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

Superstar guests, the new President and CEO of Equinix, Adaire Fox-Martin, welcome and thank you for coming.

Thank you, super star, I appreciate that.

Before we begin, do you have to do any of the safe harbor comments?

I do. So any forward-looking statements, I would just guide you towards our SEC filings. Thank you.

Could you give us the forward-looking statements now, so that we can discuss them.

No, you're okay, I will wait till the question.

Which question should I ask? So thank you so much for coming. We're really pleased to have you here for our first time in your new role as the CEO. You're not new to Equinix, but you've been on the Board for a long time, I think, 5 years or so in -- so I think I want to start to what has Charles Meyers done wrong that you need to fix?

You are speaking about my boss, of course, you understand that.

Yes, he is the new chairman...

So we have Charles in the Executive Chair role, which I'm very grateful for actually because I think that it gives us a wonderful opportunity for continuity, a wonderful opportunity for me to have a sounding board because I think everybody can appreciate that these are lonely jobs. And also the history of the company where it's being...

He has been around a long so much...

He has a lot of history. So it's really wonderful to have that access. And I think I've clicked over to 90 days this week in the role, and I've spent the last 3 months out on a listening and learning tour. So I'm not really sure what time zone I'm actually in at the moment. I came back from Australia and Singapore last weekend into Europe. A number of facilities and meeting with customers, partners and our team. And for me, I would say that observations are that the fundamentals of the business are very, very strong. This is about handing a baton to me. And for me, having the opportunity to build on those very strong fundamentals. And in the short term, my focus is listening, learning, maintaining an outside-in perspective. I think that's extremely important. Getting ready for a Board meeting that's in 2 weeks. We've been doing a short strategy sprint for that board meeting, delivering a strong Q3 and a strong close to the year in terms of short-term objectives. And then I guess from a longer-term perspective, taking the outcome from this listening tour, the things that I've heard, how that informs the work that we're presenting to the Board in the next 2 weeks and then how that helps us. And you define a North Star for Equinix for the next 5 years in terms of how we continue to deliver growth and AFFO per share.

How much distance as a former Board member, do you now understand existed between you and the business in your new seat as CEO?

Yes, it's been -- it's actually -- the difference between governance and operations is -- it is quite wide. I think when you think about Board structure, there's so much that is the administrative process that takes up time on the Board agenda. And then when teams show up, they're going to show up well, in terms of what they present, how they present and the questions that you address around the strategy. So I've certainly learned over the last 3 months that I probably knew a small proportion of the actual operations of the business. And so honestly, in reality, the last 3 months every day has been a school day. There's something that I learned that I did not know before.

So in this,, what were you hearing while you were listening that was some of the more important things that you think you need to surface to this Board meeting in a couple of weeks.

Well, I guess I'll do it maybe through the lens of the customer because I think that's the most important lens as you continue to be in service to your customer. As you continue to deliver customer value that's ultimately what drives the growth of the company. So I think some interesting pieces of feedback and some observations perhaps on where our customers are on their journey. And specifically, I guess, as it relates to the AI narrative that's so prevalent across the market at this point in time. How long did it take me?

Five minutes and 10 seconds -- you are not the fastest by the way.

Right. Okay. I wasn't the fastest. So I guess, first off, from our customers unsolicited and this -- I should say, this is a broad spectrum of customers from across numerous industries and obviously based in a number of geographies. So I would say from our customers in the first instance, there was unsolicited commentary on the quality of the services that Equinix delivers. The appreciation of the technical services and quality of Equinix and in particular, the appreciation of the team and how they engage to ensure our customers' outcomes. So I very much appreciated that. I will say, at the same time, almost at the full stock being issued at the end of that sentence was, it's expensive. You're a premium price, and that's the reason why here. But nevertheless, that was an element of commentary. Then interesting discussions about the journey that our customers are on. And I'm not talking now about customers that are cloud native or that grew up in the cloud. But businesses that are traditional businesses and have a diverse IT, infrastructure and architecture. So for many of those customers, I don't think we should be assuming that their journey to cloud is over. Many of them are still navigating that journey. And it is an interesting dichotomy in some respects because they're being asked to do more with less and take on an AI narrative at the same time. And when you think about the journey to cloud, I really think that the consumption of our customers has changed exponentially in terms of maturity. So in the early days of cloud, it was migrate to cloud, which cloud. Today, it's very much a workload decision. So understanding the particular applications, which workload where and why, right? And that means that they well may choose an AWS instance, for example, for compute because of the price power there. But that given, say, Google's capability in analytics, they may choose Google for a data analytical workload. So the vast majority of our customers are navigating a multi-cloud environment. And that adds to the degree of complexity of the environment that they have to navigate.

