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

My name is Brendan Lynch, I'm the U.S. communications infrastructure analyst here at Barclays. Very pleased to have Steve Madden, VP of Digital Transformation and Segmentation Marketing from Equinix as well as Katie Morgan, who's on the IR team. Thank you for joining us.

Thanks for having us, Brendan. And before we open up the stage, I just want to remind everyone some of the statements we may make today are forward-looking in nature, so please check at our SEC filings.

Great. Steve, maybe we'll start with you. You had digital transformation in your title, why don't you tell us what that means in the context of Equinix?

Sure. Yes, we spend a lot of time given the position where we are in the market, focusing on what's happening in the market, what's changing, what are the big compelling events. And digital transformation in terms of customers changing the way they do business and the way they use technology for revenue purposes is really what that means. So we see a lot of activity of customers changing the way they deliver and operate in a digital world electronically, which is moving their business from traditional to digital. I would say a digital company is majority revenue coming from digitized or digital sources of revenue.

Great. And when you look at your customer landscape, where would you say most of those customers are in the process of embracing the digital transformation? How much work do they have left to do?

Yes. No, so it's been a journey. I would say around 2015, there was a big push to move out of the traditional data center. Around now, more than half of the market has already actively left the sort of traditional data center. It's a mix of hybrid cloud and still public and private cloud. And there are a lot more companies that have been born during that journey in cloud and then grown. So I'd say for traditional enterprises and large-scale enterprises, that's somewhere in the middle. We've got a new market of new startups and IT services companies that have been born and the amount of -- I don't want to call them laggards, but followers, is getting smaller as they need to catch up.

Sure. And when you look at these companies, I'd imagine they're all doing a cost benefit analysis to varying degrees. What are some of the considerations that go into placing a workload either low latency colocation like Equinix or in the cloud or maybe more generic colocation that's a bit cheaper? What are the considerations that they're going through?

Yes. It really comes down to a lot of what is the benefits for where you put something in certain location. So when we have those conversations with customers, we know that different workloads have different characteristics and, therefore, lend themselves best to a different way of deployment. For us, it's extremely low latency or it's high regulation, high security because it's all dedicated private activity or high volumes, really high volumes of traffic and data. If I was going to do a rule of thumb, they typically work well within Equinix. So if it's a workload that is either latency is impactful for either revenue reasons or just user experience reasons, the high volumes are, if you're moving tons of data and traffic, doing that over the network is extremely expensive. Versus the people you're exchanging data with are in the same building or in the same campus, a cable cross-connect is a far more efficient way of doing it. And like I said, with regulatory and security, a lot of these businesses don't want that data or that business to traverse the Internet, and they don't have a common network so they use us to do that as well.

And it seems that Equinix has an expanding suite of services to address the varying needs of clients in various different positions and how they're trying to optimize their deployments, maybe we can start with kind of Equinix' bread and butter of interconnection. Your number of interconnections has doubled since 2016. The growth rate has slowed a little bit off a much higher base. I think it growth now. Maybe you could talk a little bit about what you see as the future growth rate and what's driving those interconnections now.

Well, I can tell you what's happening, and then that resembles and what they deploy. So in a lot of cases, when companies were first coming to Equinix, they were connecting to network service providers, and rearchitecting how they deploy the global backbone. And so whether they might have had 1 or 2 providers before, they typically is between 7 and 9 across the globe, different providers, a lot of interconnections go to that, about more than half in some cases. Then there's all the cloud providers. So they're not just connecting to more than 1 cloud. They're connecting to more than 1 availability is on bridge cloud as well in different regions. So a large number of interconnections, about 15% to 20% goes there. So the balance then is about another 20% left, go to others. So whether it's data providers, business partners, transaction providers, subsidiaries, et cetera, or connecting to payments platforms or what we're going to do, that's them peering or connecting to other aspects of the business they want to use. So depending on the maturity, if they're still in the early track, they'll still just connect into networks and those other connections are going to come. But as they become more mature, we start to see the amount of interconnection going up that's non-cloud, non-network is getting higher. And that growth rate is a maturity and how we come to becoming digital. So the reason why we don't attribute that to numbers of connections all the time is because bandwidth is also being aggregated into bigger pipes and bigger chunks. And you go from like 1 gig to 10 gig to 40 gig, you're still consolidating a huge amount of traffic. And likewise, we've got virtualized circuits that are getting so small, people can actually splice up different business units on the same wire. So how are they using it will change the number of connections, but the growth is continuous regards.

We talked a little bit earlier in the day about blockchain and interconnections for blockchain purposes. Maybe you could talk a little bit about your