us to get that NDR up, and we’re encouraged by the progress that we’re making. And the product development work that Paddy described, and the heightened focus on the customer success, those all contribute to the improvements and we’re banking on only the things that we control, so those improvements. We’re not banking on any macro improvement or kind of market shift to higher growth in terms of what we’re guiding.

Raimo Lenschow: Okay, perfect. That makes a lot of sense. Thank you. Well done.

Operator: Your next question comes from the line of Pinjalim Bora with JPMorgan. Your line is open.

Pinjalim Bora: Oh, great. Thanks for taking the question, and congrats on the quarter. I want to ask you, Paddy, the AI strength is definitely palpable here, but I want to ask you if you are seeing attach rate of the core DigitalOcean offering as people create applications around the AI workloads? Is that flywheel AI driving more core DO starting to happen?

Paddy Srinivasan: Great. Good question, Pinjalim. Nice to hear from you. So the way I’m looking at it is that there are two ways that this cross sell or this cross attach happens. One, our DigitalOcean — traditional DigitalOcean core customers that are now starting to use our both platform-as-a-service as well as our infrastructure to consume some of the AI models that I was talking about. So it is still very early, but we have a handful of examples of some of our top customers that are trying to leverage our infrastructure for things like fraud detection models and things like that. So we are starting to see early signs of that happening. The other thing which is really interesting is startups and other model heavy companies that are coming to take advantage of our GPU as well as our platform infrastructure that we have.

Quickly realized that for them to scale their model and deploy it and as they start getting into the inferencing mode, they need a lot of the core cloud primitives that a platform like DigitalOcean offers from compute, network, storage, bandwidth, and having a geo — a global geo footprint to get inferencing with the lowest possible latency as close to their customers as possible. So we are starting to see very healthy early signs of these model-based companies, quickly realizing that for them to go live and start getting into the inferencing mode, they need a lot more than just raw GPU horsepower. They need all of the cloud primitives. Yes there are idiosyncrasies on how storage or networking works in the world of AI. But these cloud primitives are absolutely essential as the models get deployed and get into inferencing mode.

And that’s something that we’re already seeing a lot of times.

Pinjalim Bora: Understood, thank you. One question or two parts for Matt. The paper space, I think the assumption was that it will contribute about 3 points of growth this year. It seems like you’re already at about 3 points of growth. So wondering if you’re still expecting that kind of a contribution or a little bit more. And on EBIT, it seems like the beat is not flowing through the full year. Was wondering if there was something around timing of expenses or any catch up in the second half?

Matt Steinfort: Yeah, we’re still very confident in the, our AI platform contributing 3% overall growth to the company, so that, as you said, we remain confident in, and we’re encouraged by the early signs that we’re seeing in that business, and I think that’s what gave us the comfort to increase the bottom end of the guide and to increase the midpoint. From an expense standpoint, yeah, the investments as they come on, what drives the EBITDA, some of that will happen in the latter half of the year as we increase our investment in the R&D and we invest in additional space and power to accommodate the AI growth. So as we said, in our business, looking at it on an annual basis is way better than looking at individual quarters for free cash flow and in certain extent, even gross margin and adjusted EBITDA.

Because our expenses, when we take down incremental space or power or it wouldn’t hit EBITDA, but for a gross margin, if we take down additional equipment, it’s lumpy. It has a slight negative impact on the margins, and then we grow into it in the following quarter, which is what you saw from the increase in the EBITDA from the fourth quarter to the first quarter of this year.

Pinjalim Bora: Understood, thank you.

Operator: Your next question comes from the line of Kingsley Crane with Canaccord. Your line is open.

Kingsley Crane: Hi, thank you. So I want to touch back on the Paperspace conversation as well. I believe you have plans to create a more native link, a shared dashboard between Paperspace and the core DigitalOcean interface sometime this year. So could you just talk more about what you can do from a product perspective to encourage those potential cross pollination opportunities? And just what other gating factors do you think you could address to encourage this? Thank you.

