Equinix, Inc. (EQIX) Earnings Call Transcript & Summary

[Audio Gap] for Global TMT Conference. For those of you I haven't met, I'm Mike Rollins with Citi Research, and we're pleased to welcome Equinix's President and Chief Executive Officer, Adaire Fox-Martin. Adaire thank you so much for joining us.

No, thank you very much for having me. I'm delighted to be here.

Before we get started, just a few things of note. This session is for Citi clients only, and disclosures are available at the back of the room next to the AV desk. And for everyone in the room, you can use the QR codes that are up on the screen or up on the side, and there are also some in the rows, to participate in our live survey. And your participation, your submissions are anonymous, so we'll encourage your participation for this. And with that, again, Adaire, thank you so much for joining us.

You're very welcome.

Well, maybe to get us started, you've been spending a lot of time from what we understand with employees and customers around the world. How are you seeing the operating strategy for Equinix evolving to enhance your financial performance and enhance shareholder value?

So my legal team have told me I have to make the forward-looking statement that some of what I'll talk about today is forward-looking statements, and please check out our SEC filings in order to understand the details behind there. So I just passed day 90 as CEO of Equinix. So it's been a wonderful journey for the last 3 months. And as you said, I've spent most of that time on the road out on a listening tour, meeting with our customers, our partners and, of course, the team. Understanding the dynamics of the business and very much working to keep an outside-in focus on those dynamics in order to ensure that we're delivering maximum value for customers, partners and, of course, for the important cohort of our shareholders. I guess in terms of learnings, there's been quite a lot that's occurred as a result of those conversations. I think with our customers, an interesting lens is for me to have understood just how broad the demand set is. There is, of course, a huge AI narrative in the market now. And of course, Equinix has a role to play in helping customers to deliver against their own objectives there. But I think on the whole, the demand set in customers is broader than AI, and that is something that is a good news for us. In terms of the key priorities of our customers, some of it is absolutely around the cloud journey and continuing the cloud journey. Many of them are still migrating, right? There's still a journey ahead of them. Many of them have become much more au fait with this technology than they were in the past. And so we're absolutely seeing questions about workload orientation, which applies as much to AI as it just to core workloads in the cloud as opposed to which cloud I migrate to. And this multi-cloud environment, this on-premise and cloud narrative, this migration journey are all things that I think will help underpin the solutions that Equinix offers going forward and therefore, continue to help us drive the revenue in our customer base.

Great. Well, I think that gives us a great segue into our first survey question that will queue up for our audience. And we've been asked by a number of our investors about your role and the potential benefits from AI. So we'll introduce this. How will the emerging demand for Gen AI workloads impact Equinix's multiyear financial prospects? Is it positive, neutral or negative? And so while we get our audience to respond, we will maybe just continue on with more of the broader strategy and wanted to delve into a little bit more of one of the comments that you made on the earnings call. So I think you mentioned the opportunity to enhance the go-to-market. And I'm curious, just as you think about simplifying Equinix from a go-to-market perspective, what does that mean? And how might that be different to what Equinix has been doing in the past?

Yes. So I think from a go-to-market perspective, we're very fortunate that today, there are more than 10,000 customers of Equinix using our facilities right across the world. And of course, we continue to serve the 10,000 customers that we have, 10,000-plus. And of course, we need to acquire new logos in order to continue to grow our business and grow our footprint in our existing customers. In order to ensure that we have the most effective and efficient go-to-market engine, I think the starting point needs to be the segmentation of your customer base. So actually understanding the cohort of customers that you have and then how best, from an operational perspective, you serve them. Because the cost to serve cannot be the same for someone who is generating multiple millions of revenue for you and somebody who's generating tens of thousands. And so we are in that particular process right now, redoing that segmentation. And that will allow us to look at how we engage more effectively. And allow us to look at opportunities, for example, for the channel, for distribution, for a lower cost of sale, which will give us broader reach and at a lower cost and ultimately then impact in a positive way the margins of the company. So starting with segmentation and then growing from there to understand the cohorts of customers, the best way to serve them, implement that with the right degree of system and process so that we can scale without the necessity of always throwing people at a challenge, throw a system and process at it instead.

Let's take a look at our survey results.

I'm interested to see this.

So 100% of response is positive. So as you've had a chance to have this different perspective on Equinix now being in the current role, what have you learned about how Equinix can benefit from AI, both from a revenue and cost perspective? And I think one of the big questions that investors ask is timing for Equinix because of the role that you play within the ecosystem.

