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

Joining us, we have the pleasure today of having Akamai. And from Akamai, we have Adam Karon, GM, COO of the Edge Technology Solutions Group, if that's the full title?

Right. You got it.

And then we also have Rick Eskelsen, Director of IR of Akamai. I would joke around that he's higher title than Barth, but Tom Barth is in the room. So thanks for joining us, everybody. Fairly -- got a group here, so I'll kind of open it up for Q&A if anyone's got any, but I'm going to run through a few questions here.

Adam, you got a new little business here at Akamai, I shouldn't even say little anymore, by buying Linode on the edge compute side of things. And there's a big debate here as to Linode was this SMB-focused asset, and you guys are messaging, no, we're going after enterprise here. And look, Akamai historically has had major logos here, some of the biggest IT spenders in the world. So just trying to understand why is Akamai positioning this as an enterprise-grade solution to go up against AWS and Azure and Google?

Yes. I mean, I think you're right, not a small business, the total compute business that we've got will be over $400 million this year, growing at well over 30%. That's a big business, even in the aggregate at Akamai. But what drove us to buy the company was really a time-to-market approach because we already had demand from customers, and we actually launched VM as a service. We're more bare metal for Apple a couple of years ago that they've been using. And that demand has been across our customer base for our ephemeral workloads that are delivering adjacent. And we knew that the market was quite big. And we didn't have to go head to head, at least in the early days, with the hyperscaler feature by feature to make it enterprise-grade. What we really needed was broad distribution, reliability, scalability and security, which is what -- that's kind of what we're known for. That's what we've got. And so we're really combining Linode together with what we already have in our backbone really was going to give us that enterprise capability right away. And on the other side, I think one of the questions that maybe you'd announced but definitely a lot of people would ask me is, well, Akamai is not very good at getting into the developer space, and developers are becoming more and more decision-makers in the enterprise and buying decisions. And so getting Linode actually, well, you call it SMB, they're very developer-centric. And getting access to the enterprise developer base is something we think Linode will let us do.

Got it. And one thing that's been coming up a lot more, and maybe we blame a competitor in the space that's been with us here this week, is around price increases. You're starting to see cloud vendors raise prices in the space. How are you thinking about the pricing behind Linode, which has been a differentiator as you can get a huge savings versus, say, going to AWS or Azure. But also, how are you thinking about that relative to bundling with some of the other stuff you have in terms of the go-to-market packaging?

Yes. So when you think about the way that a large compute company today like a hyperscaler goes out that has both delivery and compute, they do use delivery in many cases as a [ loss ] leader or as a bundling component. And that is one attractive component for us, which is we can bring those 2 things together as well, which gives us a competitive advantage compared to anybody other than a hyperscaler in the space. And so we are looking at that with our customers. I wouldn't say we have a specific bundle, that's like canned today. But on a custom basis, we're looking at it right now. In terms of the SMB side, I think it's a wait-and-see to see if the price increase that other folks have introduced into the market is a win or a loss for them, and we'll wait it out and kind of see. And again, that's not the primary focus to grow that small base. It's really the enterprise base, which is not going to be where you'd see the price increases that you've seen in other players really impact.

Got it. And you run the delivery side of the business for quite some time now. It's kind of been a little bit of the story this year around delivery and the fact that you guys are taking some meaningful steps in your strategy here. I think last quarter, we talked about no longer going after the peak traffic volume kind of thing. So really, I'm trying to understand -- sorry about that, everyone. How that more disciplined pricing strategy has kind of gone here and especially with some of your larger media customers that are using Akamai and maybe a handful of others.

So I mean, you said it right in that CDNs, in general, make money off of day-to-day traffic. So a consistent volume over time. It's easier to manage. It's more profitable. It's better in general. Better for the networks, too, so the people that you're buying colo from or working within the carrier space, they like that as well. And so yes, the strategy we kind of went forward with is, for a small group of large players in the market that have significant differentials between their peak and their day-to-day. We basically said we're not going to drop pricing for their -- if they want to continue to use us for peak. And we've announced, I think, at last quarter, we did do that with one of our large customers. They're still a very large customer of ours, a great partner, but we're not going to take that peak traffic from them anymore, at least for now. Now they're going to go out and they'll try to stitch together another way to achieve those peaks. But again, we're talking about a small component of the customer base, and they just happen to be these big differentiators in terms of peak versus day-to-day volume.

