Good day and welcome to the Second Quarter 2024 Akamai Technologies Earnings Conference Call. All participants will be in listen-only mode. After today's presentation, there will be an opportunity to ask questions. Please note this event is being recorded. I would now like to turn the conference over to Mark Stoutenberg, Head of Investor Relations. Please go ahead.

Thank you, operator. Good afternoon, everyone, and thank you for joining Akamai's second quarter 2024 earnings call. Speaking today will be Tom Leighton, Akamai's Chief Executive Officer and Ed McGowan, Akamai's Chief Financial Officer. Please note that today's comments include forward-looking statements, including statements regarding revenue and earnings guidance. These forward-looking statements are subject to risks and uncertainties and involve a number of factors that could cause actual results to differ materially from those expressed or implied by such statements. The factors include any impact from macroeconomic trends, the integration of any acquisitions, and any impact from geopolitical developments. Additional information concerning these factors is contained in Akamai's filings with the SEC, including our Annual Report on Form 10-K and our Quarterly Reports on Form 10-Q. The forward-looking statements included in this call represent the company's view on August 8, 2024. Akamai disclaims any obligation to update these statements to reflect new information, future events, or circumstances, except as required by law. As a reminder, we will be referring to certain non-GAAP financial metrics today. A detailed reconciliation of GAAP and non-GAAP metrics can be found under the financial portion of the Investor Relations section of akamai.com. I will now hand the call off to our Co-Founder and CEO, Dr. Tom Leighton.

