Q4 2024 Earnings Call Transcript

Good day, and welcome to the Fourth Quarter 2024 Akamai Technologies, Inc. Earnings Conference Call. All participants will be in listen-only mode. [Operator Instructions]. 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.

Good afternoon, everyone, and thank you for joining Akamai's fourth 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 those regarding revenue and earnings guidance, along with our business outlook, three to five year goals and longer-term targets. These forward-looking statements are based on current expectations and assumptions that are subject to certain 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 but are not limited to any impact from macroeconomic trends, the integration of any acquisition, geopolitical developments, and any other risk factors identified in our filings with the SEC. The forward-looking statements included in this call represent the company's views on February 20, 2025. Akamai undertakes no obligation to update any forward-looking statements, which speak only as of the date they are made. As a reminder, we will be referring to certain non-GAAP financial metrics during today's call. A detailed reconciliation of GAAP to non-GAAP metrics can be found under the financial portion of the Investor Relations section of akamai.com.

Also, as part of our ongoing commitment to transparency, we've enhanced our disclosures to provide investors with a more comprehensive understanding of our business. This quarter, we've created a new presentation available in the Investor Relations section of our website, offering a bit more information on our product portfolios, financials, and performance goals. This presentation supplements the information in our earnings release and annual filings, and we encourage you to review it. Within the presentation, you'll find an overview of select revenue and year-over-year revenue growth rates.

Please note, all growth rate percentages are reported on a constant currency basis. With that, I'll now hand the call off to our CEO, Dr. Tom Leighton.

Thanks, Mark. As you can see in today's press release, Akamai delivered solid performance in the fourth quarter with revenue coming in at $1.02 billion and non-GAAP earnings per share coming in well above our guidance range at $1.66. I'm also pleased to report that we made excellent progress on our multi-year journey to transform Akamai from a CDN pioneer into the cybersecurity and cloud computing company that powers and protects business online. For the first time in Akamai's history, Security delivered the majority of our annual revenue in 2024, surpassing the $2 billion threshold and growing at 16% year-over-year.

Our cloud computing portfolio recorded $630 million in revenue last year, growing 25% over 2023. The portion of this revenue derived from our cloud infrastructure services was $230 million, up 32% over 2023. Our cloud infrastructure services primarily consists of the compute and storage solutions that we've developed based on Linode. They also include our edge workers product and ISV solutions running on our cloud platform. Combined, security and compute accounted for two-thirds of Akamai revenue in 2024, growing 18% year-over-year. And we exceeded all our year-end annualized revenue run rate or ARR goals for the fastest-growing areas of the business, namely for our Guardicore platform, our API security solution, and enterprise revenue for our cloud infrastructure services.

These are three of the key areas that we anticipate will drive revenue acceleration for our overall business in 2026 and beyond. In the area of security, Akamai has expanded into new adjacent markets, growing beyond point solutions to provide a more holistic and comprehensive security offering. This has enabled us to expand our customer base and to better serve enterprises with a broader portfolio for protecting infrastructure, applications, APIs and user interactions in both cloud and on-prem environments. Security growth in Q4 was driven by continued strong demand for our market-leading Guardicore segmentation solution as more enterprises relied on Akamai to defend against malware and ransomware.

The Guardicore platform ended the year with an ARR of $190 million up 31% year-over-year and surpassing our goal of $180 million. In Q4, we signed our largest deployment to date for Guardicore with a leading IT services company in India. The solution covers 30,000 servers and nearly 300,000 endpoints. We also displaced a competitor segmentation offering that was falling short at major banks, both Hong Kong and in the U.S.

More than 80% of our segmentation revenue in 2024 came through channel partners, including one of the world's leading SIs, Deloitte, which wraps its services and implementation expertise around our segmentation and API security products to create value for customers that Deloitte knows well. Together, in Q4, we won a $5.8 million contract with Petrobras in Brazil to reduce the risk of a breach and ransomware attacks. We also partnered with Deloitte to help defend two large European banks from API risks. In Q4, Akamai also signed a large contract for API security with one of the biggest asset managers and brokerage firms in the U.S.

Our API security solution ended 2024 with an ARR of $57 million, up from just $1 million at the end of 2023 and exceeding our goal of $50 million. Taken together, our API security solution and Guardicore platform ended 2024 with $247 million of ARR. As we look ahead, we expect to generate continued strong growth for these products with a goal of increasing their combined ARR by 30% to 35% during the year. We also anticipate that over the next several years, the rapid growth and more meaningful amount of revenue from these new products will help offset the slower growth of our more widely adopted and market-leading web app firewall and DDoS mitigation products.

As a result and as noted in our supplemental disclosure that Mark mentioned a few minutes ago, we believe that we can maintain a CAGR of about 10% in constant currency for our security products over the next three to five years with typical M&A. This would bring us to more than $3 billion in security revenue by the end of the decade. While our security product line is performing very well, our compute product line is growing even faster and has a much larger addressable market. We've come a long way since we expanded into the cloud computing market in 2022 with the acquisition of Linode.

And we made a lot of progress last year, achieving what we set out to do in revenue growth, signing new enterprise customers, infrastructure deployment, product development, partner ecosystem expansion and migrating our own applications from hyperscalers to the Akamai cloud. We continue to sign compute customers at a rapid pace in Q4 and including two financial software companies in the U.S., two of the largest retailers in the U.S., a cybersecurity provider in Europe, an enterprise software company in Asia, one of the largest banks in Southeast Asia and an intelligent transportation system provider in Latin America.

When we measure the progress of our cloud infrastructure services, we've been looking at the results through two lenses. Our primary lens is the uptake of these solutions by larger enterprise customers, such as those doing over $100,000 in ARR. The second lens is looking at the performance of these products across customers of all sizes. At year-end, approximately 300 enterprises were spending at least $100,000 in ARR for our cloud infrastructure services, up significantly from the year before.

Collectively, these customers finished the year with an ARR of $115 million for our cloud infrastructure services, far exceeding our goal of $100 million. When you include all customers, the ARR for our cloud infrastructure services finished 2024 at $259 million, up 35% year-over-year. This is a substantial improvement from when we acquired Linode in 2022 when the ARR from these services was approximately $127 million and was growing in the mid-teens. In addition to enabling customer growth, our work to make Linode be enterprise grade has allowed us to move some of our most important products from hyperscalers to Akamai's Cloud.

This has resulted in improved performance and savings of well over $100 million per year. Many of our customers have also significantly expanded their use of our cloud infrastructure services over the last year, some by a factor of 4x or more. By year-end, 15 customers were spending over $1 million in ARR for our cloud infrastructure services, more than triple the number from 2023. And today, we announced our first customer to sign a contract committing to spend more than $100 million for our cloud infrastructure services over the next several years. We believe this is a remarkable validation of our new cloud capabilities, signaling that an extremely sophisticated buyer of cloud services is confident in our ability to execute and provide a level of service and performance comparable to or better than the hyperscalers.

