Good morning, everybody. Tim Horan, the cloud and communications analyst here at Oppenheimer. My pleasure to be hosting Ed McGowan, again, from Akamai, the Chief Financial Officer. Ed, we're gonna hop right into it unless there's any comments you wanna make before we start.
No, just thanks for having me, Tim. Good to see you again.
Great to see you. So really the story to me right now, and the most things we're getting questions about, is your cloud compute business. Congratulations, it seems to be a great acquisition. You got a lot of momentum there. Can you just talk about where you are in the ability to serve enterprise customers? What have you had to do-
Yeah
Going from kind of SMB to enterprise? How much maybe, you know, how much you've spent to CapEx or OpEx and, you know, what type of exact, you know, service improvements, you know, have you done?
Yeah, good question, Tim. So we've spent the last two years really investing significantly in the platform. So we did two things. From a physical perspective, we built out 14 new core data centers. Those, think of those as 5, 10-megawatt facilities, pretty large data centers, so expandable. We don't build out right away into the full capacity of those, but pretty good-sized sites. We connected those to our backbone, and we also connected the 11 or so sites that we acquired from Linode and made those a little bit more robust. They weren't quite enterprise-grade in terms of, you know, the connectivity and the capabilities there, so we upgraded that quite a bit from a physical standpoint, and now we're out building the 100 or so Gecko sites. In terms of CapEx, we spent, you know, $200 million in CapEx.
Some of that was for our own use, and we were sort of the first major enterprise customer. We added a lot of functionality to it, things that, you know, companies need, availability zones and things like that, and, you know, other capabilities and some compliance, so PCI compliance, SOC 2 compliance, and we're still going along that journey. Added some partners to our marketplace so that we have, you know, more robust offerings, repeatable offerings, and things like that. And now we're out selling to enterprises. We've moved, you know, close to $100 million.
There's still a little ways to go there of our own capabilities, and finding that performance in some cases is better, and, no issues as far as being able to move real production-ready workloads that have, you know, hundreds of millions of dollars of revenue behind them. And, we're ramping up that enterprise customer, count as well. As we talked about on our Q1 call, we were at $50 million in, call it ARR, with over, you know, 250 customers, and now we're, you know, projecting to be over 100 by the end of the year. So we're seeing very good receptivity from our customers and our sales reps, and very excited about the traction. So I think the 2 years is starting to pay off, and we're now starting to reap some of those rewards.
A lot of questions there for me. The $100 million, is that showing that you're saving on cloud? Where does that show up in the income statement at this point?
Yeah. So it's a little obfuscated in the sense that it's mostly in cost of goods sold. But, as I talked about, as we enter some of those big colocation data center sites, we're required to account for those under the lease accounting standard. So we've made some commitments, long-term commitments over, say, a 10-year period. You have to amortize those out on a straight line basis, so we're actually taking more expense ahead of our actual payments. So we've got about a point of margin pressure there. And then, as we build out these sites, we have colocation expense that, you know, we're not quite utilizing yet. So, we have some headwind there as well, just in terms of just getting to, you know, the efficiency out of, you know, selling into those locations as well.
So we've got, you know, we saved... We're probably about two-thirds of the way done, you know, closing in on three-quarters of the way done, and we've, we haven't really seen it shown up. If we hadn't done it, our gross margin would be down a couple of points. So unfortunately-
Got it.
You're not really seeing it in terms of net improvement because of what's, you know, the investments we're making back in the business.
Got it. So the $100 million, that doesn't show up in revenue at all. That's showing up-
No, no, no, no.
Yeah.
No.
It's almost entirely cost of goods sold, and it's a balance.
Sure.
You're saving some of your own money, but-
That's right.
Obviously, the spending's going up. Where do you think you are with the utilization of the cloud assets that you've deployed right now?
Yeah, it depends on the location. Some locations are getting, you know, fuller than others. You know, obviously, when you sell compute, it's not quite like CDN, so you have It's based on, you know, where the customers wanna deploy their applications. So, we still have a ways to go, though. Like, so I was asked on an earlier call about, you know, given we took our guidance up, do we have to take our CapEx up as a result? And the answer is no. We just have sold more into the excess that we've deployed so far. But, you know, I think as we get, you know, going into next year, we'll still add some capacity.