Is there a way to understand how much that -- how much, I don't want to say of your business, is that peak, direct peak volume.

I think Ed, who -- for those who don't know, Ed McGowan is our CFO. I think the way Ed always talked about it over the years is that -- and I think the question usually comes into how much revenue do events generate for Akamai, right? So it's Super Bowl or an Olympics, and you're talking each one of those may be single-digit millions. So this is not a large portion of the business. I don't think we've ever said a specific percentage. But this is -- I think we've said in terms of events, they're not material drivers, so I think I'd probably use the same terminology here. The peaks are not material. You're usually using a peak or an event to get quid pro quo in day to day. And so that's literally the old way you'd play the market. That wasn't working, which is why we've gone to the, hey, we're going to charge for that. If you want to use our scale, you can't then pick and choose on a day to day. So you've got to make the choice. And on those big customers, the reception has been pretty good.

It's far more impactful if you think about it from a CapEx angle as well than on the revenue side of things because building out that capacity for peak obviously comes with a cost. And if we could spend $1 on CapEx, we'd rather spend $1 in security and compute relative to delivery.

Yes. I'm going to get to security here in a bit. But investors are kind of questioning, why isn't this just further share loss for Akamai? The share -- we've gone towards more of a multi-sourcing approach over the last decade. Why isn't this just further evidence of Akamai losing share? And secondly, on this, like is there any impact from just better compression technology playing a role on the amount of bits kind of going through the Akamai network?

Two different questions, I guess. So I'll answer the first one first. I don't -- I think if you think about those small group of customers that there's peaks themselves, that share may get spread out amongst that. But I think, again, we're not talking about going out to the entire market of OTT platforms, which don't usually use us for peaks, right? They don't have events like that. They're really day-to-day drivers. So I don't see that impacting share on the material of the day-to-day driver traffic on the internet. On the other side, I don't think the 4K, 8K adoption kind of compression technology is really the impactful component today. You have kind of this hangover effect, or whatever you want to call it, from both the launch of a massive amount of OTT platforms just before COVID, and it happened to coincide with COVID, right? And so you have 2 years of just insanity that we're coming out of now. And now the question is, what is the new normal going to look like in 12 months? I think I described it earlier today as like, do we have a temporal issue, which is, for the next year, sometime in that time frame, I don't know the exact time frame, you're going to see kind of the Internet figuring out exactly what the new normal growth rate is going to be. But my expectation is that we'll return back to normalcy. The question is calling the exact date to when. Is it 12 months? Is it 6 months? Is it 18 months? I don't know. But if you take a step back, in 2 years, we're going to be back to normal. It's insane to think that more people aren't coming online. More multiplayer games are not going to come online, and people are going to take advantage of just the normal culture change will continue.

You kind of pre-answered the next question I was going to have for you around -- back 2019, 2018, we also used to reference roughly the Internet growing 25%, 30% a year. And you've seen kind of traffic slowed down generally. Tough compares, right? But is there a sense to when you think that could come back given you did start to have an extra year of traffic in '20, an extra year of traffic roughly in '21 hitting the network? Is it for '23? Or are we talking '24 as normal?

Like I said, it's hard to call the date. But I think you've watched 2 things as kind of things you'd look for ahead of that. So one is, what happens with ad-supported platform launches later this year? And do they have significant impact on subscriber growth and/or engagement time on the platforms? And I think you'll see folks like Netflix starting to talk about engagement time because they have maxed out subscribers, and now the game is, how do I get people to watch longer? Both are wins for Akamai, right? More subscribers will have increased viewing time. More engagement time as we increase viewing time, so we win on both either. So as you see the OTTs talk about that, you would kind of have a pre kind of announcement to what you'll see from us into the future. And then it's gaming, recovery in the gaming space is kind of that's headed. So it will recover. I don't know when, but I know that when I start to see those things happening probably with my kids, then I'll know.