The recent improvements that we've made to our cloud platform will enable us to do even more for enterprise customers in 2025. For example, in the last year, we expanded Akamai's core data center footprint to 41 locations in 36 cities around the world. Next week, we plan to announce that we've enabled our new managed container service in our 4,300-plus points of presence in more than 700 cities around the world. We're currently testing this service with customer workloads in over 100 cities. We've significantly upgraded our object storage solution with a 5x increase in scalability and a 10x increase in performance, making it comparable to the hyperscalers but with much lower egress fees due to the efficiencies of our unique edge platform.

We added GPUs for a variety of AI and media use cases. One customer ran a proof of concept between Akamai and a hyperscaler and then chose Akamai for text to image AI inferencing workloads. Another customer uses our cloud for AI-powered speech recognition for its in-vehicle voice assistant. And an OTT provider switched from a hyperscaler to Akamai to provide a more cost-effective platform for its ad-supported streaming TV service.

We enhanced the scale and security of our Linode Kubernetes engine product. Traditional cloud providers run Kubernetes platforms from a relatively small number of core data centers. Akamai's differentiated approach will combine the computing power of our cloud platform with the proximity and efficiency of our edge to put workloads closer to users than any other cloud provider. Building on technology that we acquired from Red Kubes last year, we released the Akamai app platform to enable developers to build and deploy highly distributed applications in just a few clicks.

And we added nine compute ISV partners last year, bringing our total to 23. Our ISV partners accounted for $36 million of cloud infrastructure services ARR at year-end. We're very excited about our opportunity for continued strong growth as we bring the power of compute to the edge with our broadly deployed network getting compute instances closer to end users with an open platform that ensures flexibility and portability, orchestrated resource deployment to ensure efficient scaling and operations and predictable pricing with an unmatched ability to minimize egress costs. I think our rapid progress in cloud computing has summed up well in an evaluation of public cloud platforms released last month by IDC.

Their worldwide public cloud marketscape for IaaS identified Akamai as a major player relative to industry peers, saying "Akamai has accelerated its journey into the public cloud IaaS space, transforming from a pure-play CDN provider into a formidable public cloud competitor." In addition, Gartner positioned Akamai as an emerging leader for Gen AI specialized infrastructure in their recent innovation guide for generative AI technologies. As we noted in the supplemental materials Mark mentioned, we're supporting a growing number of AI use cases with a special focus on inferencing. While it's still early days, we're excited about the long-term revenue opportunity, and we believe that the unique properties of Akamai's Cloud position us to be a major player in AI inferencing in the years to come.

As we look forward to the rest of 2025, our goal is to grow our total cloud infrastructure services ARR by 40% to 45% in constant currency. We believe that the accelerating growth of our cloud infrastructure services revenue will be driven primarily by enterprise customers. Given the great success that we're having with our cloud infrastructure services, we plan to focus more of our compute investments in this area. In particular, we're in the process of migrating some of our older cloud applications for tasks such as visitor prioritization, image and video management, and live streaming workflow to ISV partners who specialize in these areas, and we plan to move some or all of their workloads to Akamai's cloud.

In addition to converting former competitors into important ISV partners for our cloud, we believe this transition will enable us to focus more of our internal resources on further development and expansion of our cloud infrastructure services. The transition also means that we expect that the revenue from some of our cloud applications will decline in 2025. And as Ed will discuss shortly that we're projecting about 15% growth for our cloud computing solutions as a whole in 2025. As our cloud infrastructure services revenue continues to rapidly increase, we believe that we can reaccelerate the overall cloud computing revenue growth rate to achieve a CAGR of at least 20% on over the next three to five years in constant currency.

This would make cloud computing our third $1 billion product line by 2027. I'll next talk about content delivery, which continues to be an important generator of profit that we use to develop new products to fuel our future growth. Our unique edge platform with over 4,300 points of presence in over 700 cities continues to be a major differentiator in terms of lowering our costs, enabling massive scale and providing superior performance. And this is true not only for delivery, but also for security and compute. In security, we use the platform to provide a massive shield against all sorts of attacks without impacting performance or raising costs.

And in compute, we use the platform to provide function as a service with our edge workers product. And as we'll announce next week, we'll also use the same platform to run our new managed container service in thousands of pops across hundreds of cities. This capability is unique in the market, and it will enable our customers to get their compute workloads much closer to users. Akamai achieved substantial cost synergies by using the same physical server to support our delivery, security and now compute services in over 4,300 POPs in 700 cities.

It's a unique capability and a key reason why Akamai has been so profitable, while many of our competitors have struggled. Our installed base of delivery customers also continues to be a key contributor to our growth in security and cloud computing. As we harvest the competitive and performance advantages of offering delivery, security, and compute as a bundle on the same platform. That synergy works especially well for our security and compute customers that want delivery as a feature and see it as critical to their relationship with us over other vendors.

In Q4, we signed many deals that included security and compute solutions alongside our best-in-class delivery. And we won back delivery business for competitors at one of the leading tech players in AI and as a leading player in streaming media. And in December, we acquired select customer contracts from Edgio to offer their customers our market-leading delivery services and the opportunity to take advantage of Akamai's full range of security and cloud solutions. I'm pleased to report that we're beginning to see signs of improvement in the delivery marketplace with more customers willing to sign multi-year contracts with predictable pricing, a more stable pricing environment generally and early signs of stabilizing traffic growth.

As a result, and as Ed will discuss in a few minutes, we now expect to see the year-over-year decline in delivery revenue shrink to about 10% this year. If the favorable trends hold, we should see the decline in delivery revenue continued to lessen in 2026 and beyond. As we noted in our call last May, our largest customer is navigating political challenges and is pursuing a DIY strategy. As a result, we expect that the revenue from this customer will produce a headwind of about 1% to 2% per year on our overall revenue growth rate for the next couple of years before stabilizing at a level similar to some of our hyperscaler customers, which would be about 2% to 3% of our total revenue.

That said, I'm pleased to report that we entered into a five-year committed relationship with this customer in Q4, that includes a substantial minimum annual spend, which provides greater predictability and which reduces our exposure to their political situation in the U.S. While we're pleased with the progress that we made last year on our multiyear transformation journey, we still have work to do to reach more new customers and to cross-sell our new capabilities in security and cloud computing to our installed base. To drive greater top line growth over the next three to five years, we're transforming our go-to-market strategy to align more resources with the higher growth segments of our business and to accelerate the pace at which we add new customers.