If we get a specific large customer that, you know, might wanna deploy a, you know, $several million a month type application into one data center, we may have to add some CapEx for that. But we do have some room to continue to grow our own usage, as well as sell into, what we've already deployed.
Well, I mean, it's obviously, it's an enormous market approaching, you know, $300 billion for the Big Three.
Mm-hmm.
You know, obviously, they're growing revenue like $60-$70 billion a year on top of that $300 billion.
Yeah.
You know, it's just an enormous, enormous market.
Yes.
You know, everybody seems to want some diversification for sure.
Mm-hmm. Yeah.
You know, a lot of your customers are competing against-
Yeah, exactly.
-the big three cloud guys, which is, you know, also, also an issue. So I guess to me, it's not too surprising that enterprises wanna use you. And it's almost seems like it's in your control to grow revenue almost, I'm not saying at will, but-
Mm-hmm
... if you have the product and you have the capacity, I think you're gonna be able to sell it, I guess, is what I'm saying. So,
Yeah. Yeah, no, I'm with you on that one. I think, you know, one of the things that I'm most encouraged by is the, you know, as we get to our first $100 million of enterprise customers, I would have expected it to be more customer concentrated with, you know, a handful of sales reps going after, you know, our biggest media customers and getting them to move a significant, you know, portion of their business, but a decent chunk of revenue. What we're finding, though, is all sales regions, all reps are selling, and we're seeing a lot of repeatable use cases in a couple of different, you know, things like observability and data analytics and insight, encoding, transcoding, storage.
You're seeing, you know, the customers doing things that are latency sensitive, you know, say, complying with something like the Apple Pay standard for financial services and things like that. So we're seeing a lot of participation from many, many different reps across many different verticals. So I'm more encouraged by that because to your point, the market is enormous, and there's a lot of business that we can go after that wouldn't be a traditional CDN customer. So we are looking at, you know, augmenting the sales a bit to, you know, put some hunting regions in place that you wouldn't if you were just selling, you know, CDN and, and web security.
So your service, you think it lend itself to some lower latency products at this point?
Mm-hmm. Yep.
Is that primarily how you're winning customers, do you think?
There's a number of reasons. I'd say cost is a big one. You know, cheaper just to buy it, just... If you're just buying it straight up, it's about half of what you spend in, say, in AWS. But also, if you have data that moves around a lot, you say you're, you might even be using a hyperscaler for your origin, for your website, even some simple use case like that. And even if your CDN is getting 90%, 99% offload, that 1% of cache miss could cost you as much as you're spending on your CDN bill with egress fees. So anyone who's having data that moves around a lot, the perfect opportunity for us to go after. You mentioned folks that are concerned about, "Hey, look, I just lost the rights to XYZ, to AWS.
Why am I paying my competitor?" You know, we don't compete with our customers, so that, that's, you know, a good selling point as well. But, you know, we also have built up trust with a lot of these customers for, you know, 25 years. So, you know, to your point, folks are looking for diversification. And then, you know, so other use cases for, say, latency could be, we're seeing a lot of ad serving, ad targeting, and ad decisioning. So the closer you can do that to where you're serving the content, it's, you know, just gives the user a better experience, and you can get overall better efficiency.
Are you doing anything to support AI in the cloud and in-
Yeah.
your services yet?
Yeah. I mean, you wouldn't run... You know, we, we're not looking at, you know, building out a massive GPU farm, so someone would say, "Hey, I'm gonna go build a big training model on Akamai." We do support GPUs, and we do have them in the platform, but we do see some use cases for AI as well. It may be more, I think over time, I think you'll see more kind of inference engines, so the smaller, compute use case that you would want to have more distributed. We wouldn't necessarily run a big training model in hundreds of locations, obviously, but for that more distributed inference engine use case, and we're seeing a little bit of that today. There's not a huge market yet for that. I think that's coming, but, you know, we are seeing some AI use cases.
You know, one of the pushbacks that we get is AWS probably has 1,000 different compute instances. They have, like, 10 forms-
Mm-hmm.
of storage.