Thanks Mark. I'm pleased to report that in the second quarter, Akamai delivered continued strong momentum in compute, strong growth in our security portfolio, steady operating margins, and healthy earnings growth on the bottom line. Second quarter revenue grew to $980 million, up 5% year-over-year as reported and up 6% in constant currency. Non-GAAP operating margin was 29%. Non-GAAP earnings per share was $1.58, up 6% year-over-year and up 9% in constant currency. These results were in line with or above our guidance. Before I provide more color on our performance, I'd like to review how Akamai is evolving as we grow. As most of you know, Akamai first made its name with the invention of content delivery services, and we're still the world's leader in that market today. We stand out for providing the scale and performance required by the world's top brands as we help them deliver reliable, secure, and near-flawless digital experiences. Recent examples include delivering the Euros football tournament and the summer games in Paris for top broadcasters around the world. As we've said in previous calls, our delivery business has been challenged in recent quarters by macroeconomic and geopolitical headwinds. Our plan for delivery is threefold. First, we will remain disciplined when it comes to the profitability of traffic that we choose to serve. Second, we will continue to leverage our market leadership position and installed base of major enterprises to generate cross-selling opportunities. And third, we will continue to take steps to retain our market leadership while also reinvesting most of the cash flow from our delivery product line into the fast-growing areas of the business. In Q2, delivery accounted for one-third of our revenue, or $329 million. This is quite a change from five years ago when delivery accounted for two-thirds of Akamai's revenue. The diversification of our revenue across new markets through continuous innovation has long been a core part of Akamai's strategy for long-term profitable revenue growth. A little more than a decade ago, we expanded our business into security with the creation of web app firewall as a cloud service. We did this to meet what we recognized as a growing customer need in a way that was complementary to what Akamai was already doing for customers with delivery. The opportunity was clear to us because we listened to our customers. We created what has proved to be a very successful cloud service for web app firewall. And now, for the first time in Akamai's history, security delivered the majority of Akamai's revenue, $499 million in Q2, up 15% year-over-year and up 16% in constant currency. This amounts to an annual run rate of about $2 billion per year. Of course, we greatly expanded our security product set over the years. We now offer market-leading solutions for DDoS prevention, Bot management, account and content protection app and API security, and zero trust enterprise security led by our Guardicore segmentation solution. Customer interest in our security solutions is strong and we had many significant wins in Q2. One of the world's largest energy companies became a new zero trust customer with Akamai. One of the top three airlines in the US is moving from a legacy VPN architecture to a zero trust architecture. With Akamai, we provided our app and API protector to one of the largest providers of HR management software and services in the US and to German retailers Delife, Douglas, Wagner, and Zalando. EFAFLEX, the German maker of high-speed industrial doors, purchased our segmentation solution, as did a major stock exchange and a leading cybersecurity company in Latin America. We provided DDoS protection to one of the largest banks in the world and to a government ministry in the Middle East and at one major electric utility in Southeast Asia. We replaced a well-known competitor in a five-year deal for our web app firewall, DDoS protection, Bot management, account takeover prevention, and API security. We're especially excited about the most recent additions to our security portfolio. In Q2, we announced our new Akamai Guardicore platform, the first of its kind to enable zero trust security through a fully integrated combination of micro-segmentation, zero trust network access, multi-factor authentication, DNS firewall, and threat hunting. Its single agent and unified control console, powered by GenAI, are designed to strengthen and simplify enterprise security with broad visibility and granular controls. The new GenAI interface enables our customers' operations teams to ask questions in human language to gain information about their enterprise networks. The new Akamai Guardicore platform reflects our evolution as a security vendor, growing beyond point solutions to a broader and more comprehensive security offering. Customers tell us they want to consolidate security products and tools with vendors they can trust, and we think this will appeal to their needs. In Q2, we also closed the acquisition of Noname Security as we accelerate our momentum in the fast-growing API security market. IDC forecasts this market will grow at a CAGR of 34% to nearly $1 billion by 2027. With Noname, we believe that Akamai now has one of the most comprehensive API security solutions in the industry. Within two weeks of the close, Akamai offered Noname customers our new Edge connector, an integration with Akamai web app and API protector that works with a click of a button. Noname saw a significant increase in closed deals in Q2, including wins at some of the largest banks and insurance companies in North America and at leading software companies in Europe and Asia. We're also beginning to see a good up-sell motion with early Noname adopters. For example, one of the largest US healthcare insurers more than doubled their Noname contract in Q2 to over $1.7 million annually. About a decade after we entered the security market, we again expanded Akamai's future opportunity by developing a much broader offering in cloud computing. As many of you know, Akamai has offered function as a service in our edge platform for many years. This kind of edge computing has been used by thousands of our customers, and it is deeply integrated into our delivery and security services. But our customers asked for more. They wanted us to offer full stack cloud computing so that they could run their VMs and containers on the Akamai platform. And they wanted us to do it in a way that would be more efficient and less costly than comparable offerings provided by the hyperscalers. They wanted Akamai to do this because they were already delivering and securing their sites and apps on our platform. They liked our track record of reliability, and they knew they could trust Akamai to be a good partner. Many of them also liked the fact that we don't compete against them. Unlike the hyperscalers, adding cloud computing to our portfolio also makes good sense for Akamai. In addition to satisfying customer demand, we can reap the advantages from offering customers delivery, security, and compute on the same platform. The synergies include improved performance, seamless integration, and other operational efficiencies, bundling for cross-selling and strong customer retention, increased margins for all of our services, deepening relationships with carrier networks, capacity to quickly detect and stop massive cyberattacks at the edge, unmatched visibility into enormous volumes of traffic, and the security insights and threat intelligence that we gain as a result. If you step back and look at how the marketplace has evolved, you can see how the hyperscalers have worked to achieve a similar suite of offerings, although they've taken a different route to get there. They started with cloud computing and infrastructure as a service and then moved into security and delivery, validating our view that there is synergy in offering customers all three together. The hyperscalers also have a more centralized architecture, while Akamai has the world's most distributed cloud platform, with more than 4,100 points of presence in over 700 cities across 130 countries. We believe that being more distributed provides customers with better performance, better economics, and greater reliability. As we reported in our last earnings call, the initial response from customers to our new cloud offering has been very encouraging. The strong early momentum that we achieved in Q1 continued in Q2, with compute revenue growing to $151 million, up 23% year-over-year and up 24% in constant currency. New compute customers added in Q2 include one of the world's best-known media and entertainment brands based here in the US, the European cybersecurity company Sekoia.io, Clara Video, the video brand of the biggest Telco in Latin America, MwareTV, a technology platform for IPTV and OTT services, and a cable satellite IPTV provider that reaches almost half the households in Australia. Customers are also leveraging our ISV partners, which we call qualified compute partners, to run low-latency workloads on our compute platform. These include solutions for observability into workload behavior, cybersecurity, and large-scale events where the need to store very large sets of data makes Akamai a more attractive and cost-effective option than competitors. Our media customers can now take advantage of a full suite of media workflow offerings on the Akamai connected cloud, which provides valuable synergy with our delivery platform for more efficient image manipulation, decision-making, and video transcoding. And with Akamai's latest qualified compute partner and customer Yospace, media companies around the globe can leverage their advanced ad tech and ad strategies at scale across the Akamai connected cloud. Customers are also building new apps on our platform where low-latency data distribution and processing provide a better user experience for their customers at significantly reduced cost. One customer is training and testing the machine learning engines that power their security scanning product. Another is building an AI-powered chatbot application to improve their customer experience and streamline operations with intelligent conversational customer engagement. Such AI-powered applications are increasingly popular with recent advances in large language models. Akamai is also a very large user of our new cloud solution. As a result of migrating most of our own apps from the hyperscalers to Akamai connected cloud, we're seeing better performance and greatly reduced cost. In fact, we expect to reduce our spending on third-party clouds to less than a third of what it would have been this year had we stayed on the hyperscalers, saving us well over $100 million in annual OpEx. It sure feels good not writing a nine-figure check to your competitors every year, and this is a feeling that we look forward to providing to our large enterprise customers. In summary, Akamai has undergone a fundamental transformation. We transformed from a content delivery pioneer into the cloud company that powers and protects life online. Compute and security now generate two-thirds of our revenue, and we believe that they provide Akamai with excellent potential for future growth and profitability. And we've achieved this transformation while successfully maintaining robust margins. Because both of our fast-growing product areas, our large security portfolio, and our rapidly growing cloud computing portfolio are built upon and enabled by the foundation of our business, our highly efficient and massively distributed delivery platform. Our near-term operating margin goal remains 30%, and we see potential margin upside over time as the fast-growing areas of the business expand our profitability. Looking back at the first half of 2024, we're pleased by our strong performance in security and compute. Looking ahead, we're very excited about our potential for future growth as we integrate Noname and as our fast-growing compute offerings continue to gain traction with customers. Now I'll turn the call over to Ed for more on our Q2 results and our outlook for Q3 and the full year. Ed?