Yes. So maybe let me divide it up into a number of different elements. First of all, I think that the plans that we have are balanced and ambitious, right? When I look at how we play in the hyperscaler space, we do that through our joint venture structure that we call xScale. This was a decision that was made prior to my time with Equinix. And of course, when you have a fresh pair of eyes, it's always good to check the hypothesis on which some of those decisions were based and see if you would refresh them or no. The off-balance sheet scenario allows us to participate in the opportunity with hyperscalers, allows us to manage those unique relationships in a very particular way and also allows us to maintain a point in the supply chain of data center design, innovation and build that ensures our ongoing relevance to that cohort of customers. This is something that we're continuing to grow. In our Q2 earnings, we announced our first xScale facility in the U.S., which will be multi-hundred megawatts based in Atlanta. We have more coming soon on this topic. So xScale allows us to deal with the training requirements primarily and the large footprint requirements of the technology players and of the hyperscalers. Then in our retail business, there are a number of use cases that play out. First of all, on the retail side, across a number of our data centers today, we are ready for enterprise-grade workloads associated with generative AI. We already have some in action in our data center environment. Our partnership with NVIDIA is playing out beautifully in the context of the enterprise space. NVIDIA, as many of you know, offer 3 or 4 different components in a stack to support the execution of NVIDIA technology for our customer. Whether that's their software platform, whether what they call the silicon brains, the chips or whether that's the networking capabilities. The opportunity with Equinix and NVIDIA is that Equinix takes all that complexity, those three pieces; it takes the implementation and the systemization of those and implemented on behalf of a customer so that a customer can simply execute. And a customer can choose to do that either on a CapEx or on an OpEx basis, but we take the pain way of making that available. We have a number of customers up and running already on that environment because essentially, the Equinix and NVIDIA partnership is providing the opportunity to have private AI, private learning, private LLMs operating in our environment. A few of the first customers are in the health care, pharma space, where much of the model work that they are doing, whilst it is associated with the life-changing outcomes for their patients and customers, is very proprietary in terms of the research and development knowledge that has been utilized. So Equinix in conjunction with NVIDIA has become the place for private AI can happen for enterprise. And then the third piece is the piece where you take the proof of concepts that many companies today are in the middle of. And you take those proof of concepts and you move them into our production environment and make them part of the underpinning systems in your business. Those -- that part will absolutely require data. It will require being at the edge. It is the inferencing element, and this is where I think the strengths of Equinix that we bring today in terms of our global footprint, in terms of our colocation capabilities, in terms of our interconnect capabilities are on-ramps to so many of the different cloud providers will add value to customers as we move into the inference space. And of course, this is something that's a huge opportunity for us going forward. Now in terms of time line. How will that potentially play out? Well, obviously, training is now. We can see much of that happening both in retail and also on the xScale side. From my conversations with our customers, many of them are working through the business case, the actual business value that taking this proof-of-concept into production will release for them. And they're not always easy business cases to resolve. My take is that it will be a 12- to 18-month journey before we see the significant impact of that. And that's also, I think, something that the Gartner Group have verified in terms of their thinking. They just recently released a report that said by 2027, 80% of traditional enterprises will have at least one generative AI use case in their mainstream business process, which is a huge increase from the single-digit percents that we saw at the start of 2023. So good to see the positive outcome there and it's also something that we're working to ensure that we can deliver for our customers.

A couple of follow-ups. So with the relationship and opportunity with NVIDIA, who is selling the product to the customer? Is it indirect Equinix, NVIDIA? So who sells the product? And then what's the revenue for Equinix? Are you only focused on the colocation, power interconnection or are you doing more of a managed service for this type of product?

Yes. Great question. Thank you. So Equinix and NVIDIA sales teams are both jointly goaled on this opportunity. So our sales team are out actively supporting and presenting this opportunity to our customers. You're right that there is obviously a co-lo space and power coalescent that would come to Equinix in the way of revenue. But also, we do have a small managed services team and that team are also supporting this system managed service revenue coming through. It's still early days for how that pans out though, to be truthful.

So it's still evolving.

Yes, yes.

And when you made a mention earlier about how you're looking a little bit at the xScale with fresh eyes, what I remember from when Equinix launched xScale, I think the argument was there is a potential magnetism to doing it, where Equinix could do well with getting xScale business because of the relationships they have in retail. And then having more of the hyperscale within your ecosystem can give you more shots on goal for the enterprise side to get more share there. Are you seeing those benefits of, call it, magnetism on either or both sides of that equation?