Sure.
They have 10,000 value-added services.
Yeah. Mm-hmm.
I mean, despite all that said, probably 70% of the revenue is coming from basic processing and storage.
Yeah, that's right. That's right.
I guess, you know, is it an impediment not having, you know, all the bells and whistles at this point? Do you get much pushback?
Yeah. I mean, obviously, there'll be certain opportunities that are foreclosed, but if you think about what a, what an enterprise does, there's so many different opportunities within a customer. And it's not a zero-sum game. It's not like customers are saying, "Hey, I'm not gonna use AWS. I'm gonna move 100% to you." I don't, I don't expect that to happen. You know, maybe there's a few small startups and things like that where that may, may happen, but for a larger enterprise, I don't expect that to happen. But you can pick up ancillary use cases. Like, for example, we have some customers that are using us for observability, and they're just taking all their CDN logs and putting it into our platform and doing all the data crunching and storage.
And, you know, instead of using AWS, you could certainly use them for that, and they might be using them for five other use cases. So it's not a zero-sum game there. And I think some folks look at it as, "You've got to go convince all the developers to dump that and move to you," and that's not the case. You can pick up small use cases, small applications, and, you know, we found sometimes in going into some of our conversations that some of these developers are inside of some of our big enterprises have used Linode for their own purposes. So they might be doing, say, a demonstration, or they want to, do some kind of a test or something like that, and they're using it, or maybe it's just a hobbyist, they're using it. And so, you know, folks like the platform.
It wasn't enterprise-grade before, now it is. So, you know.
And what else have you had to do to make it enterprise-grade? And then, are you and where are you in that process? Do you think you're at, you know, 80% of the way through, you know, or?
Yeah. I mean, obviously, we're able to handle, you know, our own stuff when we were big consumers of all three platforms, and we haven't had any difficulty finding alternatives for anything that we're using. You know, Tom said on the call that we've actually designed out Snowflake and Databricks. So, you know, that we haven't found any limitations that have prevented us from using it. And, you know, I'd say, obviously, there's more capacity that's needed. There's more compliance. We're not FedRAMP compliant, for example. Our marketplace is not that robust at the moment. We're gonna obviously continue to add to that. There's, you know, other things that we need to add....
But right now, you know, we've got, like I said, using a ton of it ourselves, and we have, you know, on our way to 100 million enterprise use cases right now. So I'd say the platform might not be able to handle everything, but it can handle an awful lot.
How hard is it, or how hard was it for you guys or for your customers to, you know, transfer the data or workloads or, you know, all the processes from one of the major three to you? Or is it more new workloads, do you think?
I'd say it's a mixture of, you know, a lot of new use cases, I'd say we're getting, but, we are seeing movement of, you know, someone, say, maybe doing all their transcoding or video uplink or something like that with one of the hyperscalers and moving, you know, maybe it's several channels or some property or something like that to us. So you're seeing a little mixture, mixture of both. I'd say it does depend on how, you know, sticky the products that you're using from the hyperscaler. So if you're using some proprietary software that you have to either find a alternative for or design yourself, that can be a bit of a challenge. And obviously, as we continue to grow and add more, you know, options in the marketplace, that will become less and less of an issue.
But those workloads, you know, we're kind of telling our... You know, if you find that objection, move on, 'cause there's other things to go after. 'Cause, you know, and some folks will. We're seeing more and more of this, where you're trying to design things that you're not stuck, right? Because you want to have alternatives. So if you're using a proprietary solution from one of the hyperscalers and you have an outage, you're in big trouble, so you want to be able to port your applications. We actually. We were doing that ourselves before. Like, if I look at some of the big applications, like our portal, for example, we designed it specifically so that we could pick it up and move it from hyperscaler to hyperscaler.
That was a fairly, you know, easier thing for us to move because we designed it with the principle of wanting to be able to have alternatives.
So what is this, obviously, you're moving a lot of your workloads onto your own cloud infrastructure. You know, where are you in that process? Like, can you migrate the whole company to more of a platform strategy, where almost all your security and delivery are all operating on the same infrastructure, you know, over time? Like, where is that process?
