Akamai Technologies, Inc. (AKAM) Earnings Call Transcript & Summary

Thanks for joining us this morning. I'm Madeline, and I have the pleasure of hosting Ed from Akamai here to talk all things delivery, security and excitingly compute as well. So it should be a good half hour. And also too, this is as much more for you guys than it's for me, so if anyone has any questions, please feel free to hop in and raise your hand and we'll get a microphone going.

Perfect. So I guess, at high level, you reported 1Q results, can you just walk us through your thoughts on the print and performance for the first quarter?

Sure. Well, first of all, thanks for having me. Nice to see everyone. Yes. So the first quarter, very strong results in compute and security, which is really the future of the company and where all the growth is a little bit of a challenge with delivery, we are seeing some slowdown in Internet traffic, and that's persisted into April and why we had to take our outlook down a bit for the delivery business, but very, very strong demand for security across multiple products. In our compute business, we actually broke out the enterprise compute where all of our focus is today, and it's about a $50 million ARR business growing at 300%, and it's very healthy in terms of the participation of multiple customers, multiple use cases, about 200 customers or so we're considering to be larger customers. So very, very healthy growth there.

Maybe I will just stay with delivery for a second before going on to the future drivers of the company. But trends in delivery, I guess, first, I'd love to understand as we look forward to the back half of the year, traffic is historically is very difficult to predict. So how are you mitigating the risk on just predicting the traffic patterns?

Yes. I mean we've been in the business for 25 years. So we have a pretty good indication of what our customers are doing. We also work with all the major telcos, so we get information from them on what they're seeing because we don't carry all of the Internet but we do carry a significant amount. But with our biggest customers, which drive, 80-20 rule where a handful of customers drive most of the traffic. We got a -- we sit down with them on a regular basis and talk about what trends they're seeing, what they're expecting, if there's new product launches or if there's things they're doing, we talked about one big customer who was doing some optimizations on their site to reduce the amount of traffic that they were serving, still serving with the same experience but doing it more efficiently in terms of how many [ bits ] they serve. So we get a pretty good indication. And also, there's some seasonality that we see. Usually, Q4 tends to be your biggest season as you get kids going back to school, you've got holiday shopping season, tends to coincide with new launches of content, live sports, et cetera. So you generally know that Q4 is going to be big, hard to predict. The summer tends to be a little bit slower. What happened, recently, over this last quarter is, we did see some slowdown in the industry where we're seeing. It could be related to a number of different things. The writer strike could even have some effect on it, where we just haven't seen the release of new content. Obviously, that will pick up again at some point. Gaming has been weak. There hasn't been a lot of new games being launched. That tends to be a big driver of traffic. But in terms of being able to predict it, we get a lot of good insights from our customers. That's generally the best information to work off of.

And I think the questions on everyone's mind about your one large customer. Talk to us about the risk for the rest of this year as well as the risk of political events, if there was full ban. And would that fall into -- is it this year risk, maybe towards end of the year or 2025?

Yes. So we added some disclosure in our latest 10-Q that we were not required to do, but we thought it would be helpful for investors to understand. There is a large customer of ours in the social media space that has been banned in the U.S. That ban goes into effect next year. Now who knows if that will actually take an effect. There's a lot of -- there'll be a lot of core challenges. But what we did is, we size it up, it's a little less than 1.5% of our total revenue. So if that were to go into effect, not this year but next year and it goes to 0, that's how much revenue it would be. We haven't disclosed how big of a customer they are across whole globe. Unlikely that they get banned everywhere or we get told we can't work with them. I don't see that happening, but it's possible, you never know. Some crazy people around the country, so you never know what can happen from a regulatory perspective. But that's the associated risk, but it's not -- doesn't go into effect this year. So it's not a this-year risk.

CDN, you guys are really using it as more of that profitability driver for the company. Can you just walk us through over the next few years as you look to continue to invest in security and in compute you'll see those CDN profits, talk to us about margins on CDN.

