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

Okay. Good morning, everyone. I am Mark Murphy, software analyst with JPMorgan. And it is a great pleasure to be here with Ed McGowan, who is the CFO of Akamai. So first off, Ed, can't thank you enough. It's very generous of you to take the time out of your schedule and be here with us.

We're happy to be here. Thanks for inviting us.

Maybe we could begin, Akamai is a big company, and I'm sure most people have heard of it, but maybe you could kind of give us the 30-second overview, just for the benefit of anyone out there who is more of a generalist.

Yes, sure. So Akamai Technologies, we've been around since 1999, went public back in October '99. About $3.8 billion in revenue, roughly, a very profitable company. We're in 3 lines of business. We started off in what's called the CDN business. So you probably interact with us many times a day, and you don't even know you're doing it. We carry about 1/3 of the world's internet traffic. We make the internet work basically. We also now our largest source of revenue is in security, primarily related to application security. So as we're delivering those websites and web applications, we're protecting them from all the bad things that can happen. We also have a smaller, but faster-growing business with enterprise security, some more traditional IT security. And most recently, our newest area is in the compute business. So think of that as now you can build your application, you can secure it, and you can run it and deliver it on Akamai.

So that's a great overview. And that evolution is kind of exactly where I wanted to start. So we think back to kind of the very late '90s, Akamai got to start in delivery, you built a very large security business, and I think you just mentioned that, it became your biggest business...

Last quarter.

Last quarter by revenue. And now you're leaning very heavily, very aggressively into compute as well. Can you help us connect the dots on. So this is the evolution that's where you're heading. And then what are the synergies that we should be seeing between delivery and security and compute.

Yes, sure. So originally, we started off solving a problem where the internet itself is not built for performance. The general routing protocols don't take performance into consideration. So we built this massive overlay network on top of the internet. We're distributed into over 4,000 locations. We basically put our servers close to where the users are. So where you connect to your local ISP, we have a server that's either in that ISP or very close to it. So that when you request content and it's one of our customers, you get served as close to the end user as possible. And the reason for that is if you think about there's massive amounts of capacity at your home, a lot of people have fiber to their home. And the choke point is the core of the internet. And so by building this distributed platform, we were able to get around that issue and be able to deliver superior performance. And it also gives you an enormous amount of scale. You can defend cyber attacks from where they're emanating from and shut them down and not actually bring the data to the data center, and that's the worse thing you could do. But basically, what -- that first evolution of the business, we became the largest CDN in the world. And if you think about the CDN business, there's a lot of synergies with security. First of all, we see every Internet user many times a day, as I mentioned. We're also the largest provider of recursive DNS, which DNS is essentially the on-ramp to the internet. So we have an enormous amount of data, and we moved from the CDN business into security was a natural evolution. You leverage the data, you leverage the platform. As I talked about, we handle a lot of denial of service attacks, where you have massive amounts of infected devices that will go and attack a site, try to take it down. But we can block that site -- that attack, right where the attack is coming from. And the same server that might be delivering a video to Mark's home is blocking an attack from his neighbors house. So we were able to leverage that same platform and build the technology that delivers the security as well as the content. So for example, one of our largest security offerings is our web application firewall. So it's protecting websites and web applications from all these various things that can happen, someone taking over your site. So I'm going to trying to steal information from it. We deliver those web application firewall rules. At the same time, we're delivering content. So we don't slow down the performance as we're providing that security. And then when it comes to compute, we've always been in what's called the edge compute business. And what do I mean by that? That essentially is using Java and technologies like that to do things on the Internet at the edge of the Internet that are sort of more functions of a service. So for example, personalization or image optimization or maybe you want to run an A/B test, you're doing advertising. You want to send show one group of people on ad, different group on another ad, dynamically at the edge, understand which one performed better. So sort of lighter weight compute, if you will. Maybe it's a waiting room, you're selling tickets, so you've got a -- so a big event and people are all coming and you don't want to overload the log-in server at the data center, so you just put a nice screen up within a waiting room and that sort of stuff. So we kind of lighter weight stuff. Now we're getting into actually building the application. So VMs as a service, containers as a service, block storage, et cetera. So competing with the big hyperscalers. That's the next evolution of us.

Okay. That's a fantastic overview. If I ask you this way, what do you perceive as the kind of secret sauce for Akamai, kind of across all those layers and segmentations, is it -- is there a differentiation in the architecture of the network? Is there differentiated IP or patents? Is it the sheer number of points of presence because we're definitely aware that Akamai kind of eclipses everything else, right, in terms of the points of presence. So what is it that you think would -- what is insulating Akamai or kind of making it difficult to replicate?