Well, it's a good question. I would say, I guess technically you could, in theory, move your CDN to the x number of locations we have, but you wouldn't want to do that because it's not ever going to be as distributed as your CDN. But I think from a technical perspective, you could obviously take the 25 core locations and the 100 or so Gecko locations when they get built and run a CDN on it. It wouldn't be as good as the 4,000 location CDN we have today, so it wouldn't technically make a ton of sense.
And then, like, when you're doing web application firewall, a lot of the data crunching and stuff like that, that you do to, you know, inform the, you know, the decisions that you make or, you know, make your security products better, would be done in, you know, one of the central locations. But you want to apply the web application firewall rule at the same time you're doing the delivery. So there'd be technical reasons why you wouldn't want to do it, and performance reasons why you wouldn't want to do it. But, we're—as far as the journeys go, we're probably two-thirds to three-quarters of the way done moving what we wanted to move. You know, obviously, every time we do an acquisition, typically, most startups will use one of the hyperscalers, so you always have other stuff to move.
You know, we'll continue to use them in certain cases for small things, but, you know, our spend will be, you know, very, very manageable, a small amount of spend.
But I would, I guess, is it possible to interoperate the, the infrastructure a little more so that, you know, integrate it more? And from a customer's perspective, I guess they probably wouldn't even know about all the different-
Yeah
... components of the infrastructure that you have.
Yeah. So I would say, you know, we've connected all of our core data centers to the delivery platform, so, you know, you've got your backbone connected. So if you're using... Let's say, for example, you said, "Hey, I want to move my mobile site and, you know, run that on Linode and use that as my—use them as the origin for my mobile applications." You could use us for a CDN and be connected right away. There'd be, you know, basically very, very, very low latency because you're connected through our backbone. So there's that aspect of it. I mean, most of the hyperscalers have some type of a CDN offering. Obviously, we have a better CDN offering, and it's already connected to the platform, so you've got that advantage. And then security, it depends on what the product is, right?
So there's, you know, some benefits to, you know, running stuff on Linode, but we don't have... Say, for example, like, when people move applications to the cloud, you want your cloud security, posture management, workload protection. We don't have those offerings today, but, you know, maybe that's something we could do in the future. But, yeah-
Yeah
... so, you know, there is obviously some synergy between the platforms for sure but-
Are you worried about the Linode legacy business growth slowing, you know, as you're more focused on enterprise or maybe, you know, some of the products, focus not being as good for that SMB? Or, and have you seen any slowdown in the legacy business of Linode?
It's still growing. You know, it's not a primary focus for us. You know, when we bought it, it was $120 million or something like that. It's still growing. Every time we do something, we add something to the marketplace, we add more locations, we make the locations more resilient, we add compliance. All the SMB customers have access to that, so they could. The platform keeps getting better. As we get more well known in the marketplace, there's, you know, in the market, there's opportunity for folks to, you know, go there instead of going to, say, a DigitalOcean or some other competitor. But again, it's not a big focus. Even if that slowed down, it wouldn't really impact the growth that dramatically, you know, 'cause it's-
Okay
... you know, smallish business. It's, you know, like I said, it's bought at $120, it's still growing, so it's bigger than that, obviously, but it's not, you know, you're talking about a massive business.
So, you know, $100 million enterprise revenue exiting the year, it's phenomenal. What's the gating factor, you know, for growth going forward? I mean, for, you know, every-- I mean, do you have an... I guess the way to say it is, it seems to me if you get the capacity out there, you're going to be able to sell it. So, you know, for every dollar of CapEx that you spend on success-based cloud, you know, how much can you get in revenue off of that? And, you know, how are you thinking about how to allocate-
Yeah.
CapEx in the next few years? Yeah.
Yeah. So, so we're seeing, you know, our model suggests it's, you know, at least $1 of revenue. It could be a little more. I think the gating factor really is about the go-to-market efforts. I'd say are more of a, of an issue of just getting, getting ourselves better known, getting customers to use more applications. There's, like I said, verticals that don't, you know, really care much about their website. You think of things like, you know, bioscience, you know, healthcare, oil, natural gas, manufacturing, don't generally have big web presence, so they're not a big enterprise customer. But if we start to add some hunting capabilities or build up a channel there, huge opportunity. So that's just... You know, they don't know who we are at this point.