Thank you, Tom. Today I plan to review our Q2 results and then provide some color on our expectations for Q3 and the full year. Turning to our second quarter results, total revenue for the second quarter was $980 million, up 5% year-over-year as reported, and 6% in constant currency. Compute revenue was $151 million, up 23% year-over-year as reported, and 24% in constant currency. We continue to be very pleased with the level of enthusiasm in the market as more and more customers are leveraging our enterprise compute solutions. In particular, we are seeing a broad array of enterprise compute use cases including live transcoding, secure access, observability, object storage, real-time log aggregation, insight spatial computing, and deep learning AI models across many verticals including media, e-commerce, software, and financial services. Moving to security revenue in the second quarter, security revenue was $499 million, up 15% year-over-year as reported, and 16% in constant currency. We're very pleased by the continued performance of our Guardicore Zero Trust solution and highly encouraged by the traction we are seeing in our recently launched API security solution. It's worth noting that the Noname transaction closed in late June and the revenue contribution in the second quarter was less than $1 million. Combined, compute and security revenue grew 17% year-over-year as reported, and 18% in constant currency, representing 66% of total revenue. Moving to delivery revenue was $329 million, which declined 13% year-over-year as reported, and 12% in constant currency. The decline in delivery revenue was primarily related to the revenue-impacting items that I outlined last quarter and was in line with our expectations. International revenue was $471 million, up 3% year-over-year and up 5% in constant currency, representing 48% of total revenue in Q2, foreign exchange fluctuations had a negative impact on revenue of $5 million on a sequential basis and a negative $10 million impact on a year-over-year basis. Non-GAAP net income was $243 million, or $1.58 of earnings per diluted share, up 6% year-over-year and up 9% in constant currency, and our non-GAAP operating margin in Q2 was 29%. Moving now to cash and our use of capital as of June 30, our cash, cash equivalents, and marketable securities totaled approximately $1.9 billion during the second quarter. We spent approximately $128 million repurchasing approximately 1.4 million shares. We now have an aggregate of roughly $2.3 billion remaining in our share buyback authorizations. We also used approximately $450 million in cash in the second quarter for the acquisition of Noname. As it relates to our use of capital, our intention remains the same: to continue buying back shares over time, to offset dilution from employee equity programs, and to be opportunistic in both M&A and share repurchases. Before I provide our Q3 and updated full-year 2024 guidance, I want to touch on some housekeeping items. First, regarding the close of the Noname security acquisition, we expect this transaction to add approximately $8 to $10 million of revenue in Q3, approximately $18 to $20 million in revenue for the full year 2024. We also expect it to be dilutive to non-GAAP EPS by approximately $0.04 to $0.05 for the full year 2024, and to be dilutive to non-GAAP operating margin by approximately 30 to 40 basis points in 2024. As a reminder, our updated full year guidance includes the impact of the acquisition. Second, and specific to traffic, we expect a modest uptick in year-over-year traffic in Q3, primarily due to the Paris Summer Games. This event is expected to drive approximately $3 to $4 million of additional revenue in the third quarter, and while Q4 is typically our strongest quarter seasonally, we saw a more muted impact of that seasonality last year, and we expect to see a similar result this year. Third, the country of India recently announced plans to eliminate its digital service tax effective as of August 1, 2024. We are working with our tax advisors to determine the full impact of this tax change on our non-GAAP effective tax rate. Based on our initial assessment, we believe this could result in a small increase in our effective non-GAAP tax rate, and we have adjusted our guidance accordingly. Finally, the macroeconomic environment remains challenging and geopolitical tensions persist. Any negative developments could have a meaningful impact on our business. So with those factors in mind, I'll move to our Q3 guidance. For Q3, we're projecting revenue in the range of $988 million to $1.008 billion, or up 2% to 4% as reported, and 3% to 5% in constant currency over Q3 2023. At current spot rates, foreign exchange fluctuations are expected to have a positive $2 million impact on Q3 revenue compared to Q2 levels and a negative $5 million impact year-over-year. At these revenue levels, we expect cash gross margins of approximately 73%. Q3 non-GAAP operating expenses are projected to be $307 million to $312 million. We expect Q3 EBITDA margin of approximately 42%. We expect non-GAAP depreciation expense to be between $129 million to $131 million and we expect non-GAAP operating margin of approximately 29% for Q3. Moving on to CapEx, we expect to spend $166 million to $174 million. This represents approximately 17% of our projected total revenue. Based on our expectations for revenue and costs, we expect Q3 non-GAAP EPS in the range of $0.58 to $0.59. Guidance assumes taxes of $59 million to $60 million based on an estimated quarterly non-GAAP tax rate of approximately 19% to 20%. It also reflects a fully diluted share count of approximately 154 million shares. Looking ahead to the full year, we now expect revenue of approximately $4.02 billion, which is up 4% to 5% year-over-year as reported, and up 5% to 6% in constant currency at current spot rates. Our guidance assumes foreign exchange will have a negative $20 million impact on revenue in 2024. On a year-over-year basis, we continue to expect security revenue growth of approximately 15% to 17% in constant currency in 2024, including the contribution from the acquisition of Noname. And given the strong momentum and adoption from both new and existing enterprise compute customers, we now expect the enterprise compute annualized revenue run rate to double from the $50 million we reported last quarter to over $100 million as we exit 2024. As a result, we are now increasing our overall expected compute revenue growth to approximately 23% to 25% in constant currency for the full year 2024. Moving to profitability, we estimate a non-GAAP operating margin of approximately 29% and non-GAAP earnings per diluted share of $2.14. Our non-GAAP earnings guidance is based on a non-GAAP effective tax rate of approximately 19% to 20% and a fully diluted share count of approximately 154 million shares. Finally, our full-year CapEx is expected to be approximately 16% of total revenue. In closing, we are pleased with the traction we are seeing in enterprise compute and look forward to helping our customers migrate services to the Akamai connected cloud. Thank you. Tom and I would now be happy to take your questions.