In particular, we've already begun to raise the ratio of hunters to farmers and sales and to increase the number of specialized sellers and presales resources that support sales of our Guardicore platform, API security and cloud infrastructure services. We're also investing more in partner enablement as the channel has become a major source of revenue growth for us. Based on advice from one of the world's top consulting firms, we're also embarking on a major project to optimize our sales operating model, account coverage framework, compensation structure, pricing strategy and the way that we leverage our channel partner ecosystem.

As we disclosed in the supplemental forecast posted on Akamai's Investor Relations website, we believe that the combination of double-digit security growth, very fast growth in cloud computing, a stabilizing delivery business and a constantly improving product mix should enable us to accelerate revenue growth in the years ahead and to achieve double-digit revenue growth by the end of the decade, if not sooner. In fact, if you remove the impact of foreign exchange headwinds in the large customer I mentioned earlier, you can see that the acceleration is already underway. Excluding these two factors, revenue growth accelerated in 2024 over 2023. And as Ed will describe shortly, we anticipate further acceleration in 2025. We believe that improving our top line growth and product mix, combined with our continued efforts to improve efficiency, will help to improve operating margins so that we can meet and then exceed our goal of 30% over the next several years.

We also believe that we can resume growing our non-GAAP EPS in 2026. While we still have much to do, we're very optimistic about the future. Our cloud computing strategy is taking hold as we envisioned. Our expanded security portfolio is enabling us to deepen and expand our relationships with customers and partners. And we continue to invest in Akamai's future growth while also maintaining strong profitability.

Now I'll turn the call over to Ed for more on our results and our outlook. Ed?

Thank you, Tom. Before I begin, I want to reiterate what Mark mentioned at the start of the call, and highlight some of the new disclosures we issued earlier today. The materials we posted to the IR section of our website include a bit more detail than what we will cover in today's prepared remarks. Our aim is to provide deeper insights into our business and present our updated long-term goals.

While we do not intend to provide this level of disclosure every quarter, we will occasionally offer additional context if we believe it will be helpful. Today, I plan to cover our Q4 results, provide some color on 2025, including a few new disclosures and then cover our Q1 and full-year 2025 guidance and I will close with our long-term thoughts on revenue and profitability goals. Now let's cover our Q4 results, starting with revenue. Total revenue was $1.020 billion, up 3% year-over-year as reported and in constant currency.

We continue to see solid growth in our compute and security portfolios during the fourth quarter. Edgio contributed approximately $9 million of revenue in the quarter, which was in line with our expectations. Compute revenue grew to $167 million, a 24% year-over-year increase as reported and 25% in constant currency. In the fourth quarter, Compute revenue was comprised of $65 million from our cloud infrastructure services and $102 million from our other cloud applications.

As Tom pointed out, one of our largest customers has committed to spending at least $100 million on our cloud infrastructure services over the next few years. We expect that their workloads will begin ramping up by the end of 2025. Moving to security revenue. Security revenue was $535 million, growing 14% year-over-year as reported and in constant currency. During Q4, we had approximately $12 million of onetime license revenue compared to $5 million in Q4 of last year.

As noted in our added disclosure, our security revenue was comprised of $205 million from Zero Trust Enterprise plus API security, which grew 51% in constant currency year-over-year and $1.84 billion from all other security products, which grew 14% in constant currency year-over-year. Combined, compute and security revenue grew 16% year-over-year as reported and 17% in constant currency in Q4 and now represents 69% of total revenue. Our delivery revenue was $318 million, slightly ahead of our expectations and down 18% year-over-year as reported and in constant currency. We anticipate an improvement in delivery revenue in 2025 driven by a more positive outlook on delivery, as Tom discussed earlier.

International revenue was $490 million, up 2% year-over-year or 4% in constant currency, representing 48% of total revenue in Q4. Finally, foreign exchange fluctuations had a negative impact on revenue of $8 million on a sequential basis and a negative $6 million impact on a year-over-year basis. Moving to profitability. In Q4, we generated non-GAAP net income of $254 million or $1.66 of earnings per diluted share, down 2% year-over-year, flat in constant currency and well above the high end of our guidance range. These EPS results exceeded our guidance driven primarily by slightly higher-than-expected revenue, lower-than-expected transition services or TSA costs related to the Edgio transaction, greater savings from the head count actions we announced in Q3 and lower payroll costs due to some hiring pushing from Q4 into Q1. Finally, our Q4 CapEx was $193 million or 19% of revenue.

Moving to our capital allocation strategy. During the fourth quarter, we spent approximately $138 million to buy back approximately 1.4 million shares. For the full-year, we spent $557 million to buy back approximately 5.6 million shares. We ended 2024 with approximately $2 billion remaining on our current repurchase authorization.

It's worth noting that over the past decade, we have not only met our objective of buying back shares to offset dilution from our employee equity programs, but we have also decreased our shares outstanding by approximately 16% over that time frame. Going forward, our capital allocation strategy 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 move to guidance, there are several items that I want to highlight to help with your 2025 models. The first relates to revenue from the Edgio transaction. we now expect approximately $85 million to $105 million in Edgio revenue for 2025. In addition, we expect to have approximately $6 million of TSA costs in Q1 related to Edgio.

Approximately $4 million of those costs will be included in cost of goods sold and the remainder in operating expense. We do not anticipate any TSA costs beyond Q1. Second, we expect our CapEx to be approximately 2 percentage points higher than last year for the following reasons. First, we plan to increase spend by about 1% of revenue to accommodate the increased traffic resulting from the Edgio transaction. Second, we anticipate approximately another 1% of revenue will be used for a geo-specific infrastructure builds to support the recently signed $100 million cloud infrastructure services contract we announced earlier today.

Additionally, we expect our CapEx spend will be heavily front-end loaded with the first quarter seeing significantly higher expenditures compared to the rest of the year. This is due in part to the items I mentioned above in our plans to pull forward approximately $10 million to $15 million of CapEx into the first quarter to help mitigate the risk of potential tariffs that may be announced later this year. Third, we expect interest income to decline in 2025 due to lower cash balances resulting from recent acquisitions and our plan to retire our $1.15 billion convertible debt instrument that matures on May 1, 2025 along with expected lower investment yields as interest rates come down throughout the year.

Fourth, for 2025 we anticipate continued heightened volatility in foreign currency markets driven by unpredictable timing and magnitude of Federal Reserve policy changes and their impact on interest rates. As an example of how impactful FX can be to our results. Since our last earnings call in early November, the strength of the U.S. dollar has negatively impacted our 2025 expectations by approximately $18 million in revenue on an annual basis and negatively impacted our non-GAAP operating margin by approximately 30 basis points and negatively impacted our non-GAAP EPS expectations by approximately $0.09 per share.