Just -- that's interesting. But some would say that the more cloud adoption we see, the closer we are to the disintermediation of the Equinixes of the world. If I don't have any -- if I have the vast majority of my compute now in the cloud, is this a good thing or a bad thing for the Equinixes of the world?

That was this -- I guess this is the other interesting point was the role of on-premise and on-premise solutions in customer architecture. So I certainly got a sense that because there's a degree of depth and understanding of how you best consume that for some environments, cloud might not be the best outcome in terms of cost efficiency, effectiveness.

Because it's variable price.

Right. And also because for some customers, there's a regulatory requirement that they can't meet if they move over to a cloud landscape. So I do think that on-premise will continue to have a very significant role to play in the architecture of our customers. And you can also see the opportunity maybe for repatriation of some apps where people bring them back for that reason.

And when you say on-prem, you mean in Equinix, not on the actual prem?

It could be -- no, I mean in a data center that's either customers or ours, right. So for me, that's actually an exciting opportunity. Because when I look at the complexity of a multi-cloud environment and I look at the journey to cloud, the migration, it's an expensive journey, right? When I look at the fact that there's on-premise opportunities, and I look at the core offer of Equinix. Then I think there's an adjacency and a value proposition that Equinix can offer those customers. Because essentially, when you think about Equinix, we are at the point where the physical meets the digital, right? And at that meeting point, there are a number of bridges and those are bridges that customers need to navigate or cross. So the bridge between clouds, the bridge between on-premise and a cloud, whether that's a migration journey or a genuine bridge that you have to build for the longer term. The bridge between the core and the edge, right? The bridge between the LLM and the data. So these are all adjacencies that I think are areas where Equinix can have the right to win and the right to lead. And so I think that's an exciting opportunity for us to continue to add value to the core infrastructure of Equinix.

And that's the message to the Board in a couple of weeks?

One of them.

Other?

Look, I would just say generally that I think -- although I was on the board, and we just spoke about this operational learning that I've had -- and this I guess, there's some set of experiences that you bring. I think there's always an opportunity when there's a new leader for a fresh set of eyes to look at decisions that were made at a point in time. And to check the hypothesis of those decisions. And to check this decision was made at the point in time that the cloud era was emerging. Now we have an era of cloud/AI, right? Does the hypothesis that we base that decision on still stand? And so there's elements of work that we're doing around that, too.

So there's tons I want to ask. Some of the learnings and listening that I did during the last 2 days were super relevant to someone in your seat. So I think one of them -- let's just talk about AI because everyone wants to talk about it.

So I took you.

Now you're telling me I am stuck, okay. Okay. Now I got you pegged. So the -- I think this notion of hybrid private public cloud. So I kind of threw out this idea that this -- the cloud is going to take over the whole world. And that was a thesis when I picked up the industry 15 years ago that I heard. And it never really went away up until even maybe as recently as last -- in this past February. But in the cloud, what has happened is we've got this very robust, very lengthy evolution of the hybrid private public cloud. So now we're talking about AI. And everyone talks about how AI, it's all about training engines and large language models and these one-off kind of data centers built and bumble, I won't say the rest of it. You know what I'm saying. And what I'm hearing is that the reality is more likely to be very representative of how the cloud evolved. So AT&T would like to have the millions and millions of customer data points that they have collected over the years of customer service. And they don't want to share those in a public environment. They want to keep them proprietary in a private environment, but they need connectivity and they need robustness and security and all the things that a retail data center company can provide. Is that -- that suggests that there is more opportunity sooner in AI than maybe many people think exists for Equinix. Can you opine on that a little bit?