Paddy Srinivasan: Yeah, thank you very much for the question. That’s a really good question, something that we are discussing on a weekly basis in our product conversations. Our first goal is to nail the use cases that our customers are picking us for. So we don’t want to miss out on that opportunity to make sure that we have the right infrastructure, right GPU fabrics, right networking, different types of storage attach that model builders and trainers and extenders need, the right infrastructure for inferencing. So we are really focused on that. Over the next couple of quarters, you will start seeing really good progress on bringing these two environments together, both from an infrastructure point of view, as well as from a user experience perspective.

So we are absolutely thinking about it. But we want that cross-sell and the attach between the two worlds to happen more organically. We’re certainly not going to rush into getting our customers attached before they’re ready to do it. That is naturally happening and I would just add that over the next couple of quarters, you will see a lot of cross pollination between the two worlds because there’s just so much natural technology convergence between what the current Paperspace offering is and the DigitalOcean core platform including our platform-as-a-service called app platform. Some of our storage and Droplet infrastructure is getting upgraded to also consume the GPU infrastructure. So you will start seeing a lot of natural convergence over the next six months, and we’ll keep reporting the cross-cell motions.

But as I said, our focus right now is to absolutely nail and address the needs of customers walking in to take advantage of our platform-as-a-service and the GPU infrastructure. So stay tuned. We have lot of other exciting things to come over the next six months.

Kingsley Crane: Thank you, Paddy. That’s really helpful. And so then either for Paddy or Matt, look, to take a step back, you’re building an exciting technology business, but you’re also highly profitable with a great financial model. That’s been core to the DigitalOcean identity for a long time. I know you just had Paddy’s first Shark Day in the past quarter. As we approach the next chapter of your growth story, how is employee sentiment and then how are you communicating that profitable technology mission internally?

Paddy Srinivasan: Yeah, great question. So, outside of spending time with customers, my number two time that in terms of time spent over the last 90 days has been with our employees, also known as sharks. And the employee sentiment, I don’t want to speak out of turn, but I think it is very robust, a lot of optimism, especially seeing the product velocity pick up over the last several weeks. And I only had time to go over the tip of the iceberg. Like literally I have only talked about three or four of a dozen or more product releases just in the last four weeks. So, the innovation phase has visibly picked up across the company, and that is a massive rallying cry for the company. So the sharks are super excited to get back into technology innovation, and as we do it, we get a tremendous amount of excitement from our community, And that is a major force multiplier for us internally to see the community really step up and take notice.

And they are our best evangelists. So you can actually see the action in various threads on X and Discord and other social media. There’s a question from one of our customers. In addition to our employees jumping to answer that question, it is often the community that provides the first level support and engage in a debate with customers and prospects. So that is a major force multiplier for us internally. So I want to say that there’s a lot of excitement internally, primarily driven by the pace of innovation and we’re also accelerating some talent addition, especially in the AIML space. We’re getting the core DigitalOcean Engineering team to also contribute a lot on our AI journey and just the enhancements that I rattled off in my prepared remarks in terms of our core innovation has really energized.

Then on the go-to-market side, it is still very nascent for us. We have a very robust customer service, customer support motion, and we are just in the early stages of taking all of the innovation that we are pushing out and translating that into a sustainable drumbeat of content and community amplification. So overall I would say the company is energized and we are starting to roll in the same direction across the company.

Kingsley Crane: Thank you, that’s exciting to hear. Thanks for the time.

Operator: Your next question comes from the line of Josh Baer with Morgan Stanley. Your line is open.

Josh Baer: Great, thanks for the question. I did want to ask about the sequential downtick in the learner customer cohort segments just for any context there. And then as a follow-up, somewhat related, just wondering more broadly if sort of having — doing any strategy shifts away from maybe going after some of those smaller customers, thinking about moving more upmarket from a more to — land more strategic customers that can expand and use more products on the platform.