As a reminder, related to currency, we have approximately $1.2 billion of annual revenue that is generated from foreign currency with the euro, yen and great British pound being the largest non-U.S. dollar sources of revenue. In addition, our costs in non-U.S. dollars tend to be significantly lower than revenue and are primarily in Indian rupee, Israeli Shekel and Polish Zloty. Fifth and finally, as Tom mentioned and illustrated in our supplemental disclosure, the traction that we are seeing in our business is being obscured by the challenges facing our largest customer and the significant FX fluctuations.

If we exclude the effect of this customer in FX, our revenue growth rate would have accelerated year-over-year in 2024 compared to 2023. Now moving to guidance. For the first quarter of 2025, we are projecting revenue in the range of $1 billion to $1.02 billion, up 1% to 3% as reported or up 3% to 5% in constant currency over Q1 2024. We anticipate that Q1 revenue levels will be slightly lower than Q4 due to the following: lower revenue from our largest customer, as we discussed earlier, the negative impact of foreign exchange, reduced onetime license revenue in Q1 from Q4 levels. Two, fewer calendar days in Q1 compared to Q4, that's two less days of usage revenue.

And finally, a significant number of delivery customers have calendar year-end dates on their contracts. Therefore, the aggregate impact of all those renewals tends to have a negative impact on Q1 revenue compared to Q4 levels. Note that we expect these items will be partially offset by a full quarter's benefit from Edgio. At current spot rates, foreign exchange fluctuations are expected to have a negative $7 million impact on Q1 compared to Q4 levels and a negative $15 million impact year-over-year.

At these revenue levels, we expect cash gross margin of approximately 72%. Q1 non-GAAP operating expenses are projected to be $310 million to $316 million. We anticipate Q1 EBITDA margin of approximately 41%. We expect non-GAAP depreciation expense of $132 million to $134 million.

We expect non-GAAP operating margin of approximately 28%. And with this overall revenue and spend configuration I just outlined, we expect Q1 non-GAAP EPS in the range of $1.54 to $1.59, we are down 6% to 3% as reported and down 2% to up 1% in constant currency. This EPS guidance assumes taxes of $57 million to $59 million based on an estimated quarterly non-GAAP tax rate of approximately 19.5%. It also reflects a fully diluted share count of approximately 152 million shares.

Moving on to CapEx. We expect to spend approximately $237 million to $245 million, excluding equity compensation and capitalized interest in the first quarter. This represents approximately 24% of total revenue. As noted earlier, we anticipate that our CapEx spending will be heavily front-end loaded with the first quarter seeing significantly high expenditures compared to the rest of the year.

Looking ahead to the full-year. We expect revenue of $4 billion to $4.2 billion, which is flat to up 5% as reported and up 1% to 6% in constant currency. We are providing a wider range than past practice due to the larger scale of our business, the uncertainty around our largest customer and continuing macroeconomic and geopolitical factors. We would expect to come in at the higher end of that range if we see significant weakening of the U.S. dollar, traffic growth materially exceeds our current levels and there is no ban in the U.S. for our largest customer.

We would expect to come in at the lower end of this range if we see significant strengthening of the U.S. dollar. Traffic materially slows from current levels, and our largest customer is banned in the U.S. Moving on to security. We expect security revenue growth of approximately 10% in constant currency in 2025. In Q4 included in security of the combined ARR for our Zero Trust Enterprise and API security solutions was approximately $247 million.

We expect the combined ARR of these two products to increase by 30% to 35% year-over-year in constant currency for 2025. For compute, we expect revenue growth to be approximately 15% in constant currency. Included in compute, our cloud infrastructure services delivered $259 million in ARR exiting 2024. We expect Cloud Infrastructure Services ARR year-over-year growth in the range of 40% to 45% in constant currency for 2025. At current spot rates, our guidance assumed foreign exchange will have a negative $38 million impact to revenue in 2025 on a year-over-year basis. Moving to operating margins.

For 2025, we're estimating a non-GAAP operating margin of approximately 28% as measured in today's FX rates. Decline in operating margin for the full-year '25 is due mainly to depreciation expense being approximately 90 basis points higher than last year and the negative impact of foreign exchange year-over-year. We anticipate that our full-year capital expenditures will be approximately 19% of total revenue, up approximately 2 points from 2024 due to the items that I listed earlier. As a percentage of total revenue, our 2025 CapEx is expected to be roughly broken down as follows: approximately 4% for delivery and security, approximately 6% for compute, approximately 8% for capitalized software and the remainder is for IT and facilities related spending.

Moving to EPS. For the full-year 2025, we expect non-GAAP earnings per diluted share in the range of $6 to $6.40. This non-GAAP earnings guidance is based on a non-GAAP effective tax rate of approximately 19.5% and a fully diluted share count of approximately 152 million shares. Before closing, I want to comment on our revenue and profitability goals for the next three to five years. Cloud Infrastructure Services is experiencing significant growth, which we believe will be the main driver of our total compute revenue in the near to medium term.

Our goal is to grow total compute revenue at a compounded annual growth rate of approximately 20% over the next three to five years. Moving to security. Given the size and scale in our security business, it is natural that we will see some products reach maturity and experience slowing growth rates, while other products in emerging areas will see more rapid growth. As a result, we expect security growth over the next three to five years to be led mainly by our newer Guardicore platform and API security products.

However, some of our largest security products like Web Application Firewall or WAF and Prolexic, are examples of some of our more mature products. We've been offering these products for over a decade and have reached a very high penetration rate within our customer base. As a result, we expect the growth rate for these products to begin slowing in 2025 and continue to moderate throughout the remainder of the decade. Given these dynamics in our security portfolio, our goal is to grow total security revenue at an overall compounded annual growth rate of about 10% over the next three to five years, including our typical level of M&A.

We anticipate delivery revenue declines will moderate in 2025 with increasing stabilization expected over the next few years. In summary, we anticipate that the combination of double-digit growth of our security products, the rapid expansion in our compute services, a stabilizing delivery business and a continuously improving product mix away from delivery will allow us to achieve double-digit top line revenue growth by the end of the decade, if not sooner. As for profitability, we expect operating margins to reach 30% plus by the end of the decade. And with accelerating top line and expanding operating margins, our goal would be to deliver strong growth in non-GAAP EPS over the next few years.

Finally, while we're not providing specific guidance for 2026, we do anticipate revenue acceleration and margin expansion compared to 2025 levels. With that, I'll wrap things up, and Tom and I are happy to take your questions. Operator?

We will now begin the question-and-answer session. [Operator Instructions]. The first question comes from Roger Boyd with UBS. Please go ahead.