Yes. So CDN is also sort of the core of what enables us to be able to deliver security. We're applying our Web Application Firewall rules as we're delivering the content so that there's no latency for the end user. It also gives enormous amount of data. So there's a ton of value that we get out in the CDN business, we've got scale. So it helps us with buying power for our compute business, et cetera. We got all the great relationship with our enterprise customer. But you're right, it does spit up a lot of cash for us. We've gotten to a scale that no other CDN has been able to get to from a profitability perspective. Most of our bandwidth, for example, is free. So we don't pay for the bandwidth in a lot of cases because we carry so much traffic. We deliver our content inside a lot of the carrier networks and provide a lot of value, so they don't charge us for bandwidth. Sometimes they don't charge us for space and power either. So we've got some great economics there and that spits off of -- last time we broke it out, we talked on an Analyst Day a couple of years ago, and we gave some relative profitability. We report as one segment. We have so much cost to allocate. We couldn't do pure segment reporting. But that's -- we think it's in the 20s in terms of operating margin, 20% operating margins, it spits up a fair bit of cash. But also the security business now is at a scale and that's spitting up a ton of cash as well. So those profits are being up back into the compute business, obviously, a bit more capital intensive. We've actually moved a lot of people out of security, sorry, out of CDN engineering into the compute space, a lot of people that build the CDN in terms of procuring the data center space, racking and stacking and deploying the servers. Those folks have been moved into the compute business as well. So we're getting some scale from people as well as the profit dollars.

But maybe just to wrap up the delivery discussion quickly, where do you see delivery for Akamai in the next 2 to 3 years? Has it just become this base rate where, obviously, you leverage the delivery network to run your other products? Or are there ways to then maybe you could try and escape a little bit of a commoditization within the industry and still specialize and be a leader there?

Yes. I think for the stock to work, if you will, the future is really the security and compute, right. We're not anticipating that the CDN business is going to be a growth driver for the company. It's all the strategic reasons we talked about and the cash flow that we get from that business, the relationships, et cetera. But there will be -- I don't think that's a growth business because you've got some dynamics of large customers that renew and unit economics, prices -- the 25 years, 24 years I've been here, volumes go up, unit prices go down. I don't expect that to change. Could there be less commoditization in the market, it's possible because you're seeing a lot of players exit the market because it is hard to make money if you don't get to the scale. And there's some other competitors that have some challenged financials, but I'm not anticipating that. I'm not expecting that to happen, but I would like to see stabilization there to see that business be flat, maybe down a few percentage points when you do have a concentration of large renewals.

Great. Well, I think we should move onto security, especially given your recent acquisition of Noname. But maybe before talking about Noname, could you just give us an overview on Akamai's lens of security and how the business has performed ex that acquisition right now?

Yes. So it's been a phenomenal business for us. And I think a lot of folks have underestimated just how big the spaces that we're in are and our ability to continually innovate and get into new adjacencies. So today, the security business is roughly a $2 billion business for us on a run rate basis. It grew last quarter 21%, we gave guidance for 15%-ish roughly growth for this year. So it's been a great business for us, and it's really sort of 2 main [indiscernible] of the business. The majority of the revenue today comes from protecting customers' websites and web applications. So things like Web Application firewall, bot management. A lot of the traffic on the Internet today isn't humans, it's machines. So helping our customers identify what machines are interacting with them, what to do with those bots, if you will. We have API security, which we'll talk about in a minute, which is an area of potential huge growth for us. And the other part of our business, smaller but growing really fast is protecting the individuals who work at companies, so think of those more traditional security. So access -- instead of using a VPN, we have an access product, we have a Secure Web Gateway where we're filtering traffic going inbound and outbound or north-south. And probably the biggest growth driver for us on that side of the business is, what we call, micro-segmentation or looking at traffic East-West. So as applications and servers are passing traffic back and forth, a great way to prevent the spread of attacks. So no matter how good your perimeter is something always gets in, micro segmentation is a great way to be able to identify things and stop malware from spreading in your organization. So that's been a great growth driver for us.

And it's always interesting to hear about acquisitions from the CEO perspective, but I think people really appreciate it from the CFO perspective. So maybe with Noname, from your seat, why now? And why Noname specifically?