Matt Steinfort: Thanks, Josh. No I think the strategy remains the same which is we have a phenomenal platform for developers and entrepreneurs and small growing technical — technology companies to come and experiment and grow their business. I think the shift that you saw in the last quarter was a result of a couple things. One, we are focused very much on the builders and scalers on our platform, which are still, that’s not upmarket relative to the industry, it’s just the larger of our customers. Because we think there’s a lot of expansion opportunity there. And a lot of that feedback that Paddy was talking about when talking to customers is consistent with things that we’ve said in the past is that we think we can get a bigger share of wallet of those customers, our existing customers, by eliminating some product blockers and adding capabilities that they find valuable.

The decline in the learners, it’s again that we had 476,000 learners, the decline of 70 or 8,700 was like a point and a half of that. It’s not a material decline. And it was likely more driven by the fact that we’ve tightened our screening of those small customers that, and if you think about the bad actors that show up on platforms, hosting platforms, you’re constantly fighting battles to try to keep them off your platform. They don’t typically pay or they pay, but they’re doing things that aren’t helpful in the Internet community. So we’ve ramped that up a bit and that contributed to a kind of a lessening of the number that we added in the quarter and so the net of the learners was down a bit. But I don’t think that’s a long-term trend. I think that’s just a result of a heavy focus on builders and scalers during the quarter and some tightening around our security practice.

Josh Baer: Thank you, much appreciated.

Operator: Your next question comes from the line of Tim Horan with Oppenheimer. Your line is open.

Tim Horan: Thanks, guys. Can we focus on the GPU CapEx spend a little bit? It sounds like you’re still capacity constrained. Maybe you can go into that a little bit. And can you give us a little color of maybe the payback that you think you’ll get for this? And I know you touched on quite a bit the cross-selling capability. It would seem like more spend on CapEx here will help out the overall revenue growth with any color in what you’re thinking with GPU CapEx spend. Thanks.

Matt Steinfort: Yeah, so as we said, we were definitely capacity constrained in the first quarter and we continue to deploy the capital that we had committed at the end of last year that’s consistent with our plan and we’ll be turning that up over the course of this year to give us the ability to increase the revenue growth and that’s all part of the plan and we’re watching to see how that goes and as we learn more as Paddy described about the specific requirements of our target customers, which are different from the requirements of some of the larger customers that are buying from the hyperscalers or the large GPU farms. We’ll make good decisions about whether we should be adding incremental capital beyond what we had committed.

The return profile on the business, you’ve got to think about it in two different categories. The platform-as-a-service offering, which is what we had acquired from Paperspace is more consistent with kind of the traditional cloud offerings we have where there’s software wrappers and capabilities around the hardware. The payback on that is very similar, a little bit lower gross margins than the core business, the core DigitalOcean business, but not that dissimilar. In the GPU-as-a-service or the hardware business, I’m sure you have the same statistics that everyone else on the call has around what an H100 costs and when you load on the networking and everything else and then when you look at the kind of going rate for GPU or H100 by the hour, you get into a — you get maybe $0.50 of ARR for the dollar of CapEx that you put in place.

You see paybacks in the less than three years. And I’d say that that’s lower, clearly, the lower margin than our core service. But you’re in an interesting part of the market where the cost curve hasn’t bent yet. You’ve got one supplier who’s controlling the majority of the inventory. You’ve got high demand for that inventory. And the market prices are fairly, I’d say, consistent in what you hear across all the different providers that are out there. And so, I think that what will happen over time, we’ll see what happens to pricing over time, whether people still get the dollars per hour they’re getting today for an H100, but certainly the cost structure will improve dramatically over the coming years as you have other suppliers come into the market and people just get better at deploying these capabilities at scale.

Tim Horan: Well, and I know your pricing can vary from $2 an hour to $6 an hour. Can you talk a little bit about how the average ARPU is looking for these products or people taking more the spot market pricing as opposed to mortgage or contracts?

Matt Steinfort: I’d say it breaks into the two categories that I described. And you have, where it’s hardware-as-a-service, it’s at one end of that range and where it’s more of the platform-as-a-service, it’s at the higher end of that range. And again, it’s still early. We’ve only had the GPU-as-a-service as an offering on the hardware side since mid-January. So we, like most of the people in the market, are figuring things out as we go.

Tim Horan: Thank you.

