Hi, my name is Richard Choe. I cover communications infrastructure here at J.P. Morgan. I'd like to thank Charles Meyers for being with us here today.
Sure.
I just wanted to start off, I'm used to saying President and CEO, but now you're transitioning to Executive Chair. Can you talk a little bit about the decision to transition from President and CEO to Executive Chair and the hiring of Adair Fox-Martin to replace you?
Sure, sure. Before I start, I'll follow through on my commitment to Katie to tell you, too, that some of what I say will be forward-looking, and to check out our riveting disclosures on our website if you need to. But so yeah, Rich, actually this will probably be my last conference as the CEO. We expect Adair to be on board in early June. So it's been an incredible run. I expect I'll be pulled into service on the occasional pinch hitting, and so probably not my last one overall, but last one at least sitting in the CEO chair. So but it's been an amazing run.
I, you know, when I took the job, I think people will probably have seen some of the commentary and the video I posted on LinkedIn, et cetera, about this. But, it was, you know, when I took the job in 2018, it was, you know, I had sort of given the board a 5- to 7-year timeframe.
As you might imagine, at that time, the board was like: "Yeah, we'll talk about that later." And so, you know, as we came up on the five-year mark, and we were, you know, continuing, as all good companies do, to think through succession planning, et cetera, I said, "No, I still am kind of on that timeframe." And so as we sort of continued down that journey, it, you know, the opportunity presented itself to look at Adair as a possible, you know, successor.
And I think she brings a tremendous, you know, set of experiences to the table for us, and also happens to be somebody who was, you know, deeply, sort of involved as we continued to set the strategy, and very bought into that and committed to what we're doing, and I think somebody who understands the company and the culture and what has made us tick. And so feel really good about, you know, her, her next, the next chapter under her leadership, and looking forward to supporting her in a time that is as exciting as I think it's ever been in the data center business, and, you know, excited for what comes next.
It's funny because people have asked me, "Well, you know, how can you possibly jump off the train at this point, you know, given this?" But it's not really kind of how my brain works in terms of, you know, this more of the timing for me is less about Equinix and more about me and kind of, you know, my family and 5 grandkids, and I'm already starting to miss things, and I really would prefer not to sort of have that cycle, you know. And so, looking forward to being able to spend time with them, but also, you know, be able to continue with a company that I feel very passionate about and love the people and get to continue to do that. So best of both worlds from my perspective.
Well, and I'll follow up there, is that in the end, you're still Executive Chair-
Yeah
and you're still gonna be very involved. So as much as I think some people think you're stepping away, you're not really stepping away.
Yeah.
As you brought up, like, 2018 to now, the world has changed a lot, and then, I assume when you took the role, you saw a very positive outlook for data centers. And how do you feel about how the business has developed from when you took over to now? And let's not even talk about AI yet.
Yeah
but just the overall, where Equinix was to where it is now.
Yeah, I mean, I think that, you know, 'cause again, you could say, I mean, it's a 14-year-plus timeframe for me, you know, with the company. And I think that, over that period of time, I think we have been very consistent about what we think Equinix does distinctively well, the role we play in sort of cultivating and curating digital ecosystems, and the value that delivers for our customers. And so, when I took the CEO role, I was, you know, continued to be extremely convinced of the opportunity in front of us and the ability to invest behind that. I think we've done some really interesting things. Obviously, you know, kind of what we've done in has been, I think, very successful.
I think it was something that we, you know, we looked at that and said, "Hey, we want to continue to play in that space. We don't want to strain our balance sheet. You know, how can we solve this conundrum?" You know, I think we were. You know, we kinda led the market in terms of that structure, which is now kind of happening all around us. But, you know, a lot has been done over that period of time, and I think, but always leaning on, I think that this distinctive advantages we have in terms of our global reach, the ecosystems that reside in our, in our platform, and being able to continue to deliver that in ways that are consistent with the evolving needs of the customers. And so I think we've followed through on that.
I think we've been, you know, very agile in many respects, and I think the performance of the business has, you know, been very compelling over that period.