Yes. So if you look at API security and that's the space that Noname is in, if I look at traffic on the Internet today, the fastest growth of traffic is API traffic. Every modern application uses APIs, and it's also an area where a lot of the attackers are starting to exploit. People have open API gateways, they're not protected and you're seeing a lot more attacks. The number of attacks that you're seeing on APIs is growing dramatically, and I think we'll just continue to increase. And there's not a lot of good solutions out there. And our customers were saying to us, I need help. I want to know, first of all, what is my universe out there? How many API gateways do I have? What's my universe? How many of those can be exploited? How many are being exploited? And then help me do something about it. And there's not a lot of companies out there today that offer any solutions there. The leader in the space was Noname. And the idea with API security first is to answer that first question, which is what does my universe look like? What is the potential threat vector for me? And Noname did a great job of integrating with most of the major sources of API traffic today. So things like the major CDNs, the major clouds, Apigee, Kong, all those -- [indiscernible] all the places that have a lot of API traffic, they've done a lot of integration so that they saw a much bigger footprint of the universe of APIs that are out there. The company was growing very quickly. They had very high valuation expectations a year ago, much more realistic now. So that's going to prompt in the why now. But it was great for us to be able to pick up the market leader in that space. We had acquired a smaller company about less than year ago that, I would say, had better efficacy in terms of being able to answer some of those second questions in terms of what's being exploited and what can I do about it. And they offered a data lake where you save a lot of that information, you can do analytics on what's happening over time. So combining those 2 products together, we think we have by far the best solution in the market today. There's lot of estimates that this market will be a $1 billion market in the coming few years. And I think the combination of the 2 gives us a great opportunity to really be a major player in the API security market.

And just to talk about growth potential estimates here. I mean, with the inclusion of Noname, how does that change your projection -- prediction for growth for security?

Yes. Yes. So it's going to be -- obviously, it's the acquisition we talked about it contributing about $20 million in the back half of the year. So think about it as kind of a $40 million run rate business. We have our own API business that was growing very fast, but very small when we acquired Neosec, it was sort of pre-revenue at that point, it had a handful of customers. So with a $2 billion business, it takes a while for that to become material. So it's not going to have a major impact on growth for us this year. But over the next several years, it's going to be one of the major growth drivers for us.

And what do you think supports growth? I know you guys have mentioned before trying to get around that mid-teens growth level and really stay there for security. So what continues to support that growth? And when you look -- when you think about growth as well, how do you think about organic versus inorganic as you grow along?

Yes. So we don't -- when we give our guidance, we don't factor in inorganic unless we buy something. So we had been growing this business, generally, between 15% and 20% for last several years. Inorganic has generally added a couple of points, 2% to 4%, generally, for all the acquisitions that we do. I don't see us going and buying something significant or transformative from a size perspective, obviously, getting into -- you could argue that getting into API security like we did is a transformative to the security business over time. But it's -- in terms of the growth vectors there, the WAF, Web App Firewall, bot management, sort of the core franchise still growing at a very healthy rate. You can't deliver those type of growth rates unless you're getting that type of growth out of that business. So we're seeing very healthy growth out of that business. And the bad guys don't stop. So there's continued innovation. We continue to add value-add to each one of those products. So for example, with Web App Firewall, you're protecting the base page. We added a product called Page Integrity Manager which protects all the third-party content that resides on the website. So that was another area where people were exploiting. So that adds to the tail of that product line. We continue -- we've done bundling and adding in new features and capabilities. We add additional sites with the customer. It's not one and done. That might protect, say, the main page of Bank of America, but you might have hundreds of web applications and I'm starting to get more of those and the pie grows from that. So that business will continue to grow. With bot management, we talked about how that protects against machines, but we also have some fraud products that protect against someone, say, taking your credentials and saying, okay, this is Madeline's credential but I know it's not her. I just saw her login in New York or San Francisco, I'm now seeing somebody attempt to log in, in London with her credentials, so the stolen credential. Might know how you interact with the site, how you type and that sort of stuff, and I can identify that this is most likely not new. So that adds to the TAM of sort of the bot management or fraud use case. There will be more stuff that comes along. So the continued growth there. The big growth for us over the next couple of years will come from micro segmentation. The last time we broke that out, it was a $100 million run rate growing at 60%. We're the market leader there by far. And that's a great security posture that most enterprises don't have today but should because, like I said, things get in, this helps you stop the damage and then API security will be another big growth driver for us. We think that's an estimate to have it a $1 billion growing to beyond that over the next 3 to 5 years.