Yes. I'm sure it'll be annual holiday content releases. We'll see how that goes.

I hope so. Yes.

Back in, I think it was May, end of May, you guys hosted an Analyst Day for this segment of the business. We talked about it as down, forward or roughly flat from a revenue perspective.

That's right.

I mean, with these challenges now with traffic, what makes you confident that we can get back to that kind of range, at least medium term? And is that range even possible for -- I know you guys don't want to guide to it, but like how is that possible for '23 even?

Well, I think it's over a long term, right? So again, that's -- and we've thought about when to call it. I don't think we can call the date. But I do -- again, kind of going back to just the normal secular trends in the industry, more people are coming online. We're nowhere close to penetrating large markets. Latin America is just starting. Bit rates are still increasing in India. So there's still tremendous amount of traffic to come online. It's just calling the timing of that. So I think getting there over the time frame we described, I think super comfortable there. Does it happen in '23? I'll let you know in Q3 next year, yes.

Got you. So obviously, this business is more about how do we transform more into security and compute now. But from the delivery perspective, is there further ways you can actually optimize either the network or how you go to market or the approach here to drive better profitability or more of a cash cow kind of business for delivery?

Yes. Yes. And I mean we spend a lot of our time -- one reason why we're one of the only profitable companies in this space because this is where we spend our time making that business as profitable as possible. And it never ends, whether it's through better software, better network deployment. And I think with the compute side of this now, the leveraging of the underlying network for compute because you need that broad distribution, and the distributed compute component of the acquisition is really the differentiator we're going to have in the market. That's what we're hearing from our customers. That's what they want. And so bringing those together will allow us to like deflect some of that cost into compute, meaning we'll be building the network for compute not for delivery. And so you'll see a reduction in network build CapEx, probably network software build for that as well on the delivery side as we kind of migrate over to compute using the egress capacity we have on delivery really to fill delivery-adjacent compute opportunities. So I think you'll see both things happening simultaneously, shifting of cost and increase in capacity on the compute side going along with it.

Got it. And so last one on the delivery side for me. Zero overage, it was something we talked about, I think, June of '19, I want to say.

That's right.

Those tended to be multiyear deals. It's not something we hear a lot from you guys anymore. How has that zero overage transition for the web delivery customers gone? And why has that resulted in Akamai being argued as more of a competitive price leader?

Yes. Well, I don't know that we're not necessarily in that space, argue that way with the combination of security along with delivery. But I think as we've renewed a lot of those off deals over time, some of the customers, we've removed it off, some of the ones we've changed. So I think we kind of -- the reason you don't hear about it as much as we went with this one-size-fits-all ZOFF approach, every customer gets ZOFF. And as with any go-to-market, that's usually not the right approach. And so we've gone with a much more targeted approach with different pricing models because ZOFF was good for some customers, not good for other ones. And so we've gone with different approaches. And so that's what you'll see now, which is a much more customer approach based on markets. So you think about e-commerce versus travel versus a health care or manufacturing, they need different things, and they don't have spiked e-traffic like a commerce does during the holiday season, which what drove the ZOFF construct even just to start, right?

Yes. Circling back on the edge compute side, I thought one of the probably most important slides at that May Analyst Day was around the Linode product road map or the edge compute product road map. Happy that you're here.

Three buckets?

Yes. So what has you most excited? Or what's the most paramount to get out the fastest in order to make that enterprise-grade solution?

I mean if you look back at that slide, so if it was up here, you'd have the left-hand side, which is our core, and then the mid-tier. I guess get the expected answer, they're both super important. We absolutely have to build -- a lot of the dollars are today are in core. And so the initial workloads coming on are core specific, so getting to the 25 to 30 locations that we need for core early next year, super critical for us. The distributed 50, 100 locations, I think as described before, for a couple of other customers we built out, we had built out already a distributed computing network. It just wasn't multi-tenant or ready to use for many, many customers. So we'll still use that. For now, they get that broader distribution. Sometime later next year, you'll see more of the distributed, but they're both important. One's a differentiator, one's where the dollars are, and we need them both to make a big mark in the space by '24.