Great, thank you for taking my question. And congrats on the quarter. And congrats on the large cloud deal. Tom, I love if you could maybe expand a little bit on that, provide any more details on the nature of workloads as customers bringing over, and it sounds like it's probably coming off a hyperscaler, but any additional info you can provide on the competitive environment there as it relates to pricing performance edge presence.

I think Ed noted there's some geo-specific conditions the customer is looking for. I know that's a lot, but thanks.

Yes. They already have been using substantial cloud infrastructure services at Akamai for a variety of media applications. And we'll increase that usage. In addition, they have some special needs in Scandinavia and we are building out a data center there for them to handle substantial applications for them throughout Europe.

So they're using the services in a normal way, plus we have a special situation that we're doing a build-out for them and development for them. But ultimately, we think many of our customers will use.

Super helpful. Appreciate the color.

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

Hi, team. Thanks so much for taking my question. First is for housekeeping. I just want to clarify that the $100 million compute deal is not the same as the five year deal that you signed with your largest delivery customer? And then I have a quick follow-up to that.

They are the same customer.

Got it. Okay. Thank you. And then the other one would be, can you just give us more details outside of just the geo location and why they chose Akamai. Is it -- does it have anything to do with maybe newer products that you have work on AI or intelligence, things of that nature would be helpful. Thank you.

Yes, they chose Akamai for superior performance and improve better pricing. We are very well connected throughout Europe. Nobody has the points of presence that we do. And so we can handle requests for this customer throughout the continent and get it into the data center facility they'd like to have in Scandinavia and do that very efficiently at a favorable price point, because of our existing platform and capabilities.

The next question comes from Amit Daryanani with Evercore. Please go ahead.

Thanks a lot. And thanks for taking my question. I guess I have two quick ones, hopefully. The first one, with your largest customer, the $60 million headwind that you talked about.

I'm hoping you just maybe expand a little bit on that is that headwind going to happen irrespective of what happens with the recent executive order around their operations? I'm just trying to understand like is this more about them intending to move some of the CDN in-house? Or is it more of an executive order that $6 million headwind does it happen in respective? And it sounds like it's another $60 million in '26. Maybe just level set that kind of what's happening there and how to ring-fence that would be helpful.

And then just to operating margins. What do you think it takes for Akamai to get back to a 30% operating margin number over time? Any kind of help on revenue or cost control, that would be helpful? Thank you.

Yes. So the headwind is, by and large, because of their DIY build-out, as we've talked about. So they'll be taking on more of the services themselves. And there are political challenges, obviously, in the U.S. as a result of the five year agreement.

Our exposure there is mitigated quite a bit. As we talked about in the disclosure last year, we had about $50 million, I have about $50 million of U.S. business with this customer that obviously would be taken away if they are banned in the U.S. However, as a result of the minimum commitments made, our exposure there is less. As Ed pointed out, if they are banned in the U.S. then we would have less revenue from them than if they continue to operate.

And if they continue to operate, there would be upside to the upper end of the range. And Ed, do you want to talk about the margins?

Sure. Yes. So most of the expansion is going to come through two things. One, the mix is going to move more towards security and compute.

And then just from an operating leverage perspective, with the revenue growth, we just believe we can scale the business accordingly. And a lot of that additional growth will drop to the bottom line. The changes that Tom is talking about or mentioned on the call about go-to-market, some of that has already been funded by the action that we took before, and we're very responsible with future investments. So we expect to see some scale as we grow top line.

Great. Thank you very much.

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

Good afternoon. Thank you for taking my questions. Tom or Ed feel free to chime in on this. With regards to TikTok, you talked specifically about the delivery business and some of the minimal commitment minimum commitments they've made to you to protect you.

But I was actually curious if you can opine or comment on if that is a customer that has adopted other solutions in your portfolio, i.e., specifically anything in the security realm that we have to be mindful of eroding away. And then I have a quick follow-up on the good market as well.

Yes. Like any large customer, they pretty much use all our services. So they're a large delivery customer. They use a lot of our security services and actually a growing amount of our compute our cloud infrastructure services. And we have factored that in to the guidance we've given, both the guidance for this year and the three to five year guidance.

Great. And just from a go-to-market perspective, Tom, you were very explicit in that there's been a lot of thoughtful change and maybe a little bit of an overhaul in the go-to-market organization by way of account segmentation, price strategy, et cetera. I'm wondering how far along are you in the process of ironing out or cementing a lot of these changes, because it does seem like there might be a couple of things that blocks from an aggregate go-to-market strategy perspective. So would love to hear kind of where you are on the journey to kind of overhaul the go-to-market organization and how we should think about you coming out the other end of that?

Thank you.

Yes. Great question. I think of this as sort of a two year process, roughly speaking. And we're in the beginning stages.

We've already increased the number of our hunters and as I said, the ratio of hunting to farming. We've already increased and we'll do further increases in the specialized sales team. We are working now, thinking through pricing structure changes, contracts, potential contract structure changes, making progress in segmentation for the field. So a lot of work still underway.

And you want to, I think, position this in the context that we're transforming the company. And you look at our product set and we're evolving substantially from what was a delivery company and then a delivery and stadium web-firewall company, where most of our revenue has been and you look at the products that are really growing and driving us forward to the future. And you're looking at some intake are core segmentation and enterprise Zero Trust security, you're looking at API security, huge greenfield there for us. And best of all, you're looking at our cloud infrastructure services, which has just really taken off.

It's remarkable. I think when you see what's happened. We bought Linode about three years ago, had about $127 million ARR, growing mid-teens at best. And today, more than double now growing at 35% last year, and we're forecasting 40% to 45% this year.

And I think what's really exciting about that is the addressable market for us is literally over $100 billion. And that's probably just in our current customer base. So tremendous potential, and it's really taking off now. So that changes the company.

And it changes what we need from the sales force, what they need to be doing. Obviously, we want to protect and grow the revenue in our existing base, but we want to be out there selling those new products, both in the existing base and to a lot of new customers. And so in response to that, we're in the process of transforming the sales force to get the capabilities that we need to continue the very strong growth we're seeing in these new areas.

Thank you.

The next question comes from Frank Louthan with Raymond James. Please go ahead.

Great. Thank you. Can you give us -- let's say, can you give us an idea of what the -- of what logo growth was in the quarter? That would be great. And is there any particular group security, compute, et cetera, that saw a better or a worse growth from new logos? Thanks.

Hey, Frank, this is Ed. Actually, logo gross, obviously was probably most positively impacted from the Edgio contracts that we acquired. We added a few hundred logos there, over 100 of those were new to the company. So that was the biggest addition.

As far as where the new logos are coming from, it's mostly in security in the two new areas that Tom talked about, Guardicore and API security, and also in the compute side as well. So I would say it's probably stronger than normal new acquisition quarter in Q4, just normal course and then obviously Edgio helped quite a bit as well.