Operator: Your next question comes from the line of Mike Cikos with Needham. Your line is open.

Mike Cikos: Hey guys, thanks for taking the questions here. I did just want to circle up on the expansion of the overall portfolio of services that you guys have. I think DigitalOcean obviously very well known historically for its products with the Droplet, but wanted to see just given how this portfolio has expanded. Are you actually seeing a shift as far as where you’re landing with new customers? Are they landing on products outside Droplet or does Droplet remain that bread and butter when we think about how customers are coming to the DigitalOcean platform?

Matt Steinfort: Thanks, Mike, for the question. Well, Paddy, I think rattled off an impressive list of incremental capabilities that we’ve offered. We don’t view it really as a portfolio expansion so much as an enhancement of the capabilities that we have. So the company was started as an infrastructure-as-a-service provider selling Droplets, which are basically just compute, bandwidth, and storage. And as the company evolved over the last 10 plus years, we got into platform-as-a-service and offered managed database and Kubernetes and other capabilities that are really just kind of extensions and additional layers on top of the — of that core infrastructure-as-a-service, because our customers grew as, from individual developers and hobbyists to small technology companies and software providers that are running businesses on our platform.

And so, if you went back and listed all of the different products that Paddy had articulated, these are just incremental capabilities that these customers need as they leverage our platform-as-a-service and our infrastructure-as-a-service. They need more flexibility in how the products are configured with different ratios of compute to storage to bandwidth. They need things that make their lives easier because they don’t have giant IT organizations, so auto scaling and other capabilities that enable them to manage their infrastructure better. So we don’t view this any — really any of the products or services that we’ve offered as outside the bounds of the core target customer market that we’re serving. And it’s just as our customers are growing and evolving, we need to grow and evolve with them to keep making it simple for them to leverage the cloud.

Mike Cikos: Okay. And then the other question, more of a follow up here on the 2Q guide. We obviously have the 11% growth in hand, which is a slight deceleration sequentially. Can you just remind us what is that, I guess, Cloudways price increase that we’re going to be lapping? How much of a headwind does that represent when we think about the growth we’re looking at in Q2?

Matt Steinfort: Yeah, it’s interesting. This is why, again, coming from a different market into the software space and observing the, I’d say the obsession, but the focus on year-over-year metrics is interesting to me. Year-over-year metrics are both laggy and also what happened a year ago as a function is as important in the change in the growth quarter-over-quarter as what happened this quarter. If you look at the progression that we’ve guided from Q1 to Q2, we’re projecting increased revenue — increased incremental revenue, which implies an increase in a higher ARR than we added, incremental ARR than we added. So the current trajectory is improving. It’s not decelerating. But when you look at it, as you said, year-over-year, it is a slight deceleration.

And some of that, as you pointed out, is because we had a pop in the Cloudways business. It was about a 10% price increase a year ago. And so the 34% growth that we posted in Q1 for Cloudways is, it’s got about 10 points in there of a price increase and we’ll lap that in the second quarter. So that’ll come out.

Mike Cikos: Thank you.

Operator: Your next question comes from the line of James Fish with Piper Sandler. Your line is open.

James Fish: Hey guys. Paddy, in your opening remarks, you guys talked about discovering gaps in the product portfolio with customer conversations. What were those gaps that DigitalOcean needs to focus in on or were the release of products like backup and [caching capabilities] (ph) those gaps? And really are those gaps on the AI side too or what’s the differentiation on this infrastructure GPU-as-a-service launch against some of the larger players out there like the hyperscalers CoreWeave, Lambda, especially if we start to get supply more balanced in time?

Paddy Srinivasan: Great. Thank you, Jim, for the questions. So let me answer the first question first, which is some of the learnings that I described in my prepared remarks from a core DigitalOcean perspective. So those are what you are already starting to see in terms of our product delivery, right? So a lot of enhancements, as Matt just described, our platform-as-a-service offering is relatively newer compared to our Droplets. And we are — as we focus more and more on builders and scalers, there are some capabilities that they would love to see from us to help them scale as they grow. So you should expect us to release a lot of additional capabilities in the world of advanced reporting and management and visibility of infrastructure capabilities, security enhancements, advanced networking and global load balancing.