Maybe just quickly touching, too, on profitability in the security business, right? I think if we look across our coverage, you see kind of a mixed bag of where profitability sits. So could you just give us maybe an overview on your expectations for the security business.

Yes. So the security business for us has been incredibly profitable. We leverage the footprint of the CDN. So if we're doing, say, a denial-of-service attack, we're leveraging the capacity that we already have. There is this enormous operating leverage. We're leveraging the same sales force that sells CDN and compute also sells security. We do have some overlay, but it's relatively small teams there. So there's enormous operating leverage from the security business. When we broke it out at Analyst Day, we talked about having 30-plus percent EBIT margins or operating margins, I think it's a little bit higher than that now. Great free cash flow, very low capital intensity business for us. So that's an extremely profitable business for us, very attractive business.

Great. And the third segment of your business compute. I think that's what personally I've been most excited about. Just can you talk us through compute is 2 different segments for you right now. You've had more of this. I don't want to call it legacy, but I guess more of a legacy business and maybe now your next gen, more distributed...

Yes, exactly. Yes. So we talked about the legacy business. We sort of started with edge compute years and years ago, probably ahead of its time. And with edge compute, it's more of a kind of a lighter weight, we call it functions as a service. So for example, you might be a website that's selling tickets and you have flash crowds that show up and you want to set up a waiting room versus having sites not available or whatever. So it gives a good user experience. And you can run that using JavaScript with a pretty lightweight application at the edge. You may be doing A/B testing for advertising, showing one group one ad, another group another ad, you wanted to do some analytics on that. That was the legacy business. It's a good business. It's growing. But the big money was spent on the real computes or building the application itself. So for example, say, a streaming customer will do a lot of things with their content instead of just delivering. So we do the delivery today, but there's the encoding, the transcoding, content management, recommendation engines, all that stuff, it takes up an enormous amount of storage and compute. And we now have offerings that you can build those applications on Akamai. Now the different differentiation with us is we're going to be much more distributed. I have many more locations that you can actually run full stack applications in than what the hyperscalers do. And it's actually a really good use case that Tom talked about this call or the prior call where a financial institution was trying to comply with Apple Pay and it required 60 milliseconds of latency. And the hyperscalers they couldn't comply with that requirement. Leveraging us and being more distributed, they're able to do that. So we believe that they will carve out a nice little niche for us. We're also more less expensive than the hyperscalers. We don't charge for egress fees, for example. So if you have data that's being accessed frequently, we're a much better option for you. And while we don't have the entire stack of what you can do with some of the hyperscalers, we have a great partner ecosystem. We're going to continue to add to it. We're adding functionality. We think this could be probably the biggest revenue source over time as you look out many years into future.

Maybe just to help everyone quantify this opportunity too, I think Akamai is a perfect use case itself with the savings. So you moved a majority of your cloud spend from hyperscalers internal to this platform, so could you maybe walk us through it and the financials of that as well...