We will now begin the question-and-answer session. The first question today comes from Patrick Colville with Scotiabank. Please go ahead.

Thank you so much for taking my question. Frank and Ed, really great to be part of the hack. My story I want to talk about the business makes shift. I mean, that's where you opened your prepared remarks. Specifically, I want to focus on compute. The $100 million revenue you just called out from Akamai connected cloud is really compelling and great to see that ramping. When might that hockey stick to become an even greater revenue base? What's the trajectory of Akamai connected cloud over the coming quarters? And if we think out beyond that?

Yes, great question. And I got to say, we were very pleased to see the rapid early adoption. That's a capability that we really just started selling in earnest this year. We had some very early adopters towards the end of last year. And if we can get up to $100 million ARR by the end of the year, which we think we're going to do, that's great for the first year of the product, and then we'll see where we are at the beginning. Next year, we'll give guidance in February for the year in compute, but we're very optimistic about strong growth in compute driven by the enterprise customers and our new capability. There's an enormous market there, obviously, and so we're really looking forward to tapping into that.

Very helpful. Thank you. I guess the second part of my question I want to ask about delivery this year. You've been very clear about the headwinds to the delivery business in 2024. I appreciate you might not want to give guidance beyond 2024, wondering whether the headwinds we're seeing right now are cyclical headwinds or are they structural in nature? Thank you.

Yes, I don't think what we're seeing today persists over the long term. Traffic, I think, will continue to grow maybe at a little bit of a slower rate than it did certainly during the COVID times. But delivery, I believe, is here to stay. We are intent on remaining the market leader by a good margin. It's a very profitable business for us. We're very careful about that. We're very efficient in what we do. And it's very synergistic with our security web app firewall business and our emerging compute business. So I don't see these headwinds persisting over the long term. There are geopolitical considerations that we're worried about for next year. But I don't think this is a long-term phenomenon. And in any case, given the very fast growth of our security and compute product lines, they've nearly tripled in revenue over the last five years. So now where they're two thirds of our revenue overall, I think what you see with delivery, with sometimes challenges, sometimes good, it has a lot less impact on the overall growth rate as we go forward. Ed, do you have anything you want to add to that?

No, I think you covered it well.

The next question comes from Keith Weiss with Morgan Stanley. Please go ahead.

Excellent. Thank you guys for taking the question and congratulations on the solid quarter. I wanted to ask you a little bit about kind of parsing out the guidance, particularly the full-year guidance. If I'm doing my math right, the midpoint of the full year comes up by about $5 million for the full year. But we're adding Noname, it sounds like about $20 million in revenues for the full year. Is there a part of the equation that's coming down, a part of the business that you're getting more cautious on that makes up that difference?

No, Keith. Yes. So we included Noname in our guidance last quarter as well. So there's nothing that's changed. If anything, the business has gotten a little bit better. So our guidance has come up a bit to reflect that.

Got it, got it. And then on the expense side of the equation, the savings from moving sort of in-house from the hyperscaler is $100 million is real savings. Congratulations on that. That's quite a feat. Tom, you talked about the ability to sort of start pushing that more into operating margins in the near future. Can you give us an indication of what near future means? 2025 near future, or are we thinking two or three years out or further?

Well, yes, the savings, operating savings we're getting, as Ed said, we've been plowing that back into the business by and large so that we can invest in growth. There's more savings to come there, but we really get the upward pressure on margins, the beneficial tailwinds on margins as the mix shift continues. As we add compute customers, that is good margin for us. And it's accretive. Security, as we add customers there, it's accretive. Now, as Ed noted, today with the new security products, initially they're dilutive, but as we grow the revenue there, every customer we add, every deal we sign improves margins. So that over time, 30% is our goal, we're very close to that today, but over time, we think we have good potential to grow beyond that.

The next question comes from John DiFucci with Guggenheim Securities. Please go ahead.

