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

Good afternoon. My name is Colby Synesael, and I'm the communications infrastructure analyst here at Cowen. Welcome to day 3 of the Virtual Cowen TMT Conference. For this fireside chat, we have Akamai. And from Akamai, we have Dane Walther, who's the company's SVP and GM of their Media and Carrier business unit. And then, we also have Tom Barth and Rick from IR, who will also be participating in Q&A. This is set up for 30 minutes of a fireside chat. I'd point out there is the opportunity to ask questions at the bottom of your screen. And I have a separate screen set up, and I'd be happy to read those off to the extent I get any.

With that, we'll start. Thanks, guys, for all being here. Really appreciate it. I guess the first question deals with COVID 19. So as noted on your first quarter '20 earnings call, traffic was up about 30% in the first quarter. And I think part of that was driven by COVID-19. Can you give us a sense where you saw the most notable step up? What were some of the themes, if you will, that kind of drove that?

Yes. I mean, I think it may be sort of natural for folks to think of it, but the places where we saw the biggest traffic increases are things you'd expect when people are home or quarantining at home, a lot more consumption of media across the board, right, whether it's VOD, all the direct-to-consumer launches that are happening and sort of still tricking out while people are at home as well as gaming. I mean gaming always has driven big spikes of traffic, especially since they try to have concurrent game releases. And I think as you can imagine, with all the kids sort of home from school across the world, there's not only a higher demand, but I think you see -- if you look at that industry, they're also trying to get more releases out and trying to do new interesting things that are also driving a bunch more traffic.

Okay. So really both, not flagging one versus the other?

No. I think it would be difficult to flag one versus the other. I mean, in some sense, it's -- I was going to say business as usual, not in the sense of traffic levels, but depending on popularity, either it's a media piece of content or again, like a new launch or launch of a video game, they both tend to drive a lot of traffic, and we're continuing to see that just at higher levels.

And then shifting to the current quarter. This uptick in traffic was largely expected to continue, albeit with some moderation as we enter the summer months, such as June. Can you give us a sense whether you're seeing that traffic begin to normalize?

Yes, maybe I'll take that one. Again I think nothing has really changed in the months of April and May in terms of the market economy in general on this global pandemic. And so basically, what we saw in the months of April and May were sort of what you would have expected out of -- coming out of March, right? Now as we move into June, you're all seeing the same news as we are as people are dying to get outside, and you're seeing a little bit more of that. And I would say that there was a little bit of anticipation of all those factors in our guidance.

Okay. And then COVID-19 has had the largest negative impact on Akamai's Web Division due to your exposure to retail, travel and commerce customers that have been hit particularly hard by the state at home environment. There's also been a number of bankruptcies already, for example, in those brick-and-mortar areas such as Neiman Marcus, J. Crew, maybe even Lord & Taylor. And at the same time, it appears likely that the travel industry will face continued headwinds. I guess, Tom, how should we read into this as it relates to Akamai?

Yes. So it's sort of the things that we prognosticated, right, in that Q1 earnings call is that while we do see some real benefits of online businesses that are not associated with travel and leisure, but even in e-commerce, there are some benefits that you're seeing in terms of people like myself buying more online or at least shopping online. So there's no question that travel and leisure continues to be under duress. I mean there's just -- until people start traveling, that's not going to change, right? Now in terms of e-commerce, there's a couple of companies, as you mentioned, that are having challenges dealing with the brick-and-mortar world and that parlays into their business world, right? So they're filing for bankruptcy. And I do think that more and the more fungible type of products that you can go buy are doing fine on -- we have plenty of commerce folks where traffic is up in the months of April and May as well as March, but you do have sort of these high-end retailers that not only is their brick-and-mortar at risk, but people aren't going to be buying high-end luxury items online. That's not the way people buy for coats or expensive watches and that sort of thing. So it will be interesting to see how that works. Now again, we have -- we talked a little bit about bad debt. Some of those folks have filed for bankruptcy or forgiven potentially some of that debt. But a company like J.Crew is going to do fine online, frankly, buying khaki's rather fungible. So -- but I think in general, when we look at what that means for Akamai's longer term future, that should be good for our business, right, because they're going to need to rely on the people that provide the best web acceleration, and that's where Akamai is...