Yes. Good question. So I would say from a technical standpoint, absolutely the architecture. So if you think about all the things we talked about, whether you're trying to deliver a big web event or you're delivering an e-commerce experience. Performance does matter. And the closer you are to the user, the better performance, and that has played out over the last 20 years. And from a CDN perspective, we are by far the largest, but it's also leveraging the technologies that we have and the data that we have, and to be able to build off that platform and continue to drive services that have higher margin on top. So I think that plus the data science, just the amount of information that we have is greater than pretty much anybody out there in the market. But I also think from a business perspective, we have some pretty big moats. And what I mean by that is the CDN business is about scale. So we were the first ones to get to massive scale. We're the only profitable CDN out there today. We've been able to build higher-margin products on top. So we've got a very profitable business model. We've got great customer relationships. We've got great people. So I think it's a combination of all that, that we were sort of the first ones to figure out that. There's some challenges with the -- in the CDN market in terms of pricing and things like that. You have to innovate and continue to stay ahead and find ways to use that scale and synergy, if you will, to continue to build your business and get into higher margin, higher value products.

Okay. So there's the architecture, there's the data science. There's the fact that you've got the profitability kind of embedded into the business model.

That's right.

Okay. If we double-click for a moment on the delivery business, that -- and we look at the contraction there, that has actually moderated a bit. If I look at it for the last 2 quarters. So it was 11%, I tried to put it in constant currency. It was 11% contraction and then it moved to about 8% to 9% contraction. What was interesting was that you pointed to an uptick in traffic late in Q1 and also with more favorable pricing.

That's right.

So it felt like there were 2 good things happening there. Can you help us understand the dynamics that would be kind of driving an improvement in pricing and volume? And is it -- I mean, should we be -- are we calling a bottom? Is it too early to call a bottom? Is it a local trough, something like that?

Yes, good question. Always tough to call a bottom in the CDN business. But -- so there's always -- for the 20-plus years I've been at the company, traffic has gone up into the right. It's always grown. The Internet has grown generally 30% a year or more. And prices have come down. So there's a very tight correlation between unit economics, volume and price, and that has stayed consistent. We had a bit of a super cycle when it came to the delivery business during the pandemic. You had 3 things happen. You had a pandemic. So people are at home, spending a lot more time online. You had 4 major OTT services that came out, and you also had a console cycle, along with sort of a new gaming phenomenon, on this multiplayer gaming like Fortnite.

So it was overheated.

Overheated. So we saw massive growth in traffic in '20. And I would have -- if I was on the stage at the end of '20, I would say I would expect to see '21 traffic decline a bit or may at least not grow at the same rate, I should say. We actually saw a very strong '21 in terms of traffic. It wasn't until March of last year that we started to see traffic really drop off the growth rate. It's still growing, but at, call it, half of what it normally does. We've gotten to the point where we've overlapped that, and we're starting to see our anniversary that. We're starting to see traffic grow again. It's growing a bit faster than it was last year, not quite what it normally does, but it is an encouraging sign. And on the pricing side, and I talked about this several years ago at an event that as volumes start to moderate, pricing declines will moderate. So prices still come down because volumes are still going up, but it's moderating. So it's not nearly as bad as the renewals I was doing in '20. The challenge with the CDN business is you've got big customers that -- and the good news for Akamai is we're very -- we don't have a lot of customer concentration risk. We only have about 8 customers or 10 customers that are 1% of revenue or greater. No customers that are 10%. So when those customers all renew, you can have some disruption. So it's -- you'll always have some variability in the CDN business, especially when you have concentrated renewals with large customers. That's why it's hard to call a bottom. But the dynamics in the business are better than they were last year, certainly.

Okay. It's good to hear. What about gaming? I think you -- it was the first thing you mentioned actually when you were talking about the pandemic, and I know that over-the-top is another one I want to come back to, but is there something happening in gaming right now that gives you comfort?

Not yet. Gaming is sort of almost like a fashion, right? It's -- if you have a popular title, like when Fortnite came out, that was a massive boom for the gaming industry. And then you had a lot of me-toos that came up with similar type games. It really is about the development cycle and the popularity of games, and that's not something that we can control. So when you start to see, like there wasn't any major releases last year it was kind of a soft year, that tends to be pretty cyclical, not calendar cyclical, but cyclical in terms of there's times when there's new games that come out that are really popular and there's lots of them. And there's times when you're in sort of a lull, and we had been in that period. Hopefully, things -- I mean we're due for that to turn around at some point.