Something that I get questioned a lot of, people kinda see data centers as a commodity-
Yeah
and as something that isn't strategic. But Equinix has always kind of built its business on being, having real strategic value to its customers. Where do you see Equinix today versus what, you know, seven years ago or five years ago?
Well, I do think it's very important. One of the things that I think is most common, and you certainly see that here, depending on, you know, at investors, you know, conferences and other places, depending on how long people have followed a story or that kind of thing, is that the data center business is far from homogeneous, right? You know, and so, people talk about it as a, you know, as an industry or as a sector, and the reality is, is that, you know, it's a variety of sort of different subsectors within that. I think a lot of the energy, for understandable reasons, focuses on sort of the hyperscale opportunity, which has dominated the conversation and probably dominated the capital flows into the market over the last many years. But it is actually a business.
It's a part of the business that is more difficult to differentiate. It is, you know, it has a different return profile. It's still a very attractive business with a rich, and I think, very, enduring, demand profile to it. But it's very different than what we do in our kind of traditional retail business. And as I said, our role in that business has been around cultivating these digital ecosystems and playing a very different and more targeted role for customers as they think about hybrid infrastructure and the ability to deploy hybrid infrastructure in an increasingly cloud-centric world, and the role that we play in actually supporting the service providers who are delivering that value to customers, including the hyperscalers. In fact, we talked about that on our last earnings call.
I think people don't quite realize or maybe didn't fully appreciate, and that's why we gave the, you know, the numbers that we did on the last earnings call about, you know, the, in the retail business alone, not even counting xScale, we're well north of a billion-dollar run rate just with the hyperscalers, in terms of the important role that we play in their underlying architectures. And so, you know, I think that I think the demand profile for, you know, data centers broadly is as rich as it's ever been, and I think the specific demand around, you know, our more interconnection-centric, ecosystem-centric value proposition in the retail data center space is also as good as it's ever been.
I think that's one of the things that I run into the most, is that people are worried that... And I feel like we've been talking about this for a long time, that as the cloud gets bigger and now AI comes, that colocation and interconnection are gonna be somehow technologically displaced. And because of the amount of dollars going into investment in both the cloud and AI, people think your business isn't gonna grow as much, but we haven't seen that.
No.
So kind of what are you seeing in terms of cloud and AI growth for your business?
Well, look, I mean, cloud has been a major catalyst for our growth over the last, you know, many years. I think when we saw this trend early on in terms of cloud adoption and what the ,y ou know, we had, we were, I think one of the things we prided ourselves as always being objective about the risks and opportunities in front of the business, and that has served us very, very well. I, as we looked at it, we said, "Okay," it seemed fairly clear early on that the majority or super majority of workloads were gonna be destined for the cloud at some point.
I think rather than resist that or try to convince customers that, oh, they should not go to the cloud, I think instead, we have always had a value proposition that says, "Look, you know, workloads should go where they are best served." And instead, our role with the customer has been to simply help them understand what their needs are, understand, you know, if, you know, when their needs are different than what might be met in the public cloud, how to adapt to that, and then also how to, you know, eventually move to an environment where people have a variety of workloads across a variety of infrastructure types.
And that sort of hybrid and multi-cloud architecture has, I think, emerged in a very, you know, compelling way as the standard, you know, for enterprises. And I think we've been, you know, very successful in helping them navigate towards that. And I think AI, by the way, I think if you look at it, has many similarities to what we saw in the cloud over that timeframe. And I think, in the same way that hybrid and multi-cloud emerged as the architecture of choice, I think some form of hybrid AI is also likely to be the architecture of choice in AI.
People, I think, are right now, many people at least, using cloud-based AI services as their point of entry into AI, because it's easy to access, it's agile, it's flexible, it's variable cost in nature. But, you know, people that are operating at scale, are saying, "Okay, the unit economics, though, are better if I can do that on private infrastructure," et cetera. And so hybrid AI, I think, represents a really big opportunity, for us, and we're already seeing that, you know, in our pipeline, and not just in xScale. xScale, we are seeing it, in terms of serving the hyperscalers whose demand is, you know, kind of, you know, voracious, to say the least.
But I think we're also seeing it, you know, as enterprises think about the AI opportunity in front of them.