Thank you. I have a question on Guardicore. So, segmentation broadly seems like it's becoming more relevant in the market. Customers are more accepting of it. It's no longer like a new thing, and it has been for a while, and you guys have been there, and you bought Guardicore a couple of years ago, but frankly, it seems like an essential component to a zero-trust environment. Not everybody has it. So can you talk a little bit more about how about this business, but really about your Akamai Guardicore Zero trust platform that you just launched a few months ago, and how that sort of fits in the ecosystem of some of an enterprise when they need to protect all their assets to establish that zero-trust environment? Because it always seemed to me that this was essential, like I said, for zero trust. I guess if you can also, in addition to the technologies and what else it fits in with, can you also hit on your channel efforts regarding this platform? Because I know this is sold through the channel. I think, Ed, you said this before, but it seems like a really sophisticated solution. It's not just you're buying a firewall for somebody or something like that. If you can just talk a little bit about how.

Yes, I think you characterized it very well. Segmentation is essential. I mean you've got to lock the doors and the windows as best you can, but now if someone still gets in, and I think really the most important thing an enterprise can do is lock down everything inside, and that means Guardicore. It means having your agent on every application and on every device. And you're right, most enterprises don't have it. When you go back a few years, and I think the community at large really disfavored segmentation, and that's because the way it was done back then was really crude. You did it in hardware, it's very inflexible, hard to do. And at the end of the day, if you did it at all, you had giant segments which defeated the whole purpose. You weren't very secure because the malware would get in and wipe out an entire giant segment, and you had a big problem. And Guardicore solved that problem through software, very easy to manipulate, make updates, much more secure, fine-grained controls. And so it's been an education process for us in the marketplace. We viewed it as something that was going to be essential and that, I think, is proving out, as you noted. And of course, with the ransomware headlines and disasters, it's not surprising to see why people are waking up to what they really do need this. So now the next question is with the platform. And there, what we've done is combine the Guardicore, which is protecting inside app-to-app device-to-device communication with the employee device to internal applications. And so we've combined what's called North, South, and East, West. Now again, you go back a few years ago, they were different buyers and treated differently. But then we were thinking back then, boy, it's going to make sense to put this all together, and sure enough, we're now seeing customers say, 'Hey, we want that on the same platform. We want a single agent, not two different agents, and we got to deal with a single control panel so that the business logic can be applied to employee devices at the same way and the same time as it is to internal applications.' And so that's what the Guardicore platform is all about. We actually also combine it with DNS firewall, multifactor authentication, threat-hunting capabilities so that you can tell when you've got now where it is, what's going on into a platform which customers are excited about. And I think it's important not to underestimate the importance of a single agent to do this because that's really important real estate. And the new control panel powered by Gen AI, it's actually pretty cool. You can converse in human language, if you will, with your network infrastructure. You get much greater visibility probably much better compliance as a result, which means better security. Now your channel question, yes, Guardicore is all channel. And you're right, it is a sophisticated integration and deployment. It's not just like throw a firewall out there. And that's where the partners can really add value. And so in some cases, in many cases, the partner will derive even more revenue than Akamai will. And it's ongoing because you're growing your Guardicore, your segmentation footprint to include more applications and devices, and it's great for partners when they can add value and generate revenue. So it's a really good channel-friendly product. And of course, not easy to do per se, but a lot easier than the way segmentation used to be.

And if I could, Tom, that all makes a lot of sense to me. But it also raises the question of what are people doing? What are the alternatives? I'm familiar with Illumio and another company called Truxport, but like you mentioned, many enterprises don't even have segmentation implemented. Some do, but many don't. So are there other options that I'm missing? Is Palo acquiring the entire Palo platform? Are they suggesting that it's not necessary, or that they have something similar? I'm just trying to understand because the opportunity seems really significant here.

No, I think it is a big opportunity. And very few, relatively speaking, enterprises have it today. The early adopters are the critical infrastructure companies because they really, really have to have it. And we do compete with Illumio. They're probably our leading competitor. We believe the Guardicore solution is a lot better. We actually have our own mini firewall in the Asia. We don't have to rely on the firewall in the OS, which sometimes won't be there, it might not be consistent. We can cover a lot of the legacy systems which that's important to enterprises to get more universal coverage. I think there is a ton of greenfield. And I wouldn't be there when one of the other competitors is talking about their platforms. They probably don't spend a lot of time talking about segmentation because they don't really have a solution for it.

The next question comes from Mark Murphy with JPMorgan. Please go ahead.

Ed, what is your latest thinking on the FX headwind to the full-year revenue forecast? I'm just curious if there's been any movement there from, I think, previously, you've been looking at that as, I believe, $40 million headwind.

Mark, yes, not much of a change. The dollar moved around quite a bit during the quarter, up and down, but it's pretty much exactly the same. So for the full year, it's about $40. I gave the guidance already in the quarter in terms of the impact quarter-over-quarter and for year-over-year, but it's still about the same, just around 40 for the full year.

Okay. And then, Tom, as a quick follow-up, you mentioned a pretty wide array of the workload types that you're seeing on your cloud computing platform. And you mentioned toward the end, deep learning and AI models. I'm wondering if you can double-click on that. For instance, what are the types of model? Are you seeing LLM, text models or image models or something else? And is it possible to estimate what percentage of your cloud ARR might be relating to those newer types of AI workloads?