It sounds like implied in your answer is that while there's some negatives to that portion of the business, Web Division, because there are also some other parts that are positively impacted by COVID-19, and perhaps there's more of an offsetting nature to all of this.

Yes. Yes, I'd say that's probably on a holistic level true, but then you also have -- I read a very important article about the automotive industry and why people aren't renting cars because nobody is traveling. And the auto industry is very reliant on selling cars to the rental car agencies. So there's a lot of things that still haven't really played out in terms of a longer-term impact on the economy. But in general, I would sit back and say, yes, I'm pretty bullish on the trends of people working from home and buying things online and all these other trends that are becoming of it.

So I want to shift back to Dane in a moment, but let me just finish this off for a second, which is you guys chose to rescind your fiscal '20 guidance in the first quarter due to uncertainty caused by the global pandemic. Has visibility begun to improve? And what's the likelihood that you'll be able to reintroduce 2020 guidance on your second quarter earnings call?

Yes. So I'd call -- I put all this in the forward-looking statement sort of category. But again, I think it's still too early to call, right? I mean everything at least in Massachusetts, everything starts opening up a little bit more in earnest next week. So that means the month of June will be kind of more of people trying to get out and about. And the fact of the matter is we have no idea what's going to happen, right? We have some indications from Georgia and Florida and things like that. But we don't know if there's going to be a second wave. We don't know if this is a virus that just works itself out, and it dies on the line, so to speak. Actually, but it's a terrible use of word. But I think, in general, we will really see what happens. And we don't report until July. And we probably will know a lot more by July about how this all plays out in the current environment.

Okay. Fair enough. So shifting gears, Dane. HBO Max launched yesterday, with Peacock scheduled to follow early in the, I think, third quarter, combined with the Disney+ launch, I think investors have viewed this collectively as the material driver of growth for companies like Akamai. Can you give us a framework to think about the magnitude or size of impact these launches could have on a company like Akamai? I mean how should we start to think about this?

I mean, I think it's a little hard to put a specific number on it, but I think you hit the nail on the head of the opportunity, right? So today, a lot of this content, especially with these, I'll say mature, but I mean that in a good way, launches like Peacock and HBO Max, where is a very large popular library, a lot of that content is already being consumed today. And so perhaps you could project out how much of that content is being watched today over more standard methods, whether it's cable or even you could say Netflix, right, because that would be a place where we could potentially pull traffic into a CDM from and sort of map that out to the size of the opportunity. At the same time, it really does depend on adoption, right? So I'm pretty bullish on the fact that these are large popular libraries that have -- especially on any of the sides, right? On any of the 3 you mentioned, they've got some real blockbusters that people want to consume and consume a lot all the time. They've also got a pipeline of new content coming in. To me, the question is really going to be how successful are they in getting people to migrate to that direct-to-consumer piece of it. But at the same time, from a CDN perspective, we don't necessarily only just have a contract on the direct-to-consumer side. We've also got contracts to deliver some of that content in the more traditional message that is happening today. In terms of OTT, right, as opposed to just like a direct-to-consumer offering from those companies. So I'm pretty bullish. Like, it's hard for me to put a number on that. But I think you could think about it as these are really popular pieces of content. You could even look at stats that maybe Netflix might have reported for content that's now going to be in a direct-to-consumer and sort of maybe use that as a way to model upside opportunity for migrating over?

Maybe to make it really simple. Is there a likelihood or is it highly likely that these launches in aggregate are material to a company even of Akamai's size?

I mean material, I guess, the threshold of that, maybe I'll leave to Tom to sort of think through. I think that it can certainly have an impact for the media side of our business, but that business is also pretty big, to your point. And so, yes, I do think there is an upside there that could help us drive the business certainly faster. But I wouldn't say that we're reliant on that to drive to the numbers that we're already sort of looking at. And there's again, depending on how popular and how much they get adopted, there's definitely a lot of opportunity there that I'm pretty bullish on.