I believe so. I mean, I do feel that there's a flywheel associated with even being associated in a meaningful way. And I feel that there is an opportunity to consider not just relationships that are supplier and customer, which is, I guess, where we are today in that we're a customer of a number of the cloud service providers and they are also a customer of ours. But also to look at potentially the opportunity for our 360. So where there are go-to-market opportunities where Equinix could work more closely alongside the hyperscalers and vice versa. So that's something that we're exploring.

So we'll go to our next survey question, as we get ready to talk a bit about the core business, and we're going to go to this question: What are expectations for normalized constant currency revenue growth ex the power price increases for 2025? And this is relative to your target guidance for 2024 of 7% to 8%. So we're asking our group here under 6%, over 6% to 8%, over 8% to 10% or over 10%. And so we'll see what the growth expectations come back from our group in a couple of minutes. But as we get started with this, one other question that's been coming up is this, within the realm of AI is, is there a risk that certain of your data centers could be considered obsolete and unable to serve the higher power density requirements of these AI workloads?

So I guess, over the past 3 months, I've lost count of the number of data centers that I've walked through. And it's been an interesting journey, particularly, for example, in the San Francisco campus, where I was able to see the very first one that was built 20-plus years ago, one that was built 10 years ago and one that's just brand new. So you can almost see data center innovation over a couple of decades across that campus in action. I don't believe that we run the risk of having obsolete data center assets. First of all, my observations are that we have an industry-leading data center design and development team, who have done an incredible job over the years and continue to innovate on data centers and data center infrastructure. We have centers of excellence that are embedded in different parts across the world that look at the core infrastructure today and wonder and define how we can identify challenges of today within that core infrastructure. So for instance, in over 100 of our facilities, we will be offering and are offering liquid cooling, whether that's direct to chip or rear door, that cooling is cooling that has been placed into data centers that our production data center, so a lot associated with that particular activity. But that data centers that when they were initially designed did not envisage this kind of technology being part and parcel of their physical infrastructure. So I believe that we will be able to work through our data center footprint to ensure that we are able to offer the kind of density that is required to support the workloads associated with the AI journey of our customers.

Very helpful. And let's see what are responses are and then we'll maybe talk about the core revenue growth opportunities. So the results: So 14% at or below 6%, 43% above 6% up to 8%, 14% above 8% to 10% and 29% above 10%. So that's a bit of a spread in the responses...

I'll tell you in February '25. When we finish this year, we'll guide for next. Yes.

And of course, we respect that you're not giving guidance for '25 here today. But as we get ready for that, what are you seeing -- you made some comments on the earnings call in terms of pipeline, demand, backlog for the core retail services, what are you seeing? And is the flavor of that demand starting to change in terms of size or verticals in terms of what's making up, whether it's the backlog, the pipeline or the bookings that you're seeing?

I guess as we mentioned on our earnings in Q2, we had one of our largest gross booking quarters ever. We had record sales in terms of kilowatts. Our interconnection remains strong. The revenue is certainly going up there. Our network traffic remains strong. So -- and our pricing was very strong in the quarter. So I think very many of the measures of the business have a trajectory that is positive. When I look at pipeline and backlog, when I look at backlog in particular, we can see that we have a strong backlog, particularly as it relates to cabinets, cabinets that have been procured but not installed yet. So I think that's a good wave of momentum that we would experience in the second half. And specifically, as it relates to pipeline, our pipeline is exceptionally strong. It is, I have to say, exceptionally strong in Q4, strong in Q3, we've got a month to run, and the team have been executing well against the opportunity, but strong pipeline, as we mentioned in our Q2 earnings, particularly in Q4. And so we will have a hockey stick to the end of the year, the opportunity to execute against that. In terms of the makeup of the pipeline, we still see very strong uptake of our typical retail footprint. There is definitely an uptick in the increase of requests for large footprint installations from our customers. And we're working to navigate that as best we can. And we do have some of our metros where we are capacity constrained, and that was something that hit us, I think, in the first half of the year. Most of those metros are bringing on capacity in the second half. And I think what's very important for us as a team is to engage early with our customers so that we can actually demand shape the requirement of our customers. Because not all demand needs to be in a Tier 1 metro located in [ DX ], whatever, right? So we are really working with the team very early on to demand shape because it is workload based and to enable the customer then to move perhaps where that capacity is available. So we are absolutely seeing that. From an industry point of view, a strong demand in manufacturing in health care, very strong demand amongst the media and gaming companies. And of course, the cloud is still strong demand signals from that.