Yes, great question. I would say today, AI workload is probably a small fraction of the ARR. I think over time, there is potential for growth there. As we talked about useful in security applications, chatbots, tailoring content for commerce companies, ad targeting, recommendation engines. I would say that the models are smaller because they're more focused. The giant models sort of are used to learn everything, your chat GPT, you can ask it any question at all, have it try to be knowledgeable about everything. Those are giant models, and we're not really targeting that business. But for our customer base, they tend to be a lot more focused on what they're trying to do. Maybe it's a commerce site, maybe it's an ad site, maybe it's a security company. And they don't need to learn the whole world to really provide value. In fact, we see that with our own solution with the Akamai Guardicore platform with the control interface powered by Gen AI. Really, it's a very specific application, which means that you don't need the gigantic model to really provide the value. And that means it doesn't have to run on this giant suite of GPUs, it can run just great on our platform, which has GPUs but primarily CPU-based, which gives us much better ROI. And that works great for what our customers are looking to do in terms of their AI applications.

The next question comes from Madeline Brooks with Bank of America. Please go ahead.

Hi team, thanks for taking my questions. I want to continue on the discussion of Connected Cloud, and you mentioned some nice wins in enterprise outside of your traditional CDN customers. But I guess I just wanted to double-click on that and kind of get a little bit of color. What are those customers for nonmedia, non-e-commerce? What are the use cases that those customers are finding from Connected Cloud? And if you could just help us break out the growth of those customers maybe versus growth of your more traditional customers? Are they growing around the same in terms of their adoption of Connected Cloud?

Yes, we're seeing growth both within the base and outside the base. I think, for example, with our qualified compute partner program with observability, a lot of companies need that to know what's going on with their applications. Security companies would need that. Now, we also have a lot of media customers. And I would say that's probably the biggest segment today. We have, by design, a full media workflow ecosystem now supported through our QCP program on the platform and so a lot of media customers starting to take advantage of that. Outside of that, for example, OS and firmware patch storage, personalized waiting room experience, improved page performance with hints and so forth. We talked about with AI tailoring the site for a user based on what they've been doing so far. Real-time log aggregation and insights into your logs, observability, kinds of things. We even have a PBX, a telephone switching system running on it, 5G Internet gateway running on it. So it does broaden the base which is, I think, exciting for us in the future. But today, probably the biggest segment would be media.

The next question comes from Alex Henderson with Needham. Please go ahead.

Great. First off, I think congratulations to an order on the great results out of both security and out of the compute. And I wanted to focus a little bit on the compute side of the business because I think ultimately that's the area that needs to be proven out to the Street more than anything else. Can you talk a little bit about the mechanics around what portion of the customer base that's converting to compute is coming from internal? What portion of the compute are true new customers? How much of the growth is coming from existing customers that are increasing their upsell? Just kind of look at it as if it was a traditional stand-alone business and give us some of those critical metrics that go into analyzing the success of that business?

Yes. I'll take our first half and turn it over to Ed. And the answer will be pretty similar to the last question. I would say our biggest users and the biggest segment for compute today is our large media customers. And that's by design. And a lot of our QCP, our Quality Compute Partners, are media workflow companies, so that's sort of the biggest segment. I would say observability as a capability is a very large one as well, and that spans across all verticals and would include new customers. So we do have a bunch of customers in nontraditional Akamai verticals that are using compute. And I think over the longer run, that opens up a whole new market verticals for us. But the biggest chunk today would be existing Akamai media companies is the biggest. And Ed, do you want to add some color on that?

Yes, sure. Alex, so as Tom mentioned, we're seeing growth in both existing applications for stuff that's been on the platform for a while, but the bulk of the growth is actually coming from new customers. The new customer additions are growing very, very quickly. We broke out some numbers for you last quarter, and that continues to ramp very nicely. And we're seeing a significant increase in the pipeline. We are seeing some new customers come to the platform. And what's interesting is we're probably seeing more workloads and repeatable workloads in areas even. Tom talked about media, but outside of media. And we're seeing customers that may be relatively small CEM customers are fairly large compute customers. So I'd say it's across the board where we're seeing the growth, but it's mostly from adding new customers to the mix and then they start to ramp.

The next question comes from Fatima Boolani with Citi. Please go ahead.

I wanted to be in on the delivery guidance and the expected performance in the back half. So appreciate you experienced a lot of the traffic degradation patterns in the first half. But I'm just curious why the trajectory of the business is actually worsening in the back half? And then I have a follow-up for Tom, please.

Yes, sure. So I talked a little bit about the expectations for Q4. Just based on what we're seeing now in terms of traffic growth and what we saw last year, we're not expecting the normal hockey stick to any significant degree like we've seen in prior years. And also keep in mind, we closed the transaction with StackPath and Lumen last year, and that's all delivery revenue. So that makes sure Q4 tougher comparison if you're looking at sort of year-over-year growth rates. That's going to skew your perspective a bit. And as we talked about on the last call, there were some dynamics going on with one of our larger social media customers. The good news there is we've got a good handle on that. That's sort of playing out as we expected. But as we talked about, those are the factors that as you put that into your model, why it may look like it's deteriorating a little bit. But I'd say the biggest issue is the fact that you're anniversary-ing the StackPath and Lumen contribution from Q4 last year.