Yes, that's a really great question. So I'm going to answer the question in the context of our retail facilities as opposed to the xScale program.

Let's speak of that later, yes.

Yes, we can get to xScale later. So I'll answer the question in the context of the retail facilities. So we are already seeing the early stages of enterprise, so organizations, companies engaging with us to create and deliver private AI environment. And in a number of our data center facilities, we have the capabilities to do that. Because we have been able to even in production data centers implement the physical facilities to facilitate liquid cooling and direct chip cooling or direct-to-chip capabilities in a number of our data centers. Some of this speaks a little to the NVIDIA partnership that we have in place. So as many of you will know, there is like 3 elements to their stack. There is the software element, the networking element and then the silicon brains that drive this engine. And the offer from Equinix is we'll take this, we'll implement for you in our data center. We will help manage it for you so that you don't even have to have those skills in your organization to do that. And you can procure that either on an OpEx or a CapEx basis, whichever way is best for you as a company to procure it. And you're ready to go. And I do think there's an opportunity for Equinix to be the place where private AI happens. And the first companies that have taken on board that offering are companies that are in the pharma space, where they are running LLMs across some of their research and development focus areas. And obviously, that is such proprietary market centers of data that the private AI environment works beautifully there. So we've already seen some very early enterprise cases come into our data centers, some with NVIDIA, some without, where we have deployed the capability to allow customers to do that kind of training in our enterprise environment. So early days, but we're ready for it, and we've seen engagement on it already.

So when you say ready for it, I think, the question I get a lot is, I think, there's a misunderstanding that the whole world is going to go from CPU to GPU all in one go. But when you say ready for it, how ready, like can you -- could -- if 10% of your new business is GPU-centric in a given market? Are you that ready, or you're only already in some markets with 15%?

Yes. I don't know that I can put a percentage on it at this stage to be truthful, David. But I think there's a few things to readiness that I would like to make sure that I caveat. First of all, we have a core business that is an amazing business and has 264 data centers around the world, offering services to our customers today. Those data centers operate on a very high-level SLA. And it is that SLA that ensures for our customers that they can sleep at night, right, knowing that their systems are secure, their sovereignty, if that's what they require, is in place, et cetera, et cetera, right? In any existing facility, we would never put that SLA at risk, right? By introducing a set of workloads that would potentially bring us beyond the levels that we believe would be acceptable for us to meet the SLA with our customers. So we're conscious of that. We have identified 100 of our IBXs around the world, where we have installed liquid cooling. So across some of the halls, we have installed liquid cooling, so that direct-to-chip or rear door, the 2 main ones are available to our customers who want to have those higher dense workloads in action. It's been an interesting piece of learning for me to understand all of the work that our team have done to innovate to make that possible in existing facilities and to introduce those capabilities into existing facilities because I'm sure everybody appreciates water and electricity doesn't mix that well, right? And I think the team have done an incredible job. So there are 100 of our facilities where that capability would be available, and we will continue to look at that. But in the context of ensuring that we deliver the service to our customers that we are contracted to deliver.

So the evolution of the, call it, the hybrid private public AI is one potential growth opportunity. One thing that got brought up, Sampath from Verizon, the CEO of the consumer wireless business kind of talked about how he has one possible future is that these AI applications require distribution of the compute. And they've worked hard. They're talking their book, of course. So about the mobile edge compute. We talked to Steve Vondran, the CEO of American Tower, talking about how they are doing use cases like using towers as the new edge. Is there a risk to the demand potential at what most would consider Equinix being at the core to lose it to the edge for the lack of a better analogy? People used to think that the macro cell tower would lose share to the small cell towers eventually because you just needed more than you could do in 1 location. How do you respond to that?