I think one of the things that you have mentioned over time, or more recently, is your ability to serve, I guess, this AI hybrid architecture, both from your normal, I guess, air-cooled and existing offering, but now you're enabling liquid cooling.
Mm-hmm.
Can you talk a little bit about the new services or?
Sure.
or ability to serve that?
Well, I mean, I think one of the misconceptions has been, oh, well, the, you know, the, the-- we're seeing a rapid increase in the sort of power densities of new technologies, particularly on GPUs, and therefore, can these older data centers accommodate those, and the reality is that the answer to that is yes. And while they-- it does represent, you know, particular challenges from a cooling perspective, our ability to implement liquid cooling, which is now available in 100 of our facilities around the world, in most, in almost all the markets in which we operate, that allows you to, you know, be able to accommodate those workloads.
Now, it does change the mix of business that will exist in those over time, and so you may end up with, you know, fewer net customers or net deployments consuming the same underlying IT load. But, you know, as long as you can monetize it at the same or better returns on capital, then that's a, you know, an eventual net gain for your business. And so, that's where we are, and we are ahead of the game in terms of having commercialized, you know, liquid cooling technology solutions across our footprint. I was just in Paris last week visiting our, you know, our liquid cooling point of proof of concept in our Paris TEN facility.
And again, I think we're having, you know, putting customers through that thing on a very regular basis and starting to see meaningful uptake on that.
And something you mentioned on the call, but something that investors have been concerned about was the number of cabinet growth.
Yeah.
I feel like that's a metric, for better or worse, that has been your main metric, but it seems like now, to what you're saying, that, it might not be the best metric to focus on in terms of, where your business is going.
Yeah.
How should we think about cabinets versus-
I don't think it's that it's not the best metric. I just think it needs a more complete view and one that is informed by the changing dynamics in the business. So that's why we gave everybody the math required to think about how densities are changing, how there's a delta between sort of the density of cabs that we're churning. Because there's a fair amount of what we were, you know, we'd talk about as frictional churn in the business. As that churns out at lower densities and you replace it at higher densities, you essentially create a bit of a hole in your cab count. So it's not that I think that cabs is a bad measure, I just think it's one that we need to, you know, moderate in terms of people's expectations.
The bulk of our growth has come with more modest cabinet growth because of this dynamic of evolving density, but that is driving our, you know, our MRR per cabinet yield. And so the combination of those two things is what you need to be concerned about. But I think our ability to deliver the revenue growth that underpins our guide and our future plan is very, very healthy. And so and I actually think in many respects, cabs are the right underlying metric for us versus kilowatts, because we, unlike others, deliver...
18% of our revenue comes in the form of interconnection, and so you have to really look at it on a per cab basis, because if you sell a kilowatt, you know, if you don't have the additional services to add to it the way we do, it's a, you know, it's a very different return on capital, whereas the health of our interconnection business is a major driver of our market-leading returns. And that's why I think cabs work. We just had to, have had to kind of modify, moderate people's views of what that, what that growth would look like because of this density shift that's occurring. And so, and I, I think people now, for the most part, in meetings today, as I've talked that through, people, you know, are, are sort of, grokking that.
So it seems like people are getting more comfortable with kind of balancing the two, and that your underlying business of interconnection and cabinets are both important, and that seems to be growing at a good pace. You had another healthy quarter of signings.
Mm-hmm.
Has that trajectory changed in recent months, or kind of where are you overall in terms of your demand profile and pipeline?
Yeah, I mean, as we said on the call, I think we have an extremely strong forward-looking pipeline. We had a record Q1 as it relates to. It's the best Q1 that we've ever had from a bookings perspective. You know, and I think we're seeing digital transformation as an underlying driver of demand for digital infrastructure at peak levels. You know, people are very, very committed to that, and AI is an additional stimulant. I do think people are trying to weigh the budgets between how do they implement AI without, you know, sort of net increases in their budgets. And so I think there's some, you know, some trade-off between those things.