Yes, it's interesting. It's not like a -- we haven't seen it yet where there are step functions, right? There's sort of this gradual curve of growth. And when you think about events or even someone like Disney, what they're seeing is a very positive trends of growth that are very valuable to not only Akamai but the CDN industry. And so I think that's going to continue, and we didn't even really even talk about mobile launches or international launches. So there's a lot of different factors that continue to drive growth, but it feels like it's -- frankly, for us, it's a lot better to have this manageable growth because we have to invest in capacity and staffing against these things. So it's actually better for us not to have Disney rollout in 80 countries around the world and bring in 50 million subscribers. Although we could handle it if we had enough to have noticed that advance in capacity planning and all these things. But for us to have -- obviously, we work with these folks, 3, 6, 1 months a year in advance, of anticipation of their expectations. And so for us, it's a little bit easier to manage this growth curve that we've seen, which continues to be very favorable for us as well as other CDNS.

Okay. And I guess on the first quarter earnings call, it was noted that some customers have a tiered pricing structure built into their contracts. I guess more broadly, can you explain how you charge for companies like this? I don't want -- I mean, I know there's some variations, but maybe some of the most commonalities. Just so to the extent investors want to start thinking about how does the money actually come into the door? How do we think about the -- how you charge for stuff like this?

I mean I think in general, I mean, there are a multitude of different ways, right, depending on the companies. But the most prevalent is on traffic delivered, right? There are different flavors there. I mean you could talk about committed amount to capacity or people want to make sure that they can lock in that they would have enough traffic for their projections. But ultimately, at the end of the day, I mean, from a media perspective, it pretty much boils down to the usage and then as part of those negotiations very often, like you said, it's a very much sort of a tax table. So as usage goes up, the price points go down, and you sort of tie that together so that we can be jointly successful with our customers around the world, especially the global ones, right? There, you can also have the potential for different rate cards and different tax tables in different geographies as well, right? Cost profiles are different, usages can be different. You could potentially have a blended one. I mean there are a couple of different flavors of how customers want to consume those sorts of tax tables. But in general it's usage based and they're paying for it based on how much consumption their customers are using of their content through [indiscernible].

Is it true usage? Or is it based on peak usage?

It's bits delivered. It's bits delivered. So you would pay for something -- so as opposed to your sort of -- I'll sort of think in my head of what you might mean by like a classical network contract might be like a 95.5%, right? And so that would be like a use, that's like sort of the peak, we'll say it's 95th percentile anyway. We used to sell that way. But these days, it's really the bits delivered. So you pay for the petabytes that are delivered downstream to your end users, not necessarily for those peaks. Although there are -- we might still have some contracts that are structured like that for various reasons. But for the most part, when I say they're paying for the content delivered to their customers, it really is the bits delivered, not based on sort of peaks like a more classical maybe network contract might be.

Got it. You answered my non-technical question exactly how I was hoping you would. And I guess to that point, though, on the flip side, one of your biggest costs to deliver this traffic is your own purchase of IP transit. Are you charged for IP transit in a similar way that you charge your customers? Or is there some type of difference that could help drive margins either higher or lower when you see them inflecting, if you will?

Yes. I mean -- I think there's a pretty wide range of business arrangements that we have with all the networks that we do business with. So to simply answer your question, do we pay for transit? The answer is yes, but only we buy it and only when we use it. And that is not the prevalent way that we're delivering content to our users, right? One of the big advantages from a cost perspective of being able to and focusing on deploying deeply in networks around the world is the fact that we'll be on network in many of those cases. And so the business discussions there with those networks are much more around being able to serve their end-user as well. And in many ways, we're saving those downstream networks, their upstream transit costs, right, because they typically need to be buying transit from a big provider, whether it's CenturyLink, or potentially AT&T or BT or wherever in the world they are. And so that then opens up much different business discussions and relationships than sort of that pure, we just pay for the bits. And so it's -- we've got an entire team that focuses very specifically on driving the network costs down and optimizing them also for performance. So there's -- they've got multiple factors that they spend a lot of time and energy working on to ensure that the structure we build, manage the cost profile and the performance. And one thing that's evolved over time as well has been -- we've seen -- transit is expensive, at least at the scale that we sort of drive it at. And so we continue to also evolve our network architecture. So over the last few years, we've also been building out our own backbone that allows us to control the cost a lot better. So we might be buying a fiber wave from someone, but that's much more of a very fixed cost and then we can figure out how much we're pushing it or optimizing the usage over it. And how much we then use those links to source content from our own platform rather than having to pay for transit or even peering, right? Peering is typically not paid, but you've also got a business relationship you're negotiating there. So there are very many flavors there. Obviously, we try to drive down costs as much as we can all the time. And I think we're very successful at it. But it does take sort of a multitude of different ways to do that. And it's a little bit different in different parts of the world. Some parts of the world are very balkanized between the networks. So you just -- if you either need to be in those networks or you're paying transit, other places you can get peering, even in sort of telco hotels where you can be peering with multiple people, right? So it really does depend a little bit around the world, but certainly, we try -- I don't want to say we try to minimize using transit, we try to minimize the cost of when we're serving traffic and make sure that it's still at the performance levels our customers' desire.