Yes, happy to do that. So what got us into this business, well, we had a few customers that came to us with these distributed use cases that they couldn't comply with the requirements. And so our customers were asking us, "Hey, can I run code, not just the JavaScript-type stuff we talked about, the function as a service, but really run my application in all your locations." And we didn't have that as an offering today. So we did some custom work and we did -- we built out a few things for some customers and realized that this is a business that we need something more repeatable. That's when we ended up doing the acquisition of Linode, but also our use of the third-party clouds was massive. So we were spending over $100 million and that spend was growing 60%, 70% a year. And it is becoming a real big problem for us. So we need to bring that in-house because it's just not scaling. I mean you can imagine within the security business, the amount of data that you look at. It's massive. The amount of processing that you do to determine things like is this the right person or not and it's a machine, whatever, requires a lot of data. So we were able to move -- the majority of that business today is now running on Linode. So we are our first $100 million customer internally and it's been great for us. It's performing better than it was and the hyperscalers in terms of a latency that we would get by having to go back and forth with using a hyperscaler. So it's been a phenomenal success for us. The last time we talked about it, we said we spend about $100 million in CapEx to be able to move that spend, we depreciate that over 6 years, so you can kind of roughly do the math there. But there's an enormous amount of cost avoidance that's going into this as well. So that we're -- those products are growing very quickly so that there's actually more than $100 million of spend that's going to be saved over time. It's been a great success for us.

And what are you seeing in terms of customer demand? When we think about this more distributed layer, is the market at the inflection point where people are coming to you saying, we see what you're doing and we'd like a piece of it too? Or is there still that piece of education as well? And maybe if you could talk about -- this is a big year for -- really one of the first years for growth for this more distributed compute, what are you seeing in the pipe and the interest levels?

Yes. So we're seeing -- in terms of the areas we're winning, we're winning in those more distributed use cases. So somebody who has some performance requirements or wants to run an application closer to where the users are, we are winning a -- lot of the use cases are coming in from that. Also just being cheaper. So a lot of people that are spending a lot on egress are coming to us to eliminate that spend. But what's interesting when we broke this out, we initially thought that most of the demand would come from our media customers, right? The ones that are doing -- spending tens of millions of dollars with the hyperscaler, doing -- just needing raw compute. They have their own encoding or transcoding engines. They just need GPU. We've got some of that, obviously, but we have hundreds of customers across many verticals, which is a much healthier base of customers coming to us now. Observability is a big use case for us. So folks that might be using, say, a Datadog or some type of application like that with a hyperscaler, we have a partner that offers a solution like that. So observability across multiple verticals has been a great success for us. It's just one example. So I've been very impressed with the number of use cases, the contribution from the field in all the regions. As a matter of fact, Asia Pacific is probably the fastest growing group of [ sellers ] for us in terms of getting traction with the customers. So it's been a lot more distributed, if you will, in terms of the verticals and the use cases. So we've been very pleased with that.

Maybe expectations for this year and growth going forward, from your seat and of course, not asking for a guide, but really qualitatively too, do you see this kind of becoming the biggest piece of the business over the next 5 years? Do you think it will really accelerate a lot more quickly?

Yes. So we -- the number to watch is the number we broke out for you on the last call, which we said we have a $50 million ARR business with the enterprise compute. So this is folks that could be using a hyperscaler come to us. That's growing at 300%, so it's growing enormously fast. At some point, if you look at just the size of that market, $300 billion market, $200 billion market growing 15-plus percent, if we just carve out a small percentage of that, it would be a massive home run for us. I think we have the opportunity to do that. Is it 3 years, 5 years, 7 years? I don't know. But I think at some point, this has the potential to be the biggest source of revenue even bigger than security.

And for the investor community, when we think about how to look at demand, I think a lot of us, we tend to look at CapEx, build-out of CapEx as a leading indicator. Is that the right way to look at this business? And how did you preplan for the CapEx [ year ]?

Yes. So it is a pretty good indicator of expected demand. Obviously, there is an initial build you have to do. So last year, we spent quite a bit building out, I think, 13 more data centers, which are much bigger in scale. This year, we'll build out about 100 distributed sites, much, much smaller in terms of capacity. The interesting thing though with the compute business is you have to guess and get the distribution correct. So if I were -- there's a metric you could say maybe $1 of CapEx is $1 of recurring revenue, roughly speaking. But you don't get demand evenly throughout each one of these data centers. So I might have a data center, say, in the D.C. area that's getting much more demand so I had to build up more there. So it's not a perfect metric, but it's a fairly good indicator for us. And obviously, we are our own customers. So some of what we built that was for our own savings, but it is a pretty good indicator. So we've got the capacity to sell a lot more than what we've done so far today.