But, I think that people continue to see what we offer as a very compelling part of how they want to think about hybrid infrastructure going forward and increasing the agility that they need in a sort of this multi-cloud, hybrid multi-cloud world. And, so, overall, I think demand, underlying demand, is as strong as we've ever seen it. There are a lot of dynamics in the business in terms of... You know, I still think that, you know, we've seen a fair amount of optimization. Still, people are working through, I think, their cloud strategies in terms of, you know, what level of cloud is appropriate for them, how do they implement hybrid cloud, and where, when, how?
And I think we're, you know, we're working actively with people in that, and I do think we've seen, you know, a level of, a level of optimization around infrastructure decisions. You know, but, but I also think we're seeing the front edge of AI demand in a very compelling way. Again, with service providers, for sure, but now beginning with enterprises as well. Right now, more on the training side, but I think with a wave of inference demand coming over the next couple of years, that I think will be an even more compelling area of differentiated, you know, demand for Equinix.
Do you feel like that inference demand has started even yet, or?
It has started, for sure. I just think it is, I just still think it's, you know, it's relatively early.
Okay. You did have a delay in reporting your quarter, even though-
Yep.
I t was a strong quarter. It's good to hear that, I guess, the team was still focused, but there's been, I guess, other press reports out there about some of your accounting and-
Sure.
issues. And so can you walk through us kind of what went on and how you-
Sure.
- dealt with it?
I mean, we talked at, we talked in at some length about it in our last earnings call. But, you know, when the, you know, when the short report came out, you know, I think, you know, while we as a management team have a great degree of confidence and comfort, and a lot of the things that were sort of talked about in that were not new news in any way in terms of the emergence of cloud and the potential risks and opportunities associated with and various other factors.
But, as any public company needs to do, when there are allegations made regarding your accounting, and the audit committee needs to sort of take that situation seriously, we did exactly what you would expect us to do and launched an independent investigation. We have an incredibly experienced audit committee, you know, that I think when you look at their backgrounds and the level of depth that they bring to the table, I think you can look at them with great confidence as doing what an audit committee needs to do. They hired a leading set of independent advisors.
We, as a management team, stepped back and said: "We're gonna focus on trying to serve our customers and keep the business moving forward in the meantime." I was very confident of where we would end up. We did end up where I expected we would, which was with a, you know, kind of a strong reiteration from that investigation that our accounting practices are an accurate and solid representation of the performance of our business, both on GAAP and on non-GAAP measures. So I think now we're prepared to move things forward and continue to focus on delivering great value for our customers.
No, that makes a lot of sense. It's not like you've been doing this for a shorter period of time. Like, we talked about your experience, and so-
Yeah.
I t's not like a lot of these issues would've been new.
No.
Like, they've been around for a while.
Yeah.
So as you've been building the business and you have a very substantial development plan, how do you kind of manage how much you want to grow, how much capital you're allocating to the different regions, and the planning? Because there, there's a considerable amount of planning that needs to happen. So as you look out and you have your development table, kind of what factors are you kind of considering-
Yeah.
- the most?
Well, I mean, the reality is that our development pipeline is driven largely by, you know, the expressed needs of our customers. And so, you know, we don't really sort of guess where we're gonna go. We go where our customers tell us they need to be. And the nature of the distributed infrastructure, I think, has changed pretty meaningfully over the y ou know, it used to be that people would have potentially globally centralized infrastructure, and then it was sort of regionalized infrastructure. And now, I think, in a substantially more application-centric world and a more cloud-centric world, people have distributed infrastructure very, very widely.
Our footprint and our development pipeline, one, is about satisfying the demand signals that exist in our existing markets, and then, two, entering new markets where we believe there is sufficient, you know, already communicated customer demand for us to go into. Yeah, I think it really substantially de-risks our entry into new markets, just given where we are. I think our, you know, our history of being able to know that, you know, some number of our customers are gonna automatically put infrastructure into markets where we, you know, where we enter, you know, makes it a lot less risky in terms of how we underwrite that. But, and then, you know, and then we think a little bit...
We obviously, our capital plan also includes our xScale business, where we have now, you know, leveraged other capital from financial, you know, from capital partners that join in that with us. And I think that, you know, that is another driver of our overall capital plan. But it allows us to grow in that business without straining our balance sheet, which has been, you know, really important to us, and I think proven to be incredibly successful.