And what's your outlook for being able to retain those Edgio contracts? Do you think some of those will churn off? Or what's the likelihood of those converting into more permanent customers?

Yes. So we've gone through the migration at this point. So we've moved everybody off. The Edgio network has closed so or shut down at this point.

So we have a good idea of who's coming over at this point. We don't anticipate significant churn from what we have today. We are very selective on what we took. We didn't take all customers.

There were certain things, certain small customers that didn't make sense for us to take. There's some acceptable use policy violations that we wouldn't take and there are some economics with certain customers that we didn't want. So factoring into our guidance is what we think we'll do with that customer base. I think there's probably some upside there with upsell generally speaking, you want to land the customer, make sure they're happy and then start developing a pipeline with them.

I want to just attack them right away with the sales pitch. So pretty happy with what we were able to retain. It was pretty much right in line, maybe a little bit better than we expected.

Okay. Great. Thank you.

The next question comes from Rishi Jaluria with RBC. Please go ahead.

Wonderful. Thanks so much for taking my questions. Just two quick ones. First, going back to some of the changes in go-to-market and changes in sales, what can you do on your end to minimize the disruption from that.

What are you assuming when you think about your 2025 guide. And I ask because in the history of software, we see sales regards changes to go-to-market and they always end up being a lot more disruptive than anticipated. So maybe if you could walk us through your assumptions and what you can do to minimize that disruption. That would be helpful.

And I have a quick follow-up.

Well, the first key is to avoid an account breakage. And so we're not doing it all at once. As I mentioned, this is a two year journey. And of course, we're adding resources with specialization on the new products.

And so those folks are helping do the cross-sell and have a chance to hunt for new customers. So the key really, I think is we will minimize the breakage of the rep associations with customers in the accounts. And Ed, is there anything you'd like to add to that?

Yes, I'd just say that as a mentioned, we're working with one of the large consulting firms and bringing some expertise in to help us think through that. But absolutely, we want to be very careful with how we think about the movement from hunting to farming and how we support customers going forward. The less breakage, the better.

Got it. No, that's helpful. And then just really quickly on Edgio. I appreciate the color and the detailed guide.

As we think maybe a little bit more longer term, given Edgio's out of the fold now, and you have those contracts and assets, how should we be thinking about the pricing environment for the delivery? And can just having fewer players in the space help maybe lead to a less aggressive pricing environment? Thanks.

Yes. That really comes down to the different buckets of customers. I'd say with some of the larger high volume customers, we are seeing some rationalization to some extent, longer contract terms, which is helpful. Obviously, if you have less players, you do potentially have some better pricing, but we're already starting to see that in the market.

So I'd say it's getting a little bit better, but it's always been a volume driven business. So to the extent that volumes continue to drive -- go higher, we'll still see the general trend in pricing to go down, but hopefully, we'll see it moderate a little bit more in the future.

Wonderful. Thank you.

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

Thank you. So Tom and Ed, I think this applies both you. Business looks to be going as planned, and we appreciate the three to five year goals and actually all the information that you gave us, this is great. But the guidance for next year implies 3% growth.

And I think there's a 1% of FX headwind, so that's about 4%. I get the large customer issues, but there's always issues, right? That aside for now, how do we get comfort that growth is going to accelerate to 10%. Is it just math, because I haven't gone through it at all yet to faster growing portions of the business will more than offset the delivery business.

And are you -- and/or, I guess, are you expecting improved macro backdrop in the forecast? Or does it -- I guess, also related to this, does that 10% eventual 10% exclude any impacts from that large customer?

All right. So I'll take this one, Tom. So I'll start with the back part. The large cost, what we've done is we've modeled out what their minimum spend would be for that five year contract.

So Think of that as being the bottom of the potential there for that customer. So there's more upside there potentially. But in terms of what we're seeing for growth, I think you have to kind of break the business down into the buckets and why we showed you all that data. We'll start with delivery first.

So delivery, obviously, has been a pretty big drag. It's starting to improve. We expect it to improve this year, continue to improve. So you should start to see some stabilization in delivery over that period of time.

So that you think of that as a big drag sort of moderating significantly, so you don't have as much of a drag. So then when you take the businesses that we have, we look at security, the bigger franchises are slowing a bit, but they're still growing. So we're still getting some growth out of that. And underneath that, say, for example, in our security products, we have new fraud products and be introducing new products.

So that franchise will still grow, just won't be growing as fast as it was. But the -- between Guardicore and APIs, those are very large markets. We have market-leading products we think we'll be able to -- that included with the -- some of the changes we're making in go-to-market and the investments we're making in hunting, we'll get more productivity out of the sales force. And there's still a long way to go there.

So we think that that growth as it gets bigger, you get 10% growth out of security, that we think is pretty durable given those factors. And then there's compute. So with compute, we're going through a little bit of a transformation that Tom talked about with the sort of [indiscernible], I don't want to say the legacy, but some of the other older products in the other applications bucket where we're partnering with some of the ISV partners, which has a little bit of a near-term flattening of revenue in that bucket, but it gives us a much bigger opportunity to grow the market by partnering with those folks getting some of their own business, but also getting a product that's being invested in, et cetera.

So we have a much more robust offering. And then again, with the go-to-market investments we're making in compute, not only just with the hunters, but also with the overlay specialists. We think that will also help drive more productivity in the field. So hopefully, that answers your question.

That was really good. That's really helpful. It makes you think we need to do a lot more work around, especially compute, but it's exciting to think about Guardicore and API. I guess just one thing, one last question on the compute business.

And one of the things that I can help think about here is profitability. And it's always a question at anybody that's coming into this business. But you alluded to having some advantages entering this business given the delivery business. I guess, can you describe that in a little more detail?

I mean, I kind of get the high level. Is it as simple as you're utilizing underutilized assets across your network? Or is there something more to it?

There are several reasons. First, we probably move around more data than pretty much anybody. And we've figured out how to do that very efficiently. And so for customers of our cloud infrastructure platform, when they compare our pricing to the hyperscalers, we're a lot less.

At least for the applications where they're moving around data, they got a lot of hits. And that's a lot of the applications out there today. And so in head-to-head competition now with the hyperscalers, we can offer a much better price. Second is performance.

And as we've talked about, we have the world's most distributed platform. We are in a position to get our customers' compute instances a lot closer to end users. And not only is that good for scalability, it's great for performance, because the business logic is running close to the users. And so when we do head-to-head trials now against the hyperscalers for a lot of the applications, we perform better.

So we're able to offer better performance at a lower price point. And now with the managed container service, that's super exciting. We can actually support our customers' containers in hundreds of cities. Nobody is doing that today.