So once we get in there and start to, you know, bring our value proposition in, you'll start to see more, more selling there. So I think, I think it's, you're right. As you build it, you should be able to sell it. It's just you need people to sell it, and you need to get the message out there. So it's just, I think it's just a time aspect, you know, in terms of getting more, getting more well-known in the marketplace, getting more success, getting more applications with existing customers and, you know, building out the sales force to go after some verticals that spend a tremendous amount in that market but aren't really big traditional CDN customers.
But I would assume you're still, though, extremely underpenetrated with some of your larger-
Oh, yeah.
CDN customers, so that's-
Absolutely.
Quite a bit of opportunity there.
Absolutely. Yeah, yeah. Mm-hmm. Yeah. I mean, just to put it in perspective, you could have a one of our big CDN customers, say, one of the 1% customers, is probably spending 10x-20x on compute.
Right.
So, you know, $50 million customer is $500 million or more, maybe upwards of $1 billion. You know, we've heard a couple of our bigger customers tell us they're spending $1 billion in compute, so the market's huge.
It sounds like the gaining factor is a little bit more go-to-market than the CapEx, you know-
A little both. It's both really, right? It's, you know, you can't sell what you don't have, so you need both. But in terms of, like, building the CapEx, it's the lead times, we've been able to get down to 60-90 days. Now, at some point, I'll outgrow the existing footprint that I have with my core data centers, and I'll have to, you know, get more colocation space, and at some point, maybe I buy my own data centers. That's down the road, probably 5 years or so, but at some point, you might do that. So that is, you know, at some point, that becomes a bit of a factor, but there's a lot of room between now and then, though.
We are hearing that the demand for data center capacity is off the charts right now, probably.
Yeah.
Are you having any, you know, do you have enough visibility on supply for the next couple of years to meet your growth?
Yeah. Yeah, so we already are a very large procurer of data center space. We're also investment-grade credit, which when, you know, in the data center space, a lot of times you're going in with partners who are building out new projects. So we'll go into someone and say, "Hey, look, we'll take 5 MW or 10 MW in, you know, 3 or 4 years in your new project that you're building." And having us as a tenant, that's got enterprise, you know, to get that investment-grade credit, is very helpful for them as they finance those projects going forward. So we're an ideal customer, and we're already a big customer. You know, colocation is probably, you know, next to labor and, you know, depreciation, one of our largest expenses.
Got it. Got it. Very good. So maybe just switching gears to the... Well, I guess at this point, how much of the CapEx is going to be for success space versus, you know, building up the basic, basic infrastructure that you needed? Yeah.
I'd say most of it at this point. You know, there's some for the Gecko sites, but that's very small. Those are... You know, a lot of cases, we'll put those, just put some Linode cap, you know, rack or two of Linode capability in some of our larger CDN locations.
Mm-hmm.
You know, we've got, you know, some places in some larger countries, you have places where you have five or six telcos all coming together in big CDN PoPs. So you, you know, you could deploy there. That makes a lot of sense to do that. But that's not a ton of CapEx, so a lot of it is success-based CapEx, and we need some for ourselves as our, as we continue to grow, but, you know, that won't be dramatic relative to what we're going to be selling.
So, I mean, you know, obviously, CapEx is up a lot here lately. You know, any way to think, which, I think the street's okay with, as long as the revenue and-
Yeah
When you get to the margins are kind of coming in.
Mm-hmm.
Can you give us any help how to think about how CapEx should trend the next few years?
Yeah. So, you know, in the short term, certainly, you know, remember, you get about half of the CapEx, so we're 16% this year. Half of it is capitalized R&D. You know, that probably stays in the 7%-8% range for a while. As revenue starts to grow faster, we're not gonna continue to grow engineering at the same pace, so that maybe ticks down a bit over time. You've got CDN capacity, capacity tends to be low single digits, probably 3%-5%, you know, probably a little bit lower now that traffic's not growing as much. And then, you got 1% for your kind of back office, G&A, office spaces, and all that kind of stuff. But then the rest would be for compute.