I think we keep waiting for the U.S. and Europe to slow down because they're very developed markets, but it seems like they haven't. It seems like one of your main focuses has been Asia. Can you talk a little bit about the opportunities you are seeing in Asia?
Yeah, I mean, look, one of the great things about our business and our performance over the last, you know, many years has been the consistency of that across our regional footprint. So it really hasn't been, you know, from quarter to quarter, you've seen, you know, kind of different regions perform better, and so but it has been, you know, overall, it's been a very consistent level of performance. There was probably a little period of time immediately post the Verizon acquisition, where the Americas growth was suppressed. But as we kind of worked through, you know, kind of the integration of that and the optimization of that asset, we saw a pretty significant re-acceleration in the Americas, and then have seen pretty consistent performance across the regions.
As it relates to Asia, I think, you know, it is a more complex market and one that has had a number of moving parts in it. Specifically for us, I think the, you know, the capacity constraints that have existed in Singapore, I think have been a little bit challenging in terms of, you know, the growth in that, in that market. But it continues to, you know, so we're living a little bit more with the capacity that we have, but now have line of sight to incremental capacity, which has been granted by the government there. And then, but we do have this strong Pan-Asian footprint, which is, I think, a unique and distinctive advantage in that, in that market and continues to, to serve us very well.
So, you know, customers, you know, more than 80% of our revenue comes from customers who do business with us in all three regions of the world, that our operating regions across EMEA, APAC, and Americas. And so people feel like we're a very safe pair of hands to drive their expansion across Asia as they see a lot of opportunity, and so I expect it will continue to overindex from a growth standpoint.
Something that I think got lost a little bit in the quarter, because of dealing with a lot of the other bigger metrics, was your growth in digital services.
Mm-hmm.
Like, your incremental, you know, as you mentioned, like, a lot of it is interconnection, but you've also grown your suite of services that you're selling to customers.
Yeah.
Can you talk a little bit about how that has developed, and where do you see that going?
Yeah. I mean, I think that, you know, the way we kind of think about digital services is it's a different form factor for the same underlying value proposition. The underlying value proposition that people find compelling about Equinix is our global reach, advantaged access to the ecosystems that are important to them, an interconnection platform to facilitate this hybrid multi-cloud architectures, and then a track record of delivery that we've offered. And what digital services is, is simply a way for us to deliver those same underlying services with less friction and with more, you know, variable cost, on-demand capability that looks more cloud-like in nature than what the traditional colo approach has been. And the reality is, most of our customers do both.
You know, there are certain instances and circumstances under which traditional, you know, colocation is their preferred method, and but then the digital services have been, you know, are more agile, more on demand, and have met, you know, specific needs for them. And so if you look at Fabric, which is Fabric just really a virtualized platform for interconnection that moves beyond the traditional physical interconnect, right? You know, and that has been the most successful new product in the history of the company, continues to be incredibly compelling in terms of its overall growth profile.
Now we've added to the digital services portfolio with, you know, not only sort of interconnection and networking-type offerings, but also our metal offering, our network edge offering, and other things that I think are allowing people, you know, to have a more cloud-like experience with the... but still, you know, sort of capture the underlying Equinix value prop.
Do you think that changes as we move into more of an AI world? Do you think your digital services will become more valuable in an AI world?
Yeah
or, and-
Well, I think both will continue to be very relevant, so I think it'll be different tools for different jobs.
Mm-hmm.
So if you look at what's happening in AI right now, we're seeing probably more demand in the traditional colo, where people want to deploy DGX, for example, in a dedicated setting. You know, these are larger enterprise customers who want owner economics and who want the performance that comes with it. And they want to implement it, you know, on private infrastructure. And so that's what we've seen. We have a very healthy pipeline for that. But then you also, I think, are going to see people who want a more, you know, like our cloud-adjacent storage offerings, which are more in the digital services portfolio, is about being able to, you know, place data where you need it, intercloud, and with proximity to the cloud.