That makes sense, and it's very fair. Tom, you've contributed significantly to the compute opportunity, and there are numerous examples of the momentum you've gained in that area. However, from a broader strategy perspective, as you look to scale that franchise and integrate it into the overall revenue narrative, how are you ensuring that compute serves to enhance your existing business rather than just compensating for slower growth in the delivery franchise? What strategies do you have in place to drive wallet share growth and ensure you're continuing to meet the demand for compute among your existing delivery customers?

Yes, compute is distinct from delivery. Delivery revenue is not being redirected into compute; that's not the scenario here. The potential for compute is significantly larger than that of delivery, and I believe it will evolve into a much larger business over time. Delivery will continue to perform well for us, offering strong cash generation, cross-selling opportunities, and favorable economics for our platform. This enables us to provide compute services, especially for applications that are data-intensive, at a more competitive price compared to hyperscalers. However, it's important to note that this isn't about addressing any shortcomings in delivery. I see compute as a major driver of revenue growth for us in the future, independent of our delivery segment.

The next question comes from Jim Fish with Piper Sandler. Please go ahead.

This is Quinton on for Jim Fish. Maybe touching on that first question there. A competitor in the space recently talked about pricing pressures from some of the largest media customers getting worse over the past couple of months. Are you seeing those pricing trends in the market that's maybe driving some of that second half weakness alongside, obviously, the StackPath and Lumen impact? Or are you not really seeing these pressures given your decision to move away from these kind of lower-margin delivery opportunities?

Quinton, this is Ed. As we talked about, we had some large renewals this year. So obviously, those we've been through all those, and there is certain pricing pressure there. I'd say it's nothing different than what we've traditionally seen in the marketplace. I wouldn't say that it's significantly worse. I think the issue is just not as much traffic growth. So as you reprice a customer, you tend to see significant traffic growth. So the revenue declines don't persist as long as they have in a situation like this. So I wouldn't say if anything has changed in terms of the trajectory of the pricing. It's always been very competitive, it always will be. So that's really not the issue. The issue is just we're not seeing the type of traffic growth that we normally see.

Yes. That's really helpful. And then obviously, it's still really early here with the Noname acquisition. But any update you can provide on the integration between Noname and your kind of existing Neosec API opportunity? And maybe how you balance that go-to-market of those two platforms and how you can kind of leverage a full suite to kind of grow the wallet here within a customer?

Yes, we are now going to market with Noname. And as I mentioned, within two weeks of the close, we had a fully integrated with Akamai products, existing Akamai products, in particular, our web app firewall where a lot of the APIs would go through that. So I would say we're basically integrated today. We have some Neosec customers who we are maintaining over time, that will evolve into the Noname product with some of the capabilities from Neosec, so we get the best of both worlds.

Yes. Just to add, the Noname acquisition came with a pretty sophisticated channel as well as a number of specialists. So we have both of those, so that's going to help drive some sales. And actually, from the minute we announced the close to when we closed or announced the deal ultimately closed, we saw a nice pickup in deals closed. So not no impact on the funnel and the teams already out there selling. So very excited about that. I think we've just enhanced our go-to-market capability as part of the acquisition.

The next question comes from Jonathan Ho with William Blair & Company. Please go ahead.

Hi. Good afternoon. As you listen to customers and what they need or want from the compute side of things. Are there any core services or capabilities that you feel like you're missing or you're going to add pretty soon that are maybe potentially catalysts for even faster adoption?

I believe the most notable change today is the size of the marketplace. The hyperscalers have a vast marketplace, and we are just beginning to enter that space. We are thrilled that we now have a very competitive media workflow marketplace and a strong observability marketplace, both of which we plan to grow further. Additionally, we are expanding our distributed compute capabilities, allowing us to operate in more locations than many hyperscalers, which will enhance our performance and enable us to better comply with data sovereignty laws in various countries. This development is continuous, and we are actively improving the platform, including our storage solutions. Our ongoing efforts will persist in the foreseeable future. From our results, it’s clear that we are now in a position to actively market our offerings, and it’s encouraging to see the rapid early adoption.

That makes a ton of sense. Just in terms of the delivery business, as we continue to see this decline as a percentage of revenue, it seems to be carrying the business in terms of margins. How concerned are you over deleveraging effects and CapEx efficiency as we see sort of the two sides of the business move in opposite directions?

Yes, I said it's not a huge concern there. The margins of the new products are very high gross margin products, so that will be helpful. Sure the compute business is a bit more capital intensive. But we've been able to drive down the CapEx of the core business and delivery pretty dramatically. So you're down low single digits as a percentage of revenue for that business. That will maintain as long as the delivery, as long as traffic is not growing significantly. And part of our strategy and being more selective of the type of peak traffic to average that we're taking on the platform is by design, is to make sure that we do maintain that efficiency as we're spending more CapEx in the compute business, and we're not taking on that sort of not as profitable peak-to-average type traffic so that we can maintain a low CapEx posture in the delivery business.

The next question comes from Tim Horan with Oppenheimer. Please go ahead.