So you're probably running 15-19, somewhere in that range, and it would kind of flex up if we're seeing a lot more demand for compute. You might get on the higher end of that. So I think if we see big delivery increase, you might have a 1% or so more for delivery, but really, the CapEx is gonna be skewed mostly by what we're seeing in terms of demand for-
... compute. Well, I hope you're closer to 20%, but good luck. You know-
Yeah, I mean, I'd love to have investors feel that way. Yeah, I mean, that's. And I think, too, if we do get large opportunities where we have to do custom build-outs, yeah, that's something that we'd probably tell you, right? We'd say, like, if I had to-
Yeah
take CapEx up a point or two because I'm - I just signed a multiyear, $10s of millions a year type deal, and it's concentrated in a few locations, I'm gonna tell you that. You know, I'll say, "Hey, CapEx-
Well, if you're getting up to 20%, then I think you're gonna be growing over consolidated revenue over 10% overall.
Mm-hmm.
But I guess the big question is, not to put words in your mouth, but just mathematically-
Yeah
... you know, how are the margins on the cloud business versus, you know, your consolidated margins, you know, at this point, and where can they be longer term?
Yeah. So there, you know, we had given some numbers a couple of years ago when we did our analyst day. We're seeing that they're, you know, at scale, we think they're gonna be in that range. I don't think it's gonna be that different, 'cause we're getting an enormous amount of scale on the go-to-market side and on the operations side. So now that we're not, we're turning away some of that big build-out, peaky CDN business, we're able to redeploy the people that build the data centers and run the CDN business and build the data centers that are required for compute. We've also shifted a lot of engineering resources. So folks that are, you know, you know, worried about all the compliance and keeping up with the, you know, the standards, doing that for CDN and for compute.
So, we're seeing a lot of engineering effort, transferable skills. So, you know, a lot of the stuff we've done internally, we've hired a bunch of people, but we've been able to repurpose a lot of folks. So we're scaling down all the investment in CDN, parking it over in compute, and not having to go out and do massive hiring. And I think from the go-to-market side, there'll be, you know, some shifts and stuff like that, but I don't see a huge investment in go-to-market required to do this. Nothing that is outside the norm, you know, maybe a little bit, but nothing substantial there. So you're gonna see a lot of scale in the operations, engineering, and go-to-market. So I think you can potentially see margins expand as the model shifts more towards compute and security.
You know, on the delivery side, we are getting a lot of questions, you know, is this even a business longer term?
Mm-hmm.
Revenues are down, like, 50% from the peak, almost, you know, a few years ago. You know, I know this year you're kind of getting hit with the perfect storm, that you had a lot of repricing your contracts and volumes have slowed down a little bit.
Mm-hmm.
But, you know, I guess over time, what do you think, is this a business that can stabilize at some point? And what enables that stability?
Yeah. You know, first of all, it's a necessary evil, meaning, you know, for the security business, having the delivery platform is incredibly helpful. All the information data you get to be able to block these big attacks, to be able to deliver the security that we do and deliver the same performance at the same time is very important as well. So I don't see it going away. And every cloud platform has some type of a delivery platform, so it does go in conjunction with your compute goals. So I don't see it going away, and I do think it can stabilize at some point, and certainly not decline double digits. And what you need for that is to see volumes pick up a bit.
We're way under trend line in terms of traffic growth, what we normally see. I think it's hard to say exactly what that is. You know, there's, you know, video traffic is a little bit sluggish. You could say maybe some of that's the economy, where people are canceling subscriptions, some of it's password-sharing crackdowns, maybe the writers' strike has something to do with it. You know, we're not seeing you know, there was a dearth of content that, you know, was, or, you know, not like new content not coming out on time and that sort of stuff. That should work itself out over time, but really, it's volumes coming down and pricing moderations in terms of the reprices should align with the growth.
So, therefore, you should get to a point where it's maybe single digit decline to flattish.
Apologize, I did get about 20 questions, which we're not gonna have time for, but, I'll just try to go through a few of them. This is from the audience. Will Guardicore and Noname be enough to continue to drive security revenues in the low double digit overall for the company, do you think?
Yeah. I think, you know, those businesses, Guardicore, obviously, is getting to a pretty interesting scale. We-- last we broke it out, we said it was 100, growing at 60%. It's obviously gotten bigger, and we haven't broken it out in a couple of quarters. Noname is about a $40 million business, roughly, and we guide it to about $10 million a quarter, give or take. That, I think you've got two market-leading products, and I'd say the adjacency for API security with our install base is huge. So that should grow very, very fast. So yeah, I think as those businesses get bigger, they should be able to contribute a lot more meaningful to, you know, contribution to the growth rate. The bigger franchises are still growing at a pretty good clip.
You know, obviously, those will slow down as you get, you know, more penetrated and that sort of stuff. So yeah, I think, I think we have the, you know, the right product set to keep that growth going, you know, in the low double digits for a while.
We're probably gonna run a couple of minutes over, if that's okay with you, Ed?
Sure. Yeah. So I was—maybe a little technical difficulty. Sorry about that.
Yeah, no, it's quite all right. A lot of good questions here. You obviously bought some CDN revenues last year in StackPath and Lumen. You know, do you think there's more opportunity to do that? Are you interested in it? Did it work out pretty well for you?
Yeah, good question. I mean, if the economics make sense, we'll always look at any opportunity that we think is, you know, could create value for shareholders. There, you know, the one thing in doing, you know, say, a big acquisition there is splitting in the market, so some of that will go off to other places. You know, the dynamics in the market aren't great, so it's not something that we're seeking out and looking at actively, but if something shows up, we'll certainly take a look at it and evaluate it. If it makes sense, we'll, you know, we'll think about it.
Then a question on the gross margins. I know we touched on it a little bit, but a lot of puts and takes.
Mm-hmm.
You've been running basically 73% here lately. Do you think that's a good run rate going forward?
Yeah. I mean, I would think at some point, we should get maybe a point of improvement. I wouldn't expect it in the near term. Maybe, you know, in another year or so, we might see that go a point, but I don't think it'll go dramatically higher from there.
Okay, got it. And, the operating margins have also been around 30%.
Mm-hmm.
Sorry, they've been a little bit lower here lately.
Yep.
A little bit below 29. When, when do we get back to 30?
Yeah, Tom said near term, so you know, not gonna give any guidance out, but, you know, hopefully, you know, in the, you know, near, near term here, we should be able to do that.
And then lastly, can you update us at all on what the Olympics means for the fourth quarter, what you saw, you know, from Olympics?
Yeah, sure.
They seem to be more popular than expected, I think.
Yeah. I mean, we, we gave guidance about $3 million-$4 million. So if it's more than that, it won't be dramatically more. You know-
Okay
... it's a pretty popular event, but I don't, you know, I don't think it's gonna drive a ton of upside.
And, have you had any more, and then, you know, back to the CDN, I mean, like, every two years, we get a lot of these repricing of contracts, so it does feel like next year... I mean, how much of an impact did that have to this year's growth? Can you give us a little bit more color? And I only ask-
Yeah
... because it'll help us model next year.
Yeah. So I would say it's a meaningful hit, right? 'Cause you've got 7 of your top 10 customers all renewing at the same time. And I don't wanna get into too much specifics because then I start talking about, like, what percentage discounts we get. But it was, it was a meaningful, meaningful hit. So, you know, next year, these contracts are generally 1 to 3 years, so you have a couple that'll renew next year, but you won't have 7 of the top 10.
Okay, got it.
So it won't be, you won't have that dynamic. I think the biggest question next year is, for the delivery business, number one, what happens with the government and our big social media customer? That's out of our hands. We've disclosed it for you. It's about a $50 million, give or take, impact if that were to go against us. That's in our 10-K. We weren't required to disclose, but we did. The other thing would be, what happens with traffic growth? If we have another anemic traffic year, delivery will be disappointing. And then, you know, pricing is probably the least of my worries with variables that, you know, isn't gonna dramatically change things. It's traffic growth and what happens with the government and TikTok, that's gonna be the-
I could talk to you for a few more hours, but we're out of time. I really appreciate you participating and sharing all the information. Thanks so much.
Thank you, Tim.
Thanks, everybody.