Yes. Okay. Now what if we think about media delivery and e-commerce The e-commerce obviously went through this major pandemic gyration, it had the spike, you can pull up a stock chart of Shopify and BigCommerce and see that. And then the connected TV providers have had some fits and starts as well. What are you seeing behaviorally from -- I mean, if you can comment on both of this. I think we're just wondering what is the behavior like where you -- if you think across everything, Disney+ and Amazon and HBO and TikTok and even Apple. How does that part of it feel?

Yes. So I won't get into individual names, but just sort of at a high level. We'll start with commerce. So with commerce, while we had a boom in traffic for the commerce business during the pandemic, we had a lot of economic challenges. And one of the things I talked about that I think is the differentiator is our relationship with our customers. And we actually work with our customers, restructured a bunch of contracts, and I called all this out on the earnings call, so you can see if you want to go back and look. But we restructured contracts, we extended some payment terms did sort of the right thing by the customer to maintain that relationship. We get out of that, things are good. Customers are buying more security, which is great security business is growing. Now we're in a recession. So that business is going through another challenging environment, just from an economic standpoint, you're starting to see a few bankruptcies pop up and things like that. So that business is always really more tied to the economy than anything else. When a good economy, we tend to do better in e-commerce, bad economy, we tend to do a little bit worse. On the media side, we saw, like I said, a bit of a super cycle. You had 4 major content releases between the major OTT providers. And that just generates an enormous amount of traffic. And then what happens is that you start to see them go into different countries, right? So you start off big launch in the U.S. and Europe, then you go across Asia, across Latin America. So we're seeing some expansion, geographic expansion with the providers. We are starting to see an interesting phenomenon now with a few. We called this out on our last call that a lot of these providers will go with multiple CDNs. Okay? So it's fairly easy to build a load balancer that essentially will sign out users as they come to their site. And it can be as sophisticated as measuring performance to as easy as just round robin or just set a percentage, 10% goes to this one, 50% goes to that one. We're starting to see the number of vendors consolidate, because all of us generally will have volume-tiered pricing. So if you want to get the lowest unit rate, you consolidate volume, you get the lowest unit rate. So we're starting to see a little bit of that, which is a good trend for us because we have the highest performing CDN, we get usually the biggest share in those accounts. So that's a positive trend for us. That's also adding to some of the traffic growth.

When did that start happening Actually, a narrowing of the list of CDN providers.

I would say that's been a phenomenon over the last 5 or 6 months.

Interesting. Okay. And so -- and just to be clear on this, so you're not really -- you're not seeing the e-commerce vector pick up right now. You're saying that's economically sensitive...

Yes, check it out it's barely that...

It's a little choppier...

Yes, it tends to be seasonal. Q4 is our biggest season, obviously. Really what the impact there is more on pricing. So you tend to get more pricing pressure in a downturn than you do when things are good. So that's more of a sort of a pricing dynamic issue. This is generally a slow season for retail, e-commerce type customers. Yes, you get graduations Father's Day, but really the big one is the, end of the year holiday season.

Yes. Okay. How do you feel about the actual network capacity of your global CDN infrastructure. What I mean by that is when we think about this huge growth trend that you have, on over-the-top or what we think of as connected TV types of platforms. In online gaming, right, from a long-term perspective, is it more of an opportunity to grow? Or is it more of a hurdle in terms of -- you may have to -- you have to avoid maybe overloading in your own network during peak periods.

Yes. Good question. We've actually changed our posture a little bit. There's 2 things that drive peak. One is a major live event, so a big sporting event think Super Bowl, election, sometimes...

Will be big...

World cup. Yes, exactly. And then the other one would be a big software update. So tens of millions of users need a patch for an OS or for OS update or from Microsoft patch. Those -- or game, say, Fortnite, a big update for that. Those tend to drive your biggest peak. So what we've said is we are no longer going to chase the peaks, and we are now saying no to certain things. So it used to be someone comes to me and say, I have the rights to ex Super Bowl, World Cup. I need why in capacity. And I say, I'm not going to chase that. I'll give you what I got. This is all I have, right? And that really is what drives your CapEx. And part of the reason we have changed our posture is we used to grow so fast that the traffic that is, that if I could -- if I spent to build out for, say, a World Cup, within a year, I'd be able to fill that in. Now it's taking me longer. So it's just simple economics, and we're just saying we're no longer going to play that game. If you want to put an update out, maybe it takes you 3 days instead of 1, right, or you go to other providers to find it. So not economical as it was in the past.