Kind of a few part question on cloud. Can you use your own platform? It sounds like you're moving a lot of legacy services on there to create kind of, you think, unique services for yourself and new services or improved legacy services on that platform. And then secondly, if you look at Microsoft, Google Cloudflare, they're kind of growing cloud in the 30% range. Do you think you can kind of get there? And are you kind of maybe CapEx constrained to do that? Or just maybe talk about how you can kind of get up to your peers there. And what do you have to do in the SMB market to hit that type of 30% growth?

Yes, on the first question, we have built capabilities for ourselves as part of our migration. So we're off of Snowflake and Data bricks which we had big spends there. We're looking at over time, making our capabilities available to customers, again, a service that is more efficient, which I think is really important for customers. In terms of the growth rate, I would say that you want to compare the enterprise compute number, which we've talked about was a $50 million ARR at the end of Q1. We think that will more than double by the end of the year. That's sort of the number you want to look at that's comparable. We've got more in the overall compute number, but those are have products like image and video manager, legacy Akamai net storage, other kinds of things which aren't as comparable for what you're looking at in terms of the hyperscaler growth. So if you focus on the enterprise compute, which is really going to be the growth driver for us, that's going in a very fast percentage now and, of course, on a much, much smaller number than the hyperscalers. Ed, do you want to add to that?

No, I think that's exactly right. And that's something that, over time, as that number becomes more material, we'll start to break that out for folks to make it easy for you to see where that growth is coming from. But Tom is absolutely right. That's where the big market is. That's where we're seeing explosive growth, and we see that we think that can continue. Obviously, the pipeline is growing. We're seeing a lot more use cases than we expected. I see a great participation from all of our reps across the world. It's not just one geo. So we're very excited about it.

The next question comes from Rudy Kessinger with D.A. Davidson. Please go ahead.

Hey, great. Thanks for taking my questions on security. I guess in the second half, if I look at it adjusting out Noname, it looks like organic growth at constant currency is just about 11% to 12%. Obviously, it's been a slowdown versus the last several quarters. And in general, the security growth rate has kind of been pretty volatile over the last 5 to 6 quarters. Could you just give us the puts and takes on maybe some tough compares in the second half? I think you had some spikes in DDoS strength last year. But also just going forward, just what's the kind of right growth range that we should expect out of the security business?

Yes, sure. So just in terms of some of the puts and takes, just remember, last year, in Q3, we had a little over $6 million of license revenue, that's a couple of percentage points. So that's going to make your Q3 compare a little bit more challenging. The other thing to keep in mind, too, last year, we introduced some new security bundles for web app firewall that did phenomenally well. So we've anniversaried that, so that makes your compare a little bit more challenging. And then between Guardicore and API security as they start to ramp, we think APIs going to ramp very, very quickly. It's just a smaller number. So as you've talked about, the growth has bounced around a little bit over time. That happens as you bring new products into the market and they start to ramp up, think about when we brought Bot manager to the market. It was a small product and then go to hundreds of millions of dollars. I think the same thing you'll see with Guardicore and API security.

Yes. Okay. Regarding the delivery outlook, it seems that Fastly implied that the issue extends beyond just one social media client, as several large media companies appear to be diverting traffic to more affordable providers. However, it seems you are not observing that trend, or perhaps it was already incorporated into your guidance for the year. Any thoughts on that?

Yes, no, we're not seeing a phenomenon of someone moving to low-cost providers. As a matter of fact, there's two less of them in the market today. So we're not seeing that. As I talked about earlier, we have a tough compare with a hard Q4. And then we've just seen just a lower traffic year. Gaming has been unusually weak, and video traffic isn't as robust as it normally is. That stuff happens from time to time, but we're not seeing a shift to low-cost providers or any new low-cost providers in the market.

The last question today comes from William Power with Baird.

This is Yan Samilton for Will. Ed, just going back to that more muted Q4 seasonality you're expecting for delivery again this year. If I remember correctly, last year you were pointing to weaker trends in retail, including an uptick in bankruptcies there and then also weaker in terms of gaming. And now I know it's probably still a little early to forecast though, but is it those same verticals where you're expecting weaker traffic again this year that you're informing your expectation? Or is it some others, or maybe it's more broad-based?

Yes. I'd say it's a combination of those two verticals. We've seen sort of over time the retail seasonal or to be less and less. Some of that has to do with the Zero overage product that we offer in the market. But just in general, it just hasn't been as robust as have been, say, 4 or 5 years ago. And then from gaming, yes, still weak. I haven't seen any major launches or we're not hearing anything from our customers that would leave me to believe that Q4 will be strong from that perspective. And then also, as I just talked about several times here, just traffic in general has been a bit sluggish from a growth perspective in general. So I don't see anything that tells me that that's going to change going into Q4. So obviously, we've got a few months to go here before we get there. We'll update you when we talk again in November. But based on what I'm seeing today, just out of this, it's worthwhile calling that out to folks as they build out their models.

This concludes our question-and-answer session. I would like to turn the conference back over to Mark Stoutenberg for closing remarks.

Thank you, everyone. In closing, we will be presenting at several investor conferences throughout the rest of the quarter and the rest of the year. We look forward to seeing you at those. We hope everyone has a nice evening tonight. Operator, you may now end the call.

The conference has now concluded. Thank you for attending today's presentation. You may now disconnect.