Do you think that most network optionality, most level vendors to what you could work with in the United States, is that a fair statement?

I mean I haven't looked at everyone else that is -- I would say yes to that. But obviously, I haven't audited everyone else's. But just given the architecture we have and the deployments that I know we have in the U.S., I would say, yes. But the U.S. is also a very different place than Southeast Asia, or India or even Europe, where the importance of being deep in the networks can vary depending where in the world you are.

And how long has this been going on, this decision or thought of own your own backbone, whether it's buying dark fiber, buying lit waves, et cetera?

It's a good question. I mean we've certainly been using our own gear and building this out for the last, I want to say 3 or 4 years, I'd have to go back and check really with the network team. And it's been an evolution. And again, it started with -- if I remember, it actually started with both optimization of performance as well as cost and control. And some of it is, you might imagine, just think about sort of a big metro. I'll just pick New York City. We might have 30, 40, 50, 60 data centers in New York City, where we might be communicating a lot between our own servers. It just doesn't make sense to pay transit or even use peering or anything that is a more valuable resource in the sense of you're working on a business relationship with it than just build our own fiber connectivity to talk then you naturally go from there, you're connecting cities and then from cities to continence. And so it's been an evolution of how we're connecting that backbone around the world.

Well last question on this whole pricing dynamic. But are you seeing the reduction in pricing on a per bit basis starting to slow? In other words, I always think of pricing Moore's law, per bit pricing goes down and down and you make up for that by selling greater volumes. I have historically thought about price reduction in the, call it, 20% year-over-year type decline. Are you noticing any bending in that curve?

Yes. Let -- I'll let you jump in, in a second, Dane, before I go on to talk publicly.

Yes...

You don't want Dane to talk first?

And I'm going to go first there only because it's funny. We can talk about third parties guesses on what pricing is. But in general, for those that have been at Akamai a long, long time, there's sort of a window. And it sort of vacillates in that window. And I think right now because there is a little bit less capacity than there is demand. There's a little bit of equalization of where that pricing range is, but it's still been in that range. I mean it's an aggressive price decline every year to your Moore's law example. And there are a number of different optimizations that's run by our platform team or Dane's team that we can try and do across software and hardware and port usage and bandwidth that we -- in colocation that we try and do to try and continue to make this very affordable for our customer.

Okay. I want to switch a little bit here, international revenues. International revenue comprised 44% of the total in the quarter -- first quarter versus about 41%, call it 42% in 2019, and that's despite FX headwinds. Can you guys talk about the biggest reasons that International and U.S. growth are just so different? I mean what is it that -- is it just simply a function of maturity in the United States and if...

There's a couple of tangible reasons, and then I'll let Dane jump in here on the media side. But from a company lens, we made a very significant investment in diversification of revenue about 5, 6 years ago, where we doubled and then doubled again the sales force, right? And what we've been adding is more and more products to keep all these sales individuals fed. But we really doubled down on investment in the company, not only on our product lens, but on a go-to-market lens. And as you know, we've been adding more sophisticated type products. So we've also had to invest in sales enablement. And I think we've really improved the quality of the average quota-bearing rep dramatically and their ability to sell multiple products across multiple regions. So that's the first real fungible reason why you see such great growth in across APAC. And EMEA is, it's not one country, right? It's some countries do better in various quarters and others do better in other quarters. So it's important to have that sort of diversification. But in terms of the U.S., we house a lot of the large sort of Internet platform customers in the U.S. right? And so as you know, we went through sort of a challenge back in the back half of 2015 and '16, where you saw a lot of that content being brought in-house by a couple of those giants. Frankly, all of them have some in house capability. And what we're starting to see, and Dane, maybe talk about it in the carrier lens. But this appetite to go do it by themselves is a little bit lessened by the financial dynamics of bit pricing. So maybe you can jump on that, Dane, if you want?