And so again -- and those are the locations we've had already for the delivery business and the security business. So we're already there. So you get better performance and scalability at a lower price point. And then you combine that with the enormous market that we have a chance to tap into.

And I think probably really important as you make the models to look at our cloud infrastructure services revenue and what's happening there. We only really started selling that to enterprises last year and went from very little revenue to $115 million ARR in the total for our cloud infrastructure services, $259 million. And we're projecting that total number to grow by 40%. At least 40% to 45% this year.

So I would encourage you to keep a close eye on that because that drives a lot of growth and profitability for us going forward.

John, just to add on the profitability, there's four main synergies One is the backbone. So we obviously can leverage that. Obviously, we don't kind of track that we serve there. Number two is with operations.

So a lot of the folks that build out the CDN, also build out the data centers for our compute business, so we didn't have to go hire a big team. Number three is engineering. We talked earlier in the year that we moved about 1,000 people out of the engineering organization in our delivery business into the compute business, both in operations and engineering. And then the fourth thing is on go-to-market.

So we can leverage our existing customer relationships and our reps. We're augmenting that a bit with some of our specialists, but we don't have to go out and hire a whole new team. So there's an awful lot of operating expenses that we can take advantage of in terms of driving profitability.

Thank you guys. This all makes sense. Now it's the easy part. You just have to execute. That was a joke. It's not easy.

Exactly.

Thanks guys.

The next question comes from Patrick Colville with Scotiabank. Please go ahead.

Thanks for squeezing me in guys. So I guess this one is for probably Dr. Tom and Ed. I mean, in the prepared remarks, you touched on the reasons why compute slow.

But I just -- I'm going to my model now and the year-on-year delta is pretty stark. And so I just want to make sure I've got them completely nailed down. So I guess just to start off, this year, 25% growth in compute next year guidance of 15%. Can you just double [indiscernible] what is driving that?

Yes, sure. So there's the two pieces. There's the caught infrastructure security, which will continue to grow and actually accelerate. So that piece, it's the smaller piece of the two businesses, but that's going to accelerate pretty significantly.

When you look under the covers of the other, the application services, there's a number of things in there. For example, we have some of our legacy net storage business, which will be end of lifing that soon in the next year or two. And some customers will migrate to our cloud offering, some may decide to do something else, but that's going to be in a bit of a decline. And then we've got a lot of revenue that comes from some of these potential ISV partners.

For example, we've got some of our transcoding. We announced something with a partner called Harmonic, we're partnering with them, and we're migrating some of that business over there. They're using us now for their compute. So that's going to be shifting over time.

And there's probably four or five different buckets, whether it's image manager, video manager, our waiting room application or visitor prioritization. Those things were going to be moving towards ISV partners and focusing more of our efforts on the infrastructure services. So you're going to see flattening to maybe slightly down in that bucket, and you'll see most of the growth coming from cloud and infrastructure services.

Okay. Okay. Helpful. And then I guess as a quick one because I've been getting this a lot from investors. In terms of Akamai's exposure to the U.S. Fed. How should we think about that? And in 2024, is there any expectations for that business embedded into guidance for 2025?

So when you say the U.S. study, you talked about the Federal Reserve interest rate policy or you're talking about the federal government slightly?

The latter.

Okay. Yes. So we obviously do sell to the government. It's not a huge portion of our business.

It's way too early to tell with what the cost-cutting efforts that are going on, how that would impact us. But it's not an overly material part of the business. I wouldn't expect anything material out of that. We haven't modeled anything material, just kind of regular way business for now.

So I'm not anticipating anything.

Crystal clear. All right, thank you so much guys. I really appreciate it.

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

Hey guys. I appreciate all the details here, and increased disclosures, I think investors will as well. Maybe just following up on something you just said first, Ed. Can you just walk us through some of that transition impact on the compute side and which of the products you plan to sort of be transitioning off and how to think about the utilization of the compute network today that we need to invest this much after two years of investing as much behind the compute business?

Yes, I mean I gave you in terms of the investment percentage of revenue that we're spending for that. 1% of that of the six that we're using compute will be specifically for that big customer. And I've always talked about how from time-to-time, if we do get large customers, you do tend to sometimes get outsized demand in a particular geo. We added an awful lot of investment this year in terms of bulking up a number of locations, et cetera. We're going to continue to add to those.

We're also seeing very strong demand. So we're continuing to invest with that. But also keep in mind we did migrate over $100 million worth of our own applications. It's running well over $0.5 billion of our revenue today, and that business continues to grow.

So we're still having to provide some investment to support that, which is significantly less than what we would be paying if we were to continue to use third-party cloud providers. Now in terms of the two different components of revenue, I think I discussed the cloud infrastructure service. So I think you guys get that. That's the real fast-growing stuff.

And inside that other bucket is where you'll see that flattening out to declining. And it's really just the items I mentioned. We've got video management, image management, visitor prioritization and some various workflow components like transcoding that we have built from time-to-time for customers. There's not a ton of demand for it.

It's grown decently, but it's not a specialty of ours. So we want to make sure that we're investing in the right areas. So as we're migrating that to partners and we shift our investment into the faster-growing infrastructure cloud services, you'll see a bottoming out here this year and then the growth will reaccelerate to 20%, led by cloud infrastructure services. And also keep in mind, I've got about -- last time we broke it out, it was about $50 million of legacy storage.

So think of that as like the media companies storing like a backup origin for their video files and music file and that sort of stuff, images, et cetera. That's -- that business will be winding down over the next year or so. So that's going to have a bit of a drag on revenue as well.

Understood. And I think a lot of investors here you guys are sitting here on an earnings call, giving a three to five year framework. The last time we got this type of framework from you guys was many years ago. And I think in that framework, we had talked about a 20% constant currency security growth, for example, and we didn't hit that.

So you're talking about a 10% CAGR that includes M&A from here within security. Just what's going to drive the reason that we actually achieve these results and that this is a reasonable expectation that, that actually might be conservative as opposed to hoping that we hit that number? Thanks guys.

Well, I mean, to be fair, there were several years during that period of time when we did grow 20%. We always said 20% would include acquisitions. So we did some acquisitions along the way. We've -- look we just did -- put up 16%.

So that business has been very healthy. We've been adding a couple of hundred million of high-margin security revenue for years now. And if you do the math, we'll continue to do that. In terms of the impact of acquisitions going forward, we're not anticipating doing significantly large acquisitions, more of what we've done in the past, whether it's tuck-ins or smaller companies that are -- got a product that's in market that's about to scale like API security, where the company is looking at making a big investment in go-to-market or raising a route or looking to exit buy something like that or like Guardicore and we scale that up, and we've shown that we've got success there.