Those are the types of things that some of our advanced scalers and even builders are looking at DigitalOcean to provide because as their footprint increases and their business scales up, these are some of the things that they’re asking from us. So in my mind, these are all great opportunities for us to keep scaling our platform as our customers grow. Coming to the second part of your question, which is more around what is the differentiation from a AI, very specifically infrastructure-as-a-service perspective, as I said in my prepared remarks, this is — this feels remarkably similar to the origin story of DigitalOcean in the sense that we’re trying to democratize the accessibility of infrastructure for AI builders, extenders, and companies that are looking to deploy inferencing for their applications.

So the ease of getting started is a very durable advantage that we are looking to bring to our infrastructure-as-a-service and many customers have already started giving us feedback that it’s significantly easier to get started with our infrastructure-as-a-service. We are also looking at not just providing just bare metal GPU services, but we are adding different types of orchestration layers because it’s just not SSHing into raw H100 boxes. There’s a lot of complex things that needs to be orchestrated if you are a small startup or an ISV that specializes in say, ad tech, and you’re just looking to leverage a variety of different AI models, there’s a lot of technology that goes into building or even extending models and introducing, I’m sure a lot of you have heard of things like rags, which are a way to customize these models to make it work in your environment and take into account your context of the application.

So there’s a lot of complications that are involved, even if you’re not a model builder but a model consumer. And our software, as always, has been in the forefront of making it super, super simple. So our platform-as-a-service already does that throughout the full lifecycle of AI and machine learning development. And even our infrastructure-as-a-service goes all the way from bare metal to orchestrated abstractions to make all this easier for our customers. So we feel very good that we are taking our time, even though the momentum is building, we are still taking our time to really understand the needs of our customers at a very deep level, because what we don’t want to do is just satisfy some spiky buzz in demand. We want to build a business that is sustainable and we feel like inferencing is a very sustainable AI business model which will help us over the years to come.

James Fish: Very helpful and thorough answer. Matt, if I could sneak in one with you. You guys are reiterating your free cash flow margins at this point, which probably means that 15% to 17% CapEx range is still what you’re thinking. But GPU purchasing seems to be relatively strong. I’m going to guess that’s why CapEx was a little bit elevated versus what we were all thinking this quarter. Is it still on pace for that $50 million at this time, or should we interpret your language around potential free cash flow margins coming down in the coming quarters, as this is running ahead of plan, and really, are you thinking about using other GPU providers this year?

Matt Steinfort: So, I wouldn’t say we’re running ahead of plan on — what we had said is that we’re still deploying the capital that we had committed last year. So we have very good visibility into our capital spend on our current plan. What we had said is, we’re very encouraged by the progress that we’re seeing with the AI business and we’re learning a lot. Paddy’s only been here a couple of months. We’re evaluating the requirements of our customers and we’re learning about what customers we’re able to attract and what the requirements are and what the right configuration is from a technology standpoint. As we see continued positive signs and as we see the continued growth, we’ll make the right decisions on whether we want to spend any incremental capital beyond what was in the plan.

But we’re still on our plan. We’re not behind or ahead in any way relative to the capital intensity of our business. We’re just signaling that we’re encouraged and as we see more signs of encouragement, we’ll come back and provide updates as to whether we’re going to increase our spend any more than what we had said we would.

Operator: That concludes our Q&A session. I will now turn the conference back over to Paddy Srinivasan for the closing remarks.

Paddy Srinivasan: Thank you very much. As you just heard, we are accelerating our pace of innovation and delivering new capabilities in a very rapid cadence, which will help our customers grow on our platform. As I alluded to, there’s still a lot of work to do to take full advantage of our opportunity, but I’m very excited that we are moving in the right direction and continue to make steady, rapid, and respectable progress quarter-over-quarter. So with that, I would like to thank everyone for their time and talk to you all soon.

Operator: Thank you. That concludes today’s conference call. Thank you all for joining. You may now disconnect.