As you look at the core retail performance, when you look at what's been impacting growth for 2024, optimizations come up, it's come up in a couple of ways, some revenue churn, cabinet volumes. And so as you're looking at the business, can you share what you learned in terms of what's happening with this optimization? And if you're able to see where there is some natural sunset or normalization of these types of activities, both in terms of the dollars of MRR churn, but also the cabinets, which is your volume metric in the P x Q math for your business?

Yes, exactly. Yes. So this is, I guess, an area of ongoing debate and also something that for -- specifically for Q2 earnings, it was something that I had a very long hard look at. I think there are a couple of things that have added to specifically the cabinet metric. Our power density, I mentioned already. But I also caveated that we see capacity coming on for those metros that were impacted by power density. Secondly, we also could see, aside from capacity that in the Q2 period, we had some churn that was like unexpected at the tail end of the quarter and that impacted us, and that's obviously a once-off. And we can also see that customers are looking to ensure best value outcome for them. I think there was some pent-up demand post-COVID for this. I think during the COVID period, whether it was on the cloud side or on the data center side, people were just happy that they were running, right, in a virtual environment, and there was very little optimization done at all during that period of time. Having looked deeply at the optimization levers, obviously, one impact on our results is density, which I think we've discussed ad nauseam at the -- our Q2 earnings. But when I look at the data itself, I can see that much of the churn is within, a, number of our customers. It is predominantly in one market, which is EMEA. And it has predominantly occurred in a cohort of customers in the network services group. And when I look at that, specifically around the interconnection grooming that we've seen, I can say that that's -- or at least for the top 10 has tailed off, and that we should see the impact of positive momentum going forward based on what I told you about our backlog and so on into the second half.

And the other part of this is power, and how do you power pricing for customers. And you mentioned EMEA, which is where there's been a significant amount of the pass-through that's taken place. Are you getting any feedback from customers in terms of how they like or maybe are not satisfied with the way that you manage price -- power pricing for them? And is that something that also as that starts to normalize helps Equinix?

Not specifically, to be truthful, certainly not in the conversations that I have had. I think there's been a degree of transparency about power price increases as much as there have been decreases. And during the course of '24, there's been a return to customers when the decreases have occurred. So no, I haven't specifically had that particular narrative from the customer.

We'll go to our third survey question here, we'll queue this up. And it is, do you expect data centers to sustain positive annual pricing power for colocation? And it's a simple yes or no response. So you can use the QR codes and poll in for that. Before we get to this and the story of pricing, maybe we could talk a little bit about competition. How do you see the competitive landscape for the core retail services and has it shifted in some way over the last 12, 24 months?

So I think paranoia is always healthy. And when I look at the competitive position, I think there are absolutely some uniques that Equinix has built over the years, the 25 years that we've been in business. In many ways, the success of Equinix as a retail colocation provider has been the exception, not an or. We have a huge deep footprint in the North America theater of operation. We have a significant global footprint, so it's not an or, it's an and. And 64% of our recurring revenues comes from customers who are with us in more than 3 regions. So the global footprint and its value proposition to customers is something that is meaningful to them. So in terms of competition, it varies because it can be very specific to the location. In some markets, there are national providers, local providers who would compete with us. I don't think there is any provider that competes with us at the same global scale that we would have. But we're always very cognizant of playing to our strengths and selling to the strength that we have as opposed to looking for your competitors' weakness. That doesn't mean that wouldn't be paranoid about our position and recognize that it's something that we need to continue to nurture. The core is super important to our future growth, continuing to grow our global footprint, continuing to grow the depth in the North America market is extremely important, and we will continue to do that. So I wouldn't take it lightly, speak with a degree of humility when we speak to it, but we do deliver our high-value service to our customers, and that is a service that is appreciated by them.

It feels like in some technology services, companies want to have multiple providers whether it's for redundancy or certain efficiencies. And then other times, they're very happy with a sole provider. What do you see from your customers in terms of you being their sole provider or being one -- whether it's primary or otherwise, one of a multiple set of providers?

The technology landscape, I think, has changed a lot in the last decade, specifically with the advent of cloud technology into the mainstream for most businesses. And actually, in many ways, that has added to the complexity that customers need to manage. Because the narrative around cloud is not necessarily choosing a single cloud provider for everything that your business requires for making a decision by workload. So that might mean that for a workload, for instance, that is a high compute workload that may go to AWS, given the price point that they offer that at. And for something that requires a high degree of analytical capability that might go to Google Cloud given their functionality and capability there. So most of our customers are managing already a complex environment in terms of a multiple cloud approach. We then also have to consider the on-premise nature of some of the applications that sit within a customer's landscape. So whether you're a highly regulated business, in which case you can't move to the cloud in some instances, or there are reasons for latency, speed, security that you want to keep a piece of your IT landscape on-premise, that's also another factor for our customers. So for me, it's about making it easier for our customers, reducing the complexity of managing these multiple landscape environments. And so perhaps you can think about it as easy at Equinix, right? Because in many ways, we are the point, where the physical world meets the digital world and where there can absolutely be an adjacency between the physicality of the environments that we operate in our data centers and where digital comes to meet that. And I see this as a number of bridges that we can support our customers in navigating, the bridge between clouds, the bridge between the core and the edge, the bridge between on-premise and cloud, the bridge between the LLM and the data. So these are, I think, all interesting adjacency points for Equinix to consider as we move forward.

Great. Let's see the survey results. So 100%, yes, they expect data centers to sustain positive annual pricing power for colocation. What are the pricing dynamics that you're seeing? And how do you look at pricing power in terms of the opportunities to monetize your platform better, but also delivering more value to customers in the process?

Yes. Well, I think the last point of the question is probably the most important one, ensuring that we are delivering value to customers because then we create the outcomes that help us then ultimately drive value for our shareholders. We are seeing, of course, we know that often when we churn, we can churn for a higher yield, which when we think about the concept of churn, we have to think about it not just as negative. First of all, most of the churn that we have experienced is partial, right? It's not a customer moving away from us. And secondly, there is often the opportunity then to get a better yield for that particular cab and that particular outcome as a result of churn. So definitely some positive momentum in that area. And I'm glad that the audience see that this is something that will coexist for a little while on our journey into the future.

What are -- I mean, in addition to providing more value to customers with your ecosystem, what else would you say is influencing price at the moment? Is there an issue with power constraints that's helping you? Is it location, availability? Or are there certain factors that we should be mindful of that are contributing to the pricing that you're able to pass through?

I think it's probably a combination to be truthful, like, of all elements. I mean certainly, the power narrative is a huge one in the data center arena right now. There is much that we are doing to manage that equally in terms of how we're sourcing our power, how we're ensuring our energy future. When we have our capacity defined in our data centers, how we're orchestrating using AI, by the way, to make that as efficient and as effective as we can so that we bring down our PUE. And then at the outside end of that, how we do those two pieces in order to deliver on the sustainability objectives that we set ourselves as a company and also to deliver on being a really good member of the community and giving back to the communities that we live in and the communities that we serve, district heating, for example, the excess heating from our data center being one example of that. So I think there's a very large awareness of the power issues, the capacity that, that then knocks on in terms of supply chain for the data center industry. And I think all of those things continue to create that premium price scenario for customers who want to operate in our environment.

As you're optimizing the menu of what you offer to customers and maybe introducing more to that menu, is there one thing that you think is really underappreciated that Equinix can do differently over the next couple of years?

Well, I guess, I said before that the core element of what we do is everything, right? It is the differentiating element. And although I come from a software background, I have a deep appreciation of our data centers, our data center innovations, the way we continue to innovate on the builds, on the structure, on that core. The huge power of interconnection, today, we have 472,000 interconnections, which is speaking, I think, to the value that Equinix brings not just in terms of the physical data centers, but the way that we allow businesses to connect to each other in order to facilitate outcomes for their customers. So some of those core elements are crucial to future. And I think we just need to continue to articulate their value and then look at a way that we can augment, accelerate and extrapolate that value for our customers by making it easy with Equinix to consume that service.

And in terms of profitability and performance, AFFO per share and AFFO per share growth has been lighthouse metrics for Equinix for some time. Do you still expect that to be the future go-forward metric? And what has influenced the Board and the management team to embrace that metric maybe more than some of the others?

I think it's been a core tenant of Equinix's value proposition to shareholders that every year we deliver on an AFFO per share basis and that we continue to grow that aspect of our business. And it actually has not been a debate that I have heard in -- with much rigor at our Board room because for us, it remains the central metric of the business, AFFO per share.

Adaire, thank you so much for joining us.

Thank you. Thank you, everybody, for your time.