Yes. Well, we can't have a presentation from Akamai without talking some cybersecurity. I know that Mani is your counterpart, and I know you work well with him. So big thing out of last quarter, too, was what's going on with the security business that we saw, that drastic of a slowdown on a normalized basis, backing out M&A and FX. And it was talked about around more of the infrastructure services. So you've heard some strength from others in the space around DDoS even still being a driver of good revenue for them. So what's going on with that infrastructure side of the business or security overall that we're seeing the slowdown, and how confident do you guys feel about keeping that 20% sustainable growth rate?

I'll start with we're staying with the 20%, right, with -- that will be constant currency with M&A. And you'll see us still continuing that into the future. And I think that when you think about the size of the business, it's quite large. And so the law of large numbers gets with us here as well, which is why you kind of see some of that kind of teens numbers coming out on the security side. If you think about infrastructure, which was really our Prolexic side of the business, the majority of it sits there, so layer 3 type of attack surface kind of stuff. That side of the business is attack-driven type of activity, right? So when attacks get very high on that side, you tend to see an increase in that business. When attacks decline, you tend to see a reduction there. One of the things that we have been doing is some of the attacks were getting large in geographies that we hadn't expanded scrubbing centers to. So you see a large expansion of scrubbing centers going on around Prolexic to take more potential attack customers on the platform for Prolexic. So I think you'll see some of the strengthening in that infrastructure space. But again, kind of sticking right inside of that 20% range constant currency with M&A.

Right, right. If anyone has a question, feel free to raise your hand. I did want to ask around the go-to-market side of security because it's kind of been this big question point where it's kind of tough, right, because Akamai traditionally, you have the biggest customers out there to begin with. You do have an MSP program on the security side. But when we think about security sales, they traditionally sold through a VAR or a distributor. So how are you guys thinking about that, especially as one of your main competitors now has introduced this new partner program that brings in some of those traditional VARs and distributors?

Yes. I mean we've struggled in the past as a company in making the channel "flywheel" run, right. Whether it's reseller, SI, VARs, they're -- we've had some troubles with it. And a lot of it deals with the accessibility of the product to those channel and partners to make money around our products at ease of configuration, those types of things. The great thing was with something like Guardicore and the remake to our enterprise, our Zero Trust kind of product line, it opens up that space significantly for us, and we're seeing some traction even early on with Guardicore to do that, where partners are wanting to come in, take advantage of really a leader -- a leading product in the space of micro-segmentation and do that. We'll see the same thing on compute as well with SIs, VARs really wanted to take advantage of that as well where they can really wrap their services around it and actually make money as opposed to being a pure resale motion that we've kind of had with some of our older CDN more related product set. So I think you're going to see a lot more announcements from us in the space of partner SI kind of flywheel generation in the States and outside.

Got it. Looking forward to those.

Well, and our -- just to say one more thing there is our head of sales now, who used to run sales when we split the company for me, but now took on all of sales for Akamai, and he comes from a pure channel background. And so putting him in charge of sales was another kind of statement by Tom, our CEO, just to say, "Hey, we want the sales motion to be partner channel-focused moving forward."

And you brought up Guardicore. And I thought that was a home run in terms of the strategic positioning, and it's done fairly well underneath you guys. And when you think about you guys in the ZTNA space versus some of the others out there, it's one thing that you guys definitely check the box and the others can't. So how are you feeling about Guardicore getting that spear into that Zero Trust kind of sale? And what other differentiations do you think you guys have on the Zero Trust side that will compete against the names that are getting attention, the cloud players, the Palos, the Zscalers?

Yes. Well, first kind of goes back to the previous question you just asked, Guardicore has been hugely successful. We're going to do over $16 million this year, growing at a really rapid pace beyond what we had even thought we'd grow at. And I think the great thing about that success is it brings on partners who then say, "Oh, great, I want to bring this in as part of a complete Zero Trust solution." And so that's going to drag along our other Zero Trust product set that we've had fits and starts in driving traction there in terms of our access product or suite product set. And I think bringing that all together with those channel partners will help. The other thing we've done is consolidate that entire product line under leadership that came from Guardicore to produce a single like combined product line with a real strategy across all of our Zero Trust products under one leader. They're going to have a combined or a single pane of glass approach with a consolidated, unified client that will launch in Q1. And when you think about all that together, we're going to have a much better sales motion with our customer base and with our partners to go in and attack the whole enterprise space, making us more competitive against the few folks that you're probably thinking of when you say the other folks in the space.