Yes. I think there's 2 aspects of that, I would say, about sort of DIY or bringing in house. I mean one is definitely we're just talking about price declines. And I think that, like Tom said, I was going to say the same thing, right? It's basically in the range that I've seen for the entire time I've been here and sort of worked in it. It does fluctuate a little bit. But those price declines mean that over time, if you're looking at sort of a TCO analysis and you want to build out your own, I think it does become kind of a tough decision, right? Because then you're going to be investing in staying ahead on technologies, and staying ahead of where trends in the industry are going to be. And then you still need to be out there making relationships with all the networks, making sure you have enough capacity. So there's a real pro and con there in the equation that I think the pricing definitely weighs one way or the other. I think another big one there is just global reach. Right. I think that equation might be a little bit different. If you want to target a mature market in the sense of networks and to be able to get connectivity at good price points. I think that's easier in a market like the U.S. then if you want to say, Okay, I want to go big in the U.S., I want to go big in Europe, I want to go big all over Asia, I need to be big in India, I'm going to go big in Latam. That's a much, much, much different equation than trying to figure out if you wanted to sort of go on your own?

So we've talked about the demand environment a little bit, OTT and gaming, in particular, talked about some of the pricing trend. But ultimately, these 2 things combined to give us a sense of revenue growth. But a few years ago, senior management talked about the media segment, likely at some future point seeing a reacceleration in revenue growth as video consumption would shift increasingly more to an online format, and it seems like that's now. Do you think that it's still that's the case, we should start to see an acceleration in media growth -- revenue growth because of all this? And if not, what are some of the preventing dynamics?

Yes, the CDN business by itself, yes, for sure, right? We combine that on a product reporting level along with other parts of web acceleration and other things. But we definitely saw it in the first quarter. I don't think the world has changed much in the second quarter. So my answer to that is yes. We've always sort of thought of the CDN industry as sort of low single digits with step functions in the mid-single digits based on things like what Dane has already talked about, OTT and perhaps even more gaming.

And I would also just to add a little bit to that is that it's not just a delivery piece of it, right? Because the delivery piece of it absolutely will drive up but then there's all the other services that we spend a lot of time and energy, especially our entire security profile, that go hand-in-hand and enhance a lot of those delivery pieces. So those can also -- that traffic growth can also help accelerate potentially the other side as well.

I guess to that point, within the Media and Carrier segment, which is one of the verticals you break out, what percentage of revenue roughly would you say comes from cloud security business? And can you give us a sense of what products, what types of things these types of customers are just in acquiring?

Yes. So I'll let Dane talk about the customer relationships. But in terms of the way to look at it, and again, part of the sales enablement came with improving the quality or the capability, if you will, of the media sales reps to go sell security. And so when they were looking to continue to grow their revenues, you had an enablement of them as well as a commission structure that forced them to learn how to sell security and on the web security products. When we last gave you that number in June of 2018, it was roughly about 6% or 7% of their total revenues. And I would say, frankly, it roughly doubled in that -- so that would have been for 2017. And so in 2018, it roughly doubled. And again, it improved by a significant factor in 2019. So -- and we still think there is a lot of room in those media customers for those web security products, which just got added to earlier this week with the page integrity sort of release.

Okay. I wanted to shift over and talk about competition, and I'll bring in a question that I got as well around this. But within the media segment Dane, who do you view as Akamai's biggest competitors. And when you're in a bake off, what are the top reasons you typically see a customer go with Akamai or at least award you what you believe would be a disproportionate percentage of their traffic?

