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

Hey, great. I think we're about to get started here. Thanks, everyone, for joining us today at JPMorgan's 50th Annual TMC Conference. My name is Noah Herman. I'm a software research analyst here at J.P. Morgan. We're really happy today to be hosting Akamai CFO, Ed McGowan. Ed, thank you so much for being here today. Really appreciate it.

Thanks, Noah. Great to be here.

Yes. So I'll be bringing the audience's questions into the conversation. Feel free to submit questions through the conference website. Also towards the end of our session, you can raise your hand, someone with a microphone will come and help assist with your question. So with that, we'll start things off. Yes, again, thank you so much for joining us today. Can you walk us through Akamai's journey since inception in 1998? And how the company has evolved alongside the Internet, and I think at this point, serving 1/3 of all Internet traffic?

Sure. Yes. And it's funny. I've been here for most of that journey. I started in -- and originally the company was formed as a research project out of MIT. And the whole idea was that the Internet wasn't built for performance, the way the routing of the Internet was just a bit random, it wasn't ever designed to handle big flows of traffic and take performance into account. So Tom Leighton, our founder and Danny Lewin, his student, who looked at solving the problem of the worldwide weight and created a overlay network on top of the Internet basically, and we distribute our machines in thousands of locations. And most people think of us for sort of our original routes, which was delivering content and making content go faster, scaling the Internet to be able to watch video and download large files and things like that. But over time, the company evolves in some of the logical fashion, if -- sort of the next big journey after we got into content delivery and streaming was to get into security. So not only delivering the content, but keeping our customers and our users safe from things like denial of service attacks or defaming of websites or stealing of credentials, that sort of thing. And that business has grown up pretty significantly over the years. We sort of pioneered Edge computing, we're way ahead of our time, back in the sort of the 2000 time frame. That's obviously seen a resurgence of growth, and we just recently got into the pure-play cloud compute business with an acquisition that we did. And today, less than half of our business is from CDM and the majority of the business is in compute and security. And here we are today some 20 years later.

Yes, no, that was a great overview. And sort of related to Internet traffic, what -- coming out of the COVID-19 pandemic, what sort of permanent impact have you seen both on Internet traffic and security?

Yes. It was fascinating going through the pandemic. Obviously, going into that period of time, nobody knew what was going to happen to their business and something that no one had dealt with for over 100 years. So one of the first things that we noticed was a massive increase in our traffic from all different directions. It was primarily volumetrically driven by people watching more video, playing more video games, that sort of thing, just being at home, locked down, spending more time. It was sort of an acceleration of a trend. We were seeing more and more video being consumed online, people playing multiplayer video games and that sort of thing. So it really just accelerated that trend. So we saw about 2 years' worth of traffic growth in 1 year. And at the same time, with people going online and some people shopping online for the first time, we're doing banking online for the first time, creates a big target for the bad guys to go after. So security attacks started to increase and the demand for our security products increased pretty dramatically, and we grew our security business successfully. So over that -- we got sort of a COVID bump, if you will. And we saw that traffic sustain for about 2 years, and we're starting to see things moderate a bit as we talked about on our last call in terms of traffic growth. And I don't know if it's just a temporary pause in terms of the consumer behavior where mask mandates are going. People are going to movies again and malls and that sort of stuff. And the weather is warmer there in the northern hemisphere. So we're not seeing quite the traffic growth that we normally see. But certainly, those trends were accelerated. They're here to stay. The security business is still doing great. We just did the 26% year-over-year growth in our security business. It's on a run rate of over $1.5 billion. So that certainly helped bolster that business, and that's obviously continuing to grow at a very healthy rate.

Yes. And I think related to the CDM portion of the business, last quarter you mentioned that customer churn is very low and Akamai is continuing to take share there. Can you give us a sense of the competitive environment, why you're taking share and also just on pricing?

Yes. So to sort of step back from a customer concentration perspective, we don't have a lot of customer concentration. We're very fortunate in the sense that we've got great diversification from a product perspective, geographically as well as from a customer diversification perspective. Our top 10 customers, we define someone as a top customer as being 1% of revenue or more, last year made up about 19% of our revenue. Our churn is very low. We've had customers that have been customers of Akamai's as long as I've been here, so over 20 years. We've sort of as far as the dynamics in the business from a competitive perspective, if I break it into 2 buckets, the pure-play CDM business versus the security business, very different environments in terms of who the competitors are and also the underlying pricing dynamics in the security business. You don't see a lot of price competition or things based on price. It's really about the quality of the product. You don't see customers typically switch vendors very often. It's more of an evolution of on-premise-to-cloud security. So it's a very good trend for us. On the CDM side, we've had competitors that have been around for years, very price competitive driven market. A lot of your major streamers, software gamers, et cetera, will -- our gaming companies, software companies will split traffic among CDMs. So there really hasn't been any significant change. There's obviously a shakeup in the industry where 2 of our competitors are merging, Limelight and Edgecast. So think of like the high volumetric traffic business, a little bit of consolidation there. But really nothing out of the ordinary, over the 20 years I've been here, there's been a lot of competition for CDM, and you've got volume price dynamics that are still at play. I'd say there's really nothing different there.

And I think you kind of mentioned that you're seeing really strong growth in security, 26%. And I think it was outlined that security would continue to grow within that 20% range or so. Are you increasingly seeing your security products displace legacy services or cloud security services? And maybe if you could just touch on what are some of the key products that are really driving that growth in security.

Yes. So it's mostly legacy products that we're replacing on-premise firewalls and that sort of thing. In terms of the security products, we think of them in 2 buckets. You've got what we call our web security products. So think of that as protecting -- I'll use J.P. Morgan as an example, you're public-facing websites, your web applications, maybe it's your wealth management side, your training applications that are Internet-facing. That's a significant portion of our business. That's the majority of our business there. Think about web application firewall, distributed denial of service protection both at the IP layer as well as the application layer. Bot management protection, a lot of the traffic today on the Internet is not human, it's machines. So understanding the machines that are interacting with your sites, what to do with them, protect against credential stuffing and that sort of thing. And then we've, over the last several years, morphed into more of an enterprise security space. So think of protecting the employees of the company, how you access your internal applications. We just acquired a company called Guardicore that's in the micro-segmentation space. We've got products that would compete with the VPN for access to your site. We have a secure web gateway that will inspect traffic flows going in and out of your organization, again, protecting the employees and sort of the traditional IT securities stack.

Yes. And could you maybe just touch a little bit on why acquired Guardicore? How does that really boost your enterprise security offerings?

Yes. It's -- first of all, it's amazing technology, and it's the leader in the space. If you think about the challenge that the IT organization has, you need to look at your flows of data that go both from application to application, in and out of your organization. And we had, call it, north southbound protection with our Application Access product and our Swig, but we didn't have anything at the time that would look at east-west or as traffic would flow from application to application. And no matter how good your protections are, your endpoint security, your secure web gateway, things always get in. People click on bad links. There's stuff that gets through your e-mail filtering, et cetera. And a good way to catch that and stop ransomware is to have the visibility inside your network to see those flows as they traverse across the network, and you're able to prevent things as they get in from spreading. And Guardicore is really sort of changed the game for us in a lot of ways, where they are the market leader in segmentation. Segmentation in the old ways of doing things is very challenging. Tom showed an interesting quote on IR Day that said where CISOs used to go to die in terms of just the legacy systems and the challenges that, that brought. With Guardicore's very elegant solution at software-based, we're doing it ourselves. We're actually running IT as part of my remit now and I'm running IT, and we're implementing it ourselves. And the visibility you get out of it is incredible, and it really helps inform where you have potential threats that may not be where you would think it would be. So it gives enormous visibility. It's attracted a lot of attention from the SI community. And one of the questions I get a lot is, okay, and I get the web security business, you're well known there. But enterprise, you're not -- you're going up against different vendors, how are you going to get mindshare with the CISO? Guardicore is a very interesting journey for the SIs to come along with us because it's -- first of all, you have to identify what's in your network. You go through the implementation phase, you go through the blocking phase and then the maintain and run phase. It's a pretty big digital transformation for our customers. And it also drags along with the other products that we have to offer. So we've seen a market change in our approach with the channel and with the system integrators. I would encourage everybody to look at our Investor Relations Day materials. Our Head of Sales, PJ put our case study up there about how a system integrator's actually building out a practice around Guardicore.

And so it seems like a great -- also just cross-selling opportunity within security. Last quarter, I think 70% of your customers are taking one security product, 35% are taking 2, can you just maybe talk to the cross-selling opportunities within security? And what that could really grow to?

Yes, great question. So we've done a good job. If you look at those numbers, those have doubled over the last 5 years. So if I go back 5 years ago, about 35%, a little over 1/3 had -- one had purchased a security product. Now 70% of our customers, we've got -- we've tripled the number of customers from a percentage perspective. Obviously, that number is larger because we've grown our base but are buying 3 or more security products. So there's several ways you can grow your business. One is with the customers that have purchased a security products, say, for example, that web example I gave you, say, JPMorgan, if they're a customer buying our security products, you might put your web application firewall on your critical mission -- mission-critical apps. And then you may add applications as you go or you standardize. We're hearing a lot of customers now wanting to standardize and have all of their public-facing applications and web applications using the product. So you can grow one product horizontally across the organization to get more of the same product. And then you're adding features on top like Bot Management or Prolexic to protect from DDoS at the IP layer. And that -- leveraging that relationship with the CISO opens up the door to have a conversation about Guardicore and other products. So I'd expect to see that we'll increase the penetration rate of our security products with our existing accounts, acquiring new customers. Most of our new bookings are all led by security. And then there's a tremendous amount of wide space with adding our enterprise security products. It's a smaller percentage of our revenue today, but a much bigger market.

And as you sort of integrate more of these security products onto the platform, is there any change in the go-to-market motion? Maybe you could touch-up on what that's looking like? You also alluded to a little bit, you're working with the channel partners and other SIs. If you could just touch on that a little bit.

Yes. So right now all of our sales reps can sell our web security stack pretty easily. It's -- you're selling to the same buyer, the person who's building say the online video application is worried about that site being knocked offline or somebody stealing credentials and that sort of thing, so very natural sale to add that. On the enterprise side, we have overlay specialists that as we've acquired companies, and we've built up that capability, hiring people from traditional industries that sell to the IT buyer and then really leveraging the channel on SIs to grow that out. Really obviously, our sales force is getting more and more comfortable as they're winning and getting success with selling those other products, but pretty much everybody can sell security. So we're doing a pretty good job there. And like I said, with the enterprise stack with Guardicore in particular, we're seeing much more interest from the channel in selling that product.

Great. I sort of want -- maybe pivot over to the compute side of the business. Last quarter, you mentioned that Linode should account for about $100 million of revenue or 25% of the compute segment. Can you elaborate on what Linode brings to Akamai? And what's really fueling that growth?

Yes, great question. First of all, it's an enormous market. The way -- probably the easiest way to describe it, today, if you wanted to use Akamai's Edge computing, let's say you were in, say, the ticket business and you're selling tickets, you might have your application built on a third-party cloud on one of the hyperscalers. And you would come to Akamai for something like a waiting room application. So you're about to launch a sale for some real popular concert and you're going to get flooded with a ton of people coming to your site to go buy the tickets or whatever. Having an application that runs on the Edge is that will give user experience, put them in a queue and be able to filter them down into your application without having your application fall over, a traditional sort of Edge workers like application. With Linode, we have the ability now to build the entire application. So it could be the commerce engine, the database, all the stuff that you would think about, you would go to a hyperscaler for, you can now use Akamai with Linode. And we'll be building out capabilities and functionality. They're primarily focused on SMB. So we're going to be adding capabilities to that to make it more robust, adding capacity, adding more locations, certainly low latency applications. We're getting a lot of customers that were coming to us saying, "Hey, I want to run my full stack application on your network". This just gives us the ability to do that in a much more scalable way and give us some new features and product that opens up an enormous market for us. There's really a natural extension of where we were sort of our journey, as I was talking about, our 20-plus year journey. It was always a question of when we're going to get into the space not if.

And it seems like a great opportunity to sort of bring what Linode has been doing, improve on it and then go after the enterprise base where you have a strong foothold in. How do you see that -- or I guess, how are some of the early discussions you're having with customers that were already Akamai customers with the Linode product?

Yes. So far, the early conversations have been great. There are certain things that enterprises require that Linode hadn't been focused on. They built the company not focused on the enterprise. There are some questions that we got at IR Day about their enterprise sales force. Well, they didn't have one, right? So they were just building things that were so simple to use, you could sort of fall into a virtual machine and get your application up and running. But there are certain things that our customers are expecting, availability zones, virtual private cloud, certain compliance capabilities and being aligned with certain compliance functionality, more capacity. So what we're hearing from a lot of our customers is something that we're experiencing ourselves, which is the spend in that area has gotten out of control and growing very, very quickly. It's concentrated with a few vendors. Prices are going up in some cases. And we were looking to move all of that to our network. We're able to do it a lot faster now. And we're hearing a lot of customers that are either competing with some of these third-party clouds looking to save money, looking for another alternative, diversification of vendor as well as, like I talked about the low latency applications where you want to be in, distributed in many more than just a handful of locations. So all that's sort of lining up in our favor, and we're talking to folks that might be spending 10 to 15x more on their compute spend than they are in CDM, and that's what some of the view on our biggest customers.

Yes. And I think it's really impressive, you've kind of put out the $500 million revenue target for that by 2023. So what are maybe some of the things that investors should be focusing on near-term to sort of see how you'll be able to achieve that target?

Yes. Good question. We'll be providing additional metrics certainly as we get towards the end of the year, the run-rate revenue, number of million dollar customers, try to give some helpful metrics on how the pipeline is going as well as some reference customers to the extent that customers will allow us to say who they are, if not, we'll just give you examples of used cases in industries and that sort of thing. But for me, what I'm looking for is, what is our exit run rate coming out of the year, what does that backlog look like? Are we getting repeatable wins, what are some of the verticals where we're seeing good traction? I think there's -- if you look at PJ, our Head of sales, had a slide that showed our penetration in all of our verticals by percentage of customers, percentage of revenue. And you can see 2 things on that. One, there's certain industries that very logical that we'll be able to tap into media, commerce, banking, et cetera. Bank is a little tougher with the compliance. It's a little bit longer of a cycle, but certainly with the media and commerce, I expect that to be some early wins for us. But there's a number of verticals that we're underpenetrated in, life sciences, manufacturing, that sort of stuff. It's been a lot on third-party cloud. So tapping into our channel, inside sales motion, developing new territories and sales to go after some of those verticals that traditionally aren't big web customers, if you will, gives us an enormous opportunity.

Great. I think we'll maybe take a quick pause here just to see if anyone has any questions. You can just raise your hand and someone with a microphone will come. Anyone have any questions?

Just what's inspiring people to go to Linode from, let's say, AWS, for example?

Yes. So I'd say some of the early use cases are funding a competitor. So not wanting to go move to someone else is not competing with them. Low-latency use cases, people that say, "Hey, I need to be more distributed. I've got a low-latency use case where I need to be in more than just a dozen or so locations." And then cost savings is another big issue, right, which is I'm spending too much, I want to reduce my spend. There's a lot of synergy that we bring right off the bat. It's not -- when you think about an acquisition, you think about synergy of cutting costs here, it's leveraging our go-to-market, but also our supply chain and our network build capabilities, meaning I've got a team of techs that will go out and build CDM capacity and also build out Linode capacity. I've got relationships with thousands of different network providers, hosting providers, significant amount of my bandwidth is for free. I've got a backbone or what they call network fabric. My engineers get mad when I say backbone. But I can leverage that, and I can compete, say, for someone who's spending a tremendous amount of money on egress fees, move that application to us, you can save on the egress fees. It doesn't cost me anything because it's already a sunk cost. So I think certainly until we get more use cases under our belt and we put more platform-as-a-service capabilities. Some of the early wins will come from that cost savings and looking for diversity of supplier, not wanting to fund a competitor, that sort of stuff. And I think we -- I think the slide that Adam and Tom showed is pretty interesting, which was you've got Linode, which was sort of in the bucket of the small cloud provider, you got the hyperscalers and sort of this need in the middle for somebody who could provide similar services to hyperscaler. Maybe not the whole full-stack of everything, but a robust enough offering that's better than what's out there today as an alternative that has the credibility and the ability to scale, and can offer these types of services, along with the ability to not only just build your application, but as I talked about, you can run functions as a service on the edge. You can add security. We're going to be adding delivery capabilities for the current Linode customers to be able to get basic delivery features. We'll be launching some basic security features that people can self-provision themselves. So I think there's a benefit for the existing customers as well as our other larger customers to move over to us.

And just sort of following up on that. I guess, if you could just maybe give us a little bit more insight into how you're able to provide the offering more affordable than your peers? What is that -- I guess, what's the incremental savings for customers that maybe would be going with you versus a peer?

Yes. So it's kind of interesting to sit in this seat and do to others what's been done to us for years. But we can actually do it profitably, that's the difference. We've got a lot of people compete on price with us and lose money here. We actually -- I showed some of the margins. I encourage you guys to take a look at the slides we showed, but this is actually something that can help expand our margins. And you can see the pricing is very transparent. There's some statistics that will show how we compare to others from the hyperscalers and other options that are out there in the marketplace. And we can do that at a very, very attractive margins.

Got it. Any other questions?

I had a quick question today. It looks like over time, you spent a couple of billion in strategic acquisitions. So I'm just curious how your philosophy has been in build versus acquire, and if there's any changes to your M&A outlook on the road ahead?

Yes, great question. So obviously, last, say, 9 months or so we've done larger acquisitions than we historically have done. Guardicore was about $600 million, Linode was about $900 million. We traditionally have done acquisitions sort of in the $100 million to $300 million range. Typically, they're either tech tuck-ins or they're a product addition. I'd say probably stay with more of the traditional, we look for opportunities. I think those 2 opportunities were pretty unique. We had a chance to catch a market leader in a space where there really wasn't a significant player in that space, and we could become that market leader with Guardicore. With Linode, financially it was a great transaction. We were very fortunate. I sort of challenged Adam as we were getting -- Adam's the one who runs that business for us, he's our COO. We're getting lots of requests from customers to be able to provide more ability for them to build a full stack application. I said, look, we can do this on our own or go find me an acquisition that could be accretive, that gives me instant functionality simple and easy to use because that's the key with that business. It's got to be easy to use. Developers want to be able to get things up and running quickly. And along comes Linode, very closely held company, timing was perfect, great cultural fit. We got a significant tax benefit in the way we structured it. It was profitable and accretive. So that was, to me, was a very unique opportunity, that could give an immediate return to us and get us into this massive market. From a going-forward perspective, we'll be opportunistic in terms of build versus buy. Oftentimes, if you think of the journey to get to $50 million to $100 million in revenue, you're building it, you takes a couple of years to build a beta product, and you're out kind of working with customers, getting it to be a better product, 5, 7 years, you get to that level of revenue typically. So if I can pick up a company that's 5 years into their journey and just getting to scale, plug it in, it either fits into an existing product or I can just plug it and sell it with my sales force. Sometimes that gets you better scale. And you see that certainly with Guardicore. We've almost doubled our expectations. I talked about when we first bought it, we thought we'd do about $30 million on the Q4 call, I talked about how we think we'll do $50 million to $55 million. And we're bringing opportunities to Guardicore as a company just using Akamai as the brand and with our relationships that they never would have seen. So there's times where that can make a lot of sense versus building things. Obviously, valuations are -- can be challenging. They're kind of working in our favor as a buyer. It hasn't quite hit the private markets yet, but it will. So we produce a lot of cash and we've got plenty of cash on the balance sheet. So we'll be opportunistic and look for opportunities. We think we've done a good job there and return pretty good value to shareholders.

And I think with that too, along with the M&A, you've -- the company's consistently been a re-purchaser of stock buybacks. Any change in that philosophy or anything different from historic precedent?

No. So we've -- if you look at over the last 10 years, we've bought back about 10% of our float, so about 1% a year. There's no formula that says that's the way we're going to do things on sort of how things have worked out. We've talked about offsetting dilution. Obviously, as a tech company, stock-based comp is a big part of our philosophy. We produce $800 million to $1 billion in free cash flow. So using that to offset dilution and then opportunistically buy back shares. You can imagine working for a mathematician like Tom, we have a formula that we run that takes into consideration different parameters of where the stock is trading at, and we'll buy back more when it's undervalued and buy back a little bit less when it's fairly valued, so nothing there to really call out in terms of a change. But I think we've been pretty consistent with our balance of M&A and buybacks in terms of use of capital.

I think definitely one of the most recurring questions we're getting at the conference is the impact of businesses with higher inflation, higher interest rates than just the macro headwinds. Any thought on how that will impact the business and what assumptions you were making within your current revenue and margin outlook?

Yes. I'd say the biggest impact for us has been the dollar. So the dollar has gotten significantly stronger. So that's one of the biggest impact in terms of like macroeconomic factors. And I called out just how big it is and how profitable we are aside of the U.S. that had a pretty big EPS impact. In terms of inflation, we've done -- we're not chasing that from a cost of labor perspective. Obviously, hiring gets a little bit trickier in this environment. So we're not immune to that, but we didn't go and go outside of our norm in terms of salary increases and that sort of stuff, as we weren't also providing 1% raises either when inflation was 1%. So we're trying to stay consistent there and not chase that. A lot of our bandwidth we get, for example, is free. So there's no inflationary pressure on that. There's -- unfortunately, in this business we can't raise prices, although I'd love to do that someday. So it hasn't been a huge impact, but that's not to say that if things get worse, nobody is immune to inflation. We're lucky we work with some of the biggest customers and companies in the world, that bankruptcies and that sort of stuff probably won't have a material impact on us. Obviously, the economy gets really, really bad, who knows. But the supply chain, some delays, but we've done a pretty good job trying to stay ahead of that. But it is something we're keeping a close eye on. And hopefully, this is a short-lived recession and things get back to growth again, and we get back to normal levels of inflation. And hopefully, the dollar starts to weaken a bit.

Yes. And I kind of just want to follow up on that on your exposure internationally versus domestically. How do you think about that longer term as you sort of bring on these new capabilities on to the platform?

Yes. So we've done a phenomenal job growing internationally. We've got a huge advantage over our competitors because we've made those investments in our sales force, our support organization, services organization in the network. So I expect that to continue to be a bright spot for us. It does create more exposure because outside of the fact, we've got some development centers in Bangalore, we've got back office operations in Poland, a little bit in Costa Rica. Most of our cost is in the U.S., so you do have sort of an imbalance. So where you can hedge your balance sheet risk and that sort of thing, it's hard to -- I've done some back testing. It doesn't make a lot of sense to hedge my cash flows at this point. That's something I'll keep an eye on. But I look at international it's just an opportunity for growth more than anything.

I think too, it was at the Analyst Day sort of briefly mentioned that having Linode now that actually might provide some cost savings to Akamai itself. Maybe just -- if you could just elaborate on any of that?

Yes. So we haven't provided the number, but it is a meaningful cost and that would definitely help us. So as I talked about kind of the margin journey that we're going on, there's a balance, right? You want to be able to invest in the business and continue to grow your opportunities in front of you, our security business, our cloud business, we're going to want to invest in more capabilities there and maintain a healthy margin. And we're going to do that through mix. It's going to be the biggest driver. So higher margins for security, higher margins for compute. And Tom showed a slide that by 2026, we could be looking at only 30% of our business from delivery, 70% from security and cloud with higher margins. But there are also cost opportunities. And one of them is what we're spending on third-party cloud. I think Adam gave a really interesting example that for Guardicore, when we acquired them they're using one of the hyperscalers for some of their cloud-based product. And we had an opportunity with several customers where if we had used a third-party cloud, we would have spent $20,000 a day, and we were able to do that on Linode for virtually nothing. So it just shows you how quickly companies you can spend, but also just a small sampling of the opportunity from a cost savings perspective, but there's a significant opportunity there for us.

Yes, that's good. That's great color. I think we have time for maybe one more question. Does anyone have -- anyone have any before we wrap up here? Okay. Well, Ed, thank you so much for being here today. We really appreciate it, and thank you all for showing up. Thank you.

Thank you very much.