So I think we've got confidence in our M&A strategy. It's not going to be a significant portion, but maybe over time, it might be one of the faster-growing products as we exit the decade. And what we've been -- why we broke these businesses out for, you can see how well we're scaling a business now that's a couple of hundred million growing extremely fast. So I think we have pretty good confidence.

And again, yes, you can -- we can quibble over the 20% and maybe a couple of years, we're in high teens. But as we said, that would have included acquisitions we've been always doing. But I thought we did pretty well over that period of time. And we just thought now that we're getting to a scale of a couple of billion it made sense to update expectations as we see some of our products that we've been in the market for over a decade, slowing down.

The next question comes from Will Power with Baird. Please go ahead.

Okay, great. Tom, you provided some good examples in the prepared remarks, I think of some of the applications that are being used in your network. I mean it seemed for some time that you should be well positioned for inference at the edge. And I just I wonder if you maybe update on this, just as along the lines of the tone of conversations you're having, what kind of the pipeline looks like, just kind of the activity level you're seeing in terms of taking advantage of particularly some of these generative AI applications and inference that maybe is starting to develop or you just think it's going to start to develop?

Yes, I think there's a lot AI-powered at image normalization so that the various images have the same consistent size, quality image classification even for things like detecting disease in crops. AI-powered speech recognition in cars, AI-powered chatbots, sentiment analysis, tactic image inferencing for a variety of applications. All sorts of applications already using Gen AI on the platform. And we have partners, our ISV partners.

Some of them doing AI on the platform, web assembly, AI inferencing. So I would say it's early days, but all sorts of things being done on our cloud infrastructure services today.

Okay. And then I mean as we look at the cloud infrastructure assumed acceleration, I think you're looking for 40% to 45% ARR growth, any further breakdown, how much is inferencing influencing that? Or is that still early? And if not, what are kind of the key drivers of the acceleration there including impact of the $100 million deal, how does that factor in?

Yes, the $100 million deal doesn't do a lot this year. That doesn't really ramp up until the end of the year. So I don't think that's a material impact on our 40% to 45% projection. That will certainly help us in '26 and beyond.

And I don't think AI inferencing is a big part of the growth this year either. There could be some, but I think that will fall more into the upside category. I think it's just more of the traditional uses of compute on the platform. And keep in mind, that's a $100 billion market today, plus without getting into AI or the new large deal that we've done.

And just as examples, with our ISV partners, database at the edge, observability for all sorts of applications, media workflow, things like file transcoding and live encoding, video packaging, interactive WebRTC. These are what our ISV partners have their applications doing on our cloud infrastructure platform, digital asset management, video optimization, game orchestration, fleet management, DRM, Kubernetes activity and auto scaling ad insertion, service side ad insertion. So there's just a bunch of stuff that is just good old regular compute that's running on the platform. I would say we're -- as you can see from the examples we've given, we're selling it into pretty much all the verticals.

A lot of financial companies are starting to use it now. Media broadly construed is still, I think, the sweet spot and partly that's by design because those customers need to get great performance. They need to cut cost. They don't like writing a huge check to their biggest competitor.

And they're big Akamai customers, and they like Akamai a lot. So that's sort of the center of mass, but it goes across verticals and across a lot of use cases. And I'd say we're excited about AI. That's upside and probably comes down the road.

That's helpful. Thank you.

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

Thank you very much. Tom, interested to get your view of the ramifications for Akamai of all of the technological advancements and efficiencies that we've seen with the Deep Seek model? And whether you see any signs of that opening up kind of a broader wave of experimentation for doing edge inferencing on Akamai compute. Maybe some of these AI applications are going to become more economically viable.

And then I had a quick follow-up for Ed.

Yes. I think the Deep Seek phenomenon or announcement or whatever you want to call it, is very consistent with what we've been saying is going to happen. I think you'll see further improvements. It's great for us because it validates what we've been saying and how we've designed our cloud infrastructure platform.

In fact, there's already entities running Deep Seek on our cloud infrastructure platform along with other models. And it does mean that it is a lot less expensive. It doesn't need the giant core GPUs. It can be run on lighter weight GPUs.

There'll be a lot of use cases where you want that on the edge. And I think it's great and very much what we expected to happen. And I think you'll see more developments along those lines.

Okay. That's great to hear. And Ed, as a follow-up on the security CAGR where you're expecting about 10% in the next three to five years, that's fairly close, I believe to the overall market growing something like 12% to 14%. And you said that it includes a typical level of M&A.

If we think about it organically, should we pencil out something like maybe mid- to high-single-digits organically? And then if you're able to find some of these tuck-ins, you've done an incredible job picking those out and executing on those recently and they're growing like a weed. And then maybe the inorganic piece gives you a couple of few points there over the next three to five years. Is that a decent framework?

Yes, I think that's a good way to think about it, Mark. I mean obviously, at the scale we're at now over $2 billion, and we talk about even the bigger revenue companies we've acquired have been $20 million to $30 million. That's insignificance about a point in total from inorganic. In terms of the near term, there's going to be mostly from organic.

But yes, probably maybe one point, something like that as you get out a couple of years. And then obviously, if we picked a good company like we did with Noname or we did with Guardicore, as you get out further, we consider that organic growth at that point that might be contributing because it's a faster-growing area if we get into a hot space that's growing quickly. But in terms of like the -- to make a number in any given year, we don't anticipate getting much more than maybe 1% from an acquisition inorganically.

Okay, understood. Thank you so much.

I understand that we have time for one last questioner and that will come from Jonathan Ho with William Blair & Company. Please go ahead.

Excellent. And let me echo the thank you for the additional disclosure. Just one question from me. How concerned should we be with potential tariff impacts.

And can you maybe pass through higher costs that are associated with that? I know you're trying to maybe accelerate some investment ahead of time particularly given your need to invest on the compute side over the long run, it does seem like there's may be some exposure here. So I just wanted to understand the implications there. Thank you.

Yes. I mean it's obviously tough to call, just given we don't know what the ultimate end game is here in terms of towers. But to the extent we do have the ability to move supply chains around. And we're looking at obviously doing that.

There was some talk about Canada and Mexico. We do get some server builds out of the there. But -- so we fast forward it so. I gave you the number, $10 million to $15 million.

So it's not significant. In terms of can we pass some of these pricing costs along, we're certainly exploring that as part of the work we're doing with the consulting firm we hired overall pricing strategy is part of it. And to the extent that it becomes somewhat immaterial in any way, we certainly would have to bake that into our pricing. But it's tough to say at this point because we just don't know what the final end state is.

Thank you.

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

Thank you, everyone. As usual, we will be attending investor conferences throughout the rest of the quarter. We look forward to seeing you there and discussing everything that we talked about today. So thanks again for joining us tonight. I know it's a long call. We hope that you have a rest nice evening. Operator, you can end the call now.

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