Well, I guess a couple of responses. One, I'm very conscious that there are a whole series of uniques that Equinix has as a colo data center provider and that those uniques have been hard won and hard-earned over 25 years of history. We have the largest global footprint, the deepest footprint in the North American market, the highest density of interconnects, more cloud on-ramps than any of our competitors. So I think the team to bring it back to Charles, have done an amazing job of building this unique competitive differentiated environment for Equinix. I would also say that I am healthily paranoid about understanding where the markets can move and how we need to ensure that we are equipped to address market moves and technology shifts. And paranoia is a good thing, I think. And humility is a good thing as you navigate business outcomes and drive towards business objectives. I think it's probably too early to call an outcome on the question that you pose against that backdrop because I think there's going to be elements that will be very use case dependent. And we haven't yet understood, I think, the plethora of use cases. We understand some, but not all. And I think that the true value of this capability, and let's remember, AI is not new. A lot of companies have been doing AI for a long time. The new piece is the generative piece. And I think a lot of these use cases have yet to form. And certainly, in my conversations with customers, one of the elements that they are struggling with is actually the business case associated with the business outcome for many of these use cases. So healthily paranoid always ensuring that we maintain our value proposition to the customer. I don't know that there's a blanket answer to that question because I do believe that it will be use case dependent in some respects. And we will continue to watch, learn and evolve.

So as we kind of transition the conversation to the business, kind of at the intersection of your former day job working for the Google Cloud and your new role as CEO. And the very obvious humongous urgent demand for new data center builds and the part of the hyperscalers. And we've seen that from like the Lumens of the world who are doing $5 billion transactions to connect data centers that haven't even been built yet, where you dip your toe into that world in a very gentle way is xScale, which represents about only 1% of revenue.

1% to 2%.

1% to 2%. And so it's weird to start the business conversation by talking about 1% of the business, but it is the thing that everyone wants to talk about particularly, this [indiscernible]. So are you -- one of the things you said is to kind of bring fresh eyes to decisions that have been made. And I remember for the very first time when DC10 got built for 4 different players and you signed a 10-year deal. And it's okay, we're going to -- it's going to work out, it's 10% and then you kind of changed tack and did this xScale thing. What do you think of xScale? Is it the right path? Or if it's not, what should that look like?

So there's a few elements I guess...

Sorry, there are about 10 questions in the...

No, no, but it's a great question. And so it gives me the opportunity to have a discussion around xScale, which I think is a very important aspect of the business. First of all, just to make sure everybody is on the same page with me here, the decision was made to do xScale off-balance sheet, right, through a series of JV partners. And that has given us the opportunity with our restructure to enable us to deploy our capital in service to our retail business, right, to ensure that we're still continuing to grow our global footprint that we're still continuing to grow our presence that we're still continuing to grow our depth and our capacity, particularly in the North America landscape. And if I think about the business model and the role that we have for global customers, 64% of our customers operate in more than 3 regions, in all 3 regions with us. So that global footprint is important. The hyperscalers are a huge cohort on a huge contingency in the technical landscape of today. And I do think it's important that we consider our relationship with the hyperscalers. We just announced in Q2, and I'm sure everybody is aware, the first multi 100-megawatt facility, which we will put in Atlanta. We just secured the land and the power to do that. And we will have another announcement coming of this context in the not-too-distant future. So we are increasing our lean in to the xScale model. But I think it's a balanced and ambitious approach to that market. It allows us also to manage the relationships in a way that is unique. So it allows us to have a one-on-one relationship or, in our case, probably many to one because we have teams that would work with each of the hyperscalers that might have a different structure to how we manage relationships with customers on the retail side. So we're also able to manage the relationship. It also gives us the basis to think about what I would describe as a 360, a 360 relationship. So today, we're customers of each other and we are partners to each other -- I beg your pardon, customers and suppliers to each other. So we have got a cloud presence in our retail facilities, as you know. We have cloud presence in our xScale environment. We also are a customer of the cloud companies for many of the products that we use. And so the third part is, how do we partner? Where are the go-to-market opportunities that we could jointly align on that would bring Equinix value into the cloud world and vice versa. And I think there are perhaps a couple of those for us to begin exploring. And then there is the flywheel effect. All engaging with this market, not just from a momentum perspective, but also from a supply chain perspective because given the multitude of builds that are occurring and most of the investment happening, as you know, in the wholesale and hyperscale space, given the amount of builds, it's putting pressure into the supply chain for elements like the coolers and chillers and the air conditioners and the other pieces that you use to go into the build of your data center environment. So working in the xScale platform also gives us a massive relevance in the supply chain discussions in terms of the priority that you have for those items as you come into your retail business. So there's a flywheel effect there to that I think is important.

So that's it. So I guess what I'm hearing -- so what I used to hear, I think, was that the reason xScale kind of was born and existed was so that when your big, valuable important customers came to you and said, "I want to do something huge." You had an answer to say, well, "I'm not going to do that on balance sheet, but I got a way to get this done." I'm not going to say no to my biggest customer.

That's still valid.

And now what I hear you kind of saying is leaning into that, there's a little bit more of a services element, something that you work together. You mentioned this also with respect to AI. So this brings me to the second smallest business, which is the Equinix Digital Services side. Is that something you want to kind of lean into? And we've only gotten snippets of data about how important and how fast. It's been mentioned in all the Analyst Days and someday, it's going to be 5%, 6%, 10% of revenues, but it sounds like you might want to get there quicker.

So thank you for the question. I really appreciate it. But I wanted to maybe just make sure that I'm very clear about something before I answer it. Because I've come from an Oracle and SAP, a Google background, I know that there's probably a lens where people would look at me and goes, software, right. I want everyone to understand that there is, for me, a massive appreciation of the value in our core, in the physical data center footprint that Equinix runs and operates today.

[indiscernible] of the physical...

Yes. And that we will absolutely continue to invest, grow and evolve that core asset of Equinix, right? That being said, as you said already, the company has had a digital services strategy and building out some product areas to support our customers. And I think it's fair to say that it's probably quite nascent in terms of an impact to revenue in the short term. But there are a few things that I think are important. First off, as part of this last 3-month period of time, whilst being out and about in order to prepare for this Board meeting in 2 weeks, we've been doing a 12-week sprint on our strategy. As I said, creating the North Star for the next 5 years and then looking at what that looks like going forward. And as part of that, we used some externals to help us as well as some internal capability. And as part of that, one of the things that we've mapped is the 10 services that customers require in the infrastructure networking data center space. And I'll give you an example like, so a service at high level might be something that has storage. But underneath storage, there would be like 3 or 4 nuances up to -- not the products, but the type of storage that a customer may require. And we've done some very detailed mapping because we use this 12-week process to survey a broader customer base than the customer base I've been able to get to. We use this mapping to look at where is there the absolute adjacency for Equinix here? So which pieces would a customer be absolutely comfortable acquiring from us and which pieces maybe not so, of these 10 pieces that they buy. And then when we look at those 10 pieces, which pieces do we partner on because there's already incredible partnerships, and we just make it easy at Equinix to acquire that through our platform. And which pieces could we differentiate on because maybe it's something that we built, right? Now there is already work that is underway. Some of you would be aware of Fabric, Fabric Router, our multi-cloud networking capabilities. So there's already some work underway on some of those areas that we identified as adjacencies. So this wouldn't be a start from scratch, right? But I think there is a piece of work that we need to do just to build out the capabilities of those a little bit more. But already, we're seeing great traction with Fabric. Not enough yet to say, well, here we're at this percent of revenue. But enough to say that we're in the thousands of customers who are using Fabric to do virtual interconnections as opposed to the physical ones that we might do in the data center. We're facilitating ease of managing the network out to the edge within Network Edge component. So there's already some really great take-up of those products, but it is quite nascent to the overall revenue picture. But I think that it adds to the value proposition of Equinix, the stickiness of the solution and making it easy for our customers to consume. And then it speaks to giving us a structure to build out the ecosystem and the partner network that we will require to continue to grow. So I hope that answers your questions.

No, that's super helpful. These are kind of the nuggets that we would like to hear from the new person in charge. So let's talk about the real business, the whole business. June of last year, big event, Analyst Day, expectations, 8% to 10% revenue growth were set. And then the beginning of 2024, the expectation was 7% to 8%. And some of that had to do with post-pandemic grooming, slightly elevated churn, which kind of persisted through the first half of 2024. And in order to get to the guidance target, which is below the original target, we have to have some improvement in the second half. And obviously, there's a couple of different components to that. There's gross adds, there's churn, there's price, volume. What gives you the confidence that second half improves relative to first half as we sit here almost at the end of the third quarter.

Well, I think as we said on our Q2 earnings call, we had record gross bookings, right? We certainly had the cabs billing debate on a topic on a number of conversations.

It's been a pet topic of mine...

I guess I know. On a number of calls and a number of follow-up calls. And I guess, just a couple of points on that. So record gross bookings, highest interconnects sold since the end of the pandemic, a backlog, which is an important measure, a backlog of cabinets that are booked but not yet installed, right? So these are all very positive indicators of momentum going forward. In addition to one of our challenges around the cabinets billing metric was churn, which I'll speak to in a moment, but also capacity, right? We saw that the density changed. We went from a density that was in the 4s to one that's almost at the 6, right? And -- so that changed, which is a good thing because same space, more yield. But we had got in the first part of the year, key markets that were capacity constrained. And when I look at each of those markets, this capacity coming on in the second half for each of those.

What are the 2 or 3 that are most important?

Yes, Singapore, Tokyo, New York, so you can see a number of them coming on. We mentioned also on Q2 earnings that we had a very deep pipeline in the second half. Obviously, our job is to execute against that, and we have the team focused on doing just that, so that we end the year strong. And then let me come to the topic of churn. It is an important topic. And there are a few thoughts from me on this topic. First off, I do think that for a lot of companies, there was some pent-up optimization that needed to occur post COVID. During COVID, we all went virtual. Everything went virtual, everything went remote. And there was a degree of just being grateful to be able to continue to operate in a virtual remote way. And there wasn't a lot of optimization that was done in environments in order to keep those environments running. So I do think just right across the industry, there's a general pent up, there needs to be some tidy up from that era. Also, I think, generally, CIOs across the world have been asked to do more with less. And so you do look at where can you take cost out of the business. I think it's important that we all understand and appreciate that not all churn is bad, right? There are opportunities where you can lean in because you're actually creating more value for your customer, right? There are opportunities where if a customer churns out, you can actually get a higher yield when you turn a new customer in. And I think that there are opportunities for us to think about this in the context of our customers' journey with us and how we, as a company, engage on that topic, right? And I think there's a huge opportunity for us to engage earlier in the process and to really hone in and then on upsell opportunities for the value that we saved for the customer. So a more valuable relationship for the customer with Equinix in the long run underpinned. But I think overall, from a numbers perspective, the guidance that we gave for churn was 2% to 2.5% and that guidance stands. And actually, in Q3, we probably did better than we expected -- but at the end of -- not Q3, at the end of Q2, we had...

You can talk about Q3, come on. That's part of the forward-looking stuff you said you were going -- I have been asking all the questions.

You've been trying. And at the end of Q2, we had a churn that we hadn't anticipated, which impacted that, which is the StackPath liquidation, which was quite material at the end of the quarter. So we were within the range and better than we expected actually without that.

Do we need to brace for something in the third quarter either fall out from that or some other event that we should be aware of?

We're executing to our plan.

Great to hear. So I guess my last question, and this is just more of a housekeeping item is and your listening and learning tour with -- inside the business, you shared a lot. And you're listening and learning business with the investors, I know you and I had a conversation after the 2Q, I think that there's a desire for there to be some sort of marriage of the traditional 4K [indiscernible] cabinet multiplied by the MRR number equals revenue when we've got cabinets that don't grow, but power density does and the revenue yield is different. I feel like I'd be...

Remiss if you did not raise that issue, yes.

Having a better vocabulary...

Yes. So look, this was -- obviously, we had a lot of discussion about this, both with various different investors across the end of the quarter about is this the right P x Q measure, given that so many things have changed. It's a big internal debate, right? I think from a company perspective, Equinix is exceptionally transparent in terms of the numbers and the number of elements of data that we share with our investors because obviously, we want people to see our transparency and to trust the information that we are providing. And we are debating about whether that remains to be valid or whether we change up that...

Still working on it. .

Yes. I mean I think one of the things that we said also at the end of Q2 is that we will provide data on density that will maybe help with that calculation in that view a little bit better. But that's definitely something that I'm working on with Keith.

Wonderful. It's great to know you're working on it. Welcome to the Equinix family. Thank you so much.

Thank you so much. Thank you. Thank you very much.

Thank you all for joining. Appreciate it.

Thank you.