Yes, I think there's any number of competitors. I mean there's certainly the historic ones that we've had for a while, Level 3 or CenturyLink and Limelight, and there's plenty of bake-offs there with those customers. There are some newer entrants. But I would say that in many cases, the bake-off or the reasons people go with us, there's a multitude of them, right? They can be from performance. But often, you've got customers that are definitely very focused on performance. And especially if you think about direct-to-consumer, I will say there's always a cost concern there. I mean, you see it in sort of price compression. But there is a very, very big focus on performance, right? And you really need to meet their targets and in fact, exceed them in many cases in order to win that share. So it's not even necessarily during the sales cycle where you're winning a bigger share, it's all the time, right? Because many of these bigger companies are doing performance analysis all the time. And they can adjust and swing their multi-CDN environment based on what their end users' performance is being. So it's not only that you're winning it during bake-offs. The most important thing is to also be winning it all the time, right? You have to provide consistent performance at scale and globally. I mean, for us, that's certainly a big reason why I think we're differentiated from some of our competitors just in breadth and reach of getting to consumers around the world. It's also a breadth of portfolio, right? From a media perspective, we've -- as we talk -- we focus a bit more on sort of the delivery and the bit delivery aspect of it. But if you think about the security products like we just touched on, they've got websites where people go in and sign up, and they've got websites where people log in and sort of all those, whether it's done through an API and an app or people are going to the website, they need protections on all of that because if those functions fall down, then no one's pulling down content anyway, right? So they use a fairly full portfolio, especially on the websites, the provision apps, identity management, web application firewall, bot manager, very similar things that you would expect from sort of an e retailer and how they want to protect their site and make sure that it's functioning well for actual people who are actually going to buy something, the media customers have very similar concerns. The lens is slightly different. And so if you look at that breadth of ability that we're able to offer, on top of the global reach and the performance we're able to give, I mean, I think that is really what differentiates us and gets us a big share with some of our biggest customers.

And I guess just to fill that out, and this is from the audience, it says what does Akamai attribute Fastly's success to? Is it merely pricing? Or is it programmatically -- at the edge, a true differentiator for Fastly versus Akamai?

Well, I mean, again, I think the APIs are important. Akamai has similar capabilities in that front. I think their success is it's a very competitive marketplace. They are primarily focused in the United States, where the backbone is a little bit more -- a little bit more subjective to a centralized sort of data center, and so they are competitive. But many of the things that we hear from them are something we heard from Limelight many years ago and there's just one more competitor in the space. I think we don't talk about pricing. I'm sure they don't either, but they're in a marketplace where customers are splitting some traffic, and they -- it's a rising tide of all ships. And so we are benefiting from a lot of those trends, and you should expect all the other CDNs to benefit from similar trends.

My last question, and if we could be brief because we're a little bit past, but I think it's important. Dane; in your response, you talked about quality as one of the differentiators. And I think one of the things that kind of expands on that is the relationship you guys have with your carrier partners. Is that something you just talked a little bit about and how that kind of fits into that opportunity to have a perhaps better quality experience?

Yes. I mean I think, as I touched on before, we are in thousands of locations around the world. And what we do is spend a lot of time and energy to see where customer demand for our customers' content is coming in from. And then we try to deploy as close to those end users as possible, right? I mean just from a simple perspective of being on the same network and being able to serve the content directly to them on their backbone or even perhaps in the same data centers that their connectivity terminates in, just gives you that closer first-mover advantage of being able to have better performance. It also allows you to have better scale. Before we touched on transit a little bit, at some point, if you're pushing a ton of traffic into a network, those networks only have so many links as well, whether it's paid transit, whether it's peering. And if you start congesting those, now you're going to impact every single end-user in that network. And so our ability to serve ideally, all of the customers from on-net, we wouldn't even necessarily notice congestion, perhaps upstream if we had a rich retrieve content. But certainly much less than anyone who is centralized and really either buying specific links into that network or relying on transit through a third-party or peering, they're going to be really limited by that specific link, which is already, in many cases, further away from a network perspective than our servers are going to be.

Guys, with that, we are out of time. Thank you so much, Dane, Tom, Rick. Really appreciate you guys doing this for us and be safe. And I hope to see all of you in person in the not-too-distant future.

I hope so. Thanks a lot.

Have a nice day, everyone. Thank you.

Bye.