Right. So just broadly speaking, whether it's any of the 3 buckets or even bundled deal together, you're hearing a lot about are you seeing elongating sales cycles. Is this something you guys are seeing? How are you guys trying to combat it? Is it an approach where, like, hey, let's just -- we have the delivery side of things, let's just try to land the core security, and then maybe next year, we'll talk about Zero Trust? Or how are you guys approaching the kind of elongating sales cycles, if you're seeing any?

We're not really seeing that yet, right? So at least we haven't yet, either in the -- definitely not in the security space, for sure. And I think people aren't willing to take risks in the security space, they need to protect their infrastructure. And I think, give Tom Barth a little credit, his line is always like, "Nobody wants to be the one that gets fired for not spending the money on security even in an elongating cycle like this." But we haven't seen it in compute. I think the great part about where we're going to play in the compute lens is there's a cost savings play to move to us. And so even what you saw in the news last week, I think it was Netflix announcing they're going to go undergo with third-party cloud kind of savings approach. And you're going to see that as other companies come under pressure, it will be good for Akamai because people will use us for that in that space. So not yet.

We're watching Europe pretty closely. I mean, I think like many other companies you probably heard from, I think there's concern about the economic outlook for Europe as you go into the winter months with the energy prices and things like that. So not seeing it, but definitely watching, especially Europe, in particular.

Yes. Just have a couple of minutes left here. Akamai has done a couple of larger deals. In fact, 2 of the largest ever within a very short time span. I know there's going to be some digestion period going on, but how are you guys thinking about that M&A, which Tom Leighton has always talked about as complementing or supplementing some of our R&D investments.

Yes. I mean it was unique, as you described, to do 2 very large deals simultaneously almost at last year. I think we're going to continue the same acquisitive approach that we've had in the past, opportunistically looking for both enhancements to the security product line where it makes sense as well as expanding potentially past services on the compute side of the house, or even infrastructure if we see those out there. But I don't think you'll see us do anything of quite that size in the near term at all.

Are there areas of -- especially your side of the house, that you guys feel there might be a hole in the portfolio that could be an area for inorganic?

Yes. I mean it's hard to say. I mean, the good thing about the platform is that it came with a lot of really great bones to build on. And so a lot of stuff can be done organically. But again, to me, time to market is really important. So if opportunistically, something shows up and it can accelerate our time to market, we're going to go after it.

Makes sense. Any questions out there? Got time for one last one. Otherwise, I can squeeze mine in. Something we're trying to do at the conference here is try to understand automation and the way that it's playing a role. So internally and externally, how is Akamai playing in this sort of automation of infrastructure.

Well, I mean part of our general COGS cost savings approach over the years has been to continually automate away activities. And I think one of the things that you'll hear Tom talk about a lot is just the small number of people that actually manage and run our robust many thousand server network. And you look at our NOC at any given time, maybe it has 8 people in it, but it's running the entire network. And I think that type of automation is what allows us to continue to be significantly profitable in the space, and we'll continue to do that. In terms of external, I do think that the -- our ability to get really close to the edge has been something we've talked about more in relation to IoT devices, but you can imagine that coming into play as you need to communicate to devices that are doing robotics or automation where they're going to need to communicate and drive compute, but maybe have decision-making that needs to be much closer to the device that's operating itself. And that's been a challenge if you need to backhaul a decision and data all the way back to some central infrastructure to make that decision. It kind of changes what you can and can't do in terms of automating, let's say, an emergency response on a train or in a car or any type -- anything like that. So I do think that you'll see that more as a future edge case, probably in the next few years in automation. But near term, all that is really internal.

Got it. Well, I think you timed it pretty well, right?

Six seconds. Yes.

And so I appreciate everybody for joining, but especially to the Akamai team here, Adam and Rick. Thank you guys.

Thanks for having us.

Thanks, Jim.

Thanks, everyone.

Thank you.