Maybe just in general, when you think about growth and investment in the business. Akamai has been a very profitable company for a while. It's deeply rooted in your DNA. So how are you thinking about these fast-growing opportunities like security and compute balancing that with your financial profile?

Yes. So we've always said we want to operate the company at a 30% operating margin or better. Obviously, when you're in an investment cycle, you're going to come off of that. I think we've done a really good job of getting back to 30% or better very quickly. I think when I look at the operating leverage, especially in the security market, you have the opportunity to expand margin a few points. Also, once compute gets to scale, it's not quite at scale yet, but when it gets to scale, we think we can deliver very attractive operating margins there. So there's potential over time, if you're not -- especially if you're not in a big investment cycle to improve margins beyond 30% at some point.

Maybe could you just give us a kind of a click down into what those levers are for improving margin?

Yes. So obviously, the third-party cloud savings was pretty significant. There is some inefficiency though in the build-out of our colocation that we talked about where when you enter into some of these long-term leases, you have to make commitments over a 5-, 10-year period and the accounting standards require you to put that on the balance sheet, amortize that cost out over a period of time. So I've got some noncash expenses right now that are putting pressure on my gross margin line. So over time, those will start to kind of go over that cliff and we'll start to get better efficiency of the colo that I have today. So you get some margin expansion and some -- hopefully some gross margin expansion there as well. It's just a timing issue there. So that's one lever. But really, the big thing comes from the scale that we get across our go-to-market engine, we're starting to get a lot more leverage out of the channel today. I think we -- our biggest cost really is people. So people cost is the biggest cost of the business. We've done a good job of driving out our real estate costs. We're basically work-from-home company. We have the option to go to the office but most people are working remotely today. So we've driven out a lot of cost there. So it's really just about getting the scale on revenue for the investments we're making in compute and that should really start to get to the bottom line at some point and expand our margins.

Since we have just a few minutes left, I'll see if anyone in the audience has a question.

These [indiscernible] but talking to potential customers, it seems like there's still perception that limits [indiscernible].

Yes. So pre-Guardicore, I'd absolutely agree with you. So that was -- they used to say segmentation was where CISOs or CIOs would go to die because you had to physically segment your market. There wasn't -- your network. There wasn't a good solution. We have an agent-based solution that can run on -- in any environment on your data centers, on the cloud, et cetera and that scales enormously. Like we're seeing thousands and tens of thousands of agents now that are able to run. So this product scales enormously well. One of the first things that Pavel, our CEO of Guardicore, he's been with us now for about 3 years. One of the core challenges we had and one of the disciplines we have is being able to scale. So the product scales very, very well. And so now you're right, in the traditional sense, it was very hard to do. Now we're seeing massive scale. We have some financial institutions as customers that are getting to very impressive scale. And even the closest competitor in the market we've been winning some deals, takeaways from them because they can't scale, but we've been able to figure that out.

So what's the driver or you said 30% operating margin. Why? Why not taking more risks? And in this economy and in such a large, I would say, cybersecurity companies that are growing much faster so what's the logic behind it?

Well, we had an investor come in and make it that -- we had no choice we had to get to 30% operating margin. But when you put all the pieces together, the company is growing in single digits today, and investors expect from us that type of profitability profile. I think they will trust us if we come down a bit on margins to be able to invest back in the company. But if you're going to make a big change in your margins, you've got to be able to drive significant top line growth. And we've got some challenges with the CDN business, which is 1/3 of our business that's not growing. So it's going to slow down your growth. But also, we're at a scale now, roughly $4 billion as a company that -- we've got the type of business that should be generating significant cash and profit. But I don't think the investors would tolerate it if we were to significantly come off of 30%. If we're point or 2 off, that's probably fine but if we're going to go take our margins in half, for example, they want to see a return right away. And that's very hard to do at this scale.

I think we're out of time. But Ed, it was so great to have a conversation with you today. Thank you so much. And we'll be around if anyone else has questions.

Thank you, Madeline. Really a pleasure to be with you today, and thanks for everyone joining.

Great.