You know, and because that's the most frequent question that we're hearing from customers is, okay, the first thing they're talking about in AI is: What should I do with my data? And I think that, you know, the digital services proposition on cloud-adjacent storage and what we're doing with our partners, whether that be Pure or NetApp or Dell or HPE, is saying: Look, you decide what technology provider you want. You decide where you want to place your data. But if you want it to be not in a cloud, but proximate to the cloud, then Equinix is a perfect solution for that.
I think that's an area, really our digital services of areas of focus have been, you know, sort of the evolution of interconnection and multi-cloud networking, and then this cloud-adjacent storage offering, both of which I think are seeing, you know, good uptake in the market. And again, those sort of more cloudy, you know, type digital services for us are actually meaningfully overindexing on growth relative to the broader business.
Oh, that makes a lot of sense. Something that I get a lot of questions on, and I think people don't always appreciate, is how latency and connectivity are getting just more and more important. And then it seems like in a more AI world, the connectivity needs to be faster and better, and more of it. Is that what you're seeing trend-wise, or-
Well, I-
Do you think it changes-
Again, a little, a little bit of the answer, again, is it depends. Because, like, on large-scale model training, much of that can be done in locations that it's a less latency sensitive application.
Mm-hmm.
Now, I, in terms of eventually when you distribute that, you know, that model, and do inference, depending on the nature of that and what, how real time it is, and, and who it's, you know, serving, what customers it's serving, that's when I think you start to see, you know, some of these same latency and performance benefits that distributed infrastructure provides. And so, so it, it really depends. I think we're seeing some of both of those things, but I definitely think that, you know, you know, the realities of distributed infrastructure and the need to have a, applications perform, you know, for the users to, to whom they're relevant, has been a, is, is a key driver of why customers, you know, choose Equinix and, and, and get a different differentiated level of performance.
Something you touched on earlier, there's a lot of demand on the hyperscale side. It's taking up a lot of resources, a lot of capital. You participate in xScale a little bit, but, as you look out now, and you see your resources and capacity, and it's becoming more scarce, how, how are you approaching pricing and the ability to make sure you're getting the proper return on all the investments you're making?
Well, I would say that there, again, this is back to the segmentation of the broader data center opportunity. I would say that pricing in the xScale market and pricing in the retail market are two very different things. And, it, you know—and so in the xScale market, I do, you know, I think it's largely driven by sort of supply, demand, you know, kind of, stability. And, but it is one where I think there's less, you know, that because you have a very concentrated set of buyers, and they, you know, they dictate price in a more, you know, in a more powerful way, I guess.
I think that, you know, it's still a market that has, you know, attractive returns, but, you know, but is one that, you know, I think has a different level of pricing power available to it. In the retail market, I think pricing is, you know, is different. And I think we have, you know, it's much more price to value. And, you know, when we talk about pricing in that market, it's not just about list pricing, which I think we have seen rise over the last several years, but it's more about yield. And it's why our MRR per cabinet figure, which we talk so much about and which has performed so exceptionally well, is really an indicator of pricing yield.
Not necessarily just of list price, because pricing yield in terms of MRR per cab is driven by list price, by interconnection density, and by, you know, power density. And all three of those contribute to what has been a very rising tide, as it relates to MRR per cab.
I think on your last call, you called out the power density increase a little bit more than in the past. Do you feel like that's a tailwind for that MRR as you kind of move forward?
It is a tailwind for MRR per cab, but I would say it's just a shifting dynamic that is, you know, it's still part of, I think, a generally stable return on invested capital profile. So it's not like, you know, higher density implementations return at a much higher return on capital level. I think it just, the form of that looks slightly different. It takes more power in smaller spaces. You know, and so I think, you know, exactly how we design data centers going forward, I think we're likely to build phases with more power, but probably, you know, not as much space. And that's just a differing dynamic. But the underlying economics and the unit economics of the business remain pretty consistent.
And so you feel very comfortable about your ability to build at much higher densities at this point?
Yeah. Yeah, I mean, Well, not only do I feel comfortable in building new facilities at higher densities, I feel very comfortable in delivering implementations in existing facilities at higher densities as well.
Great. I think we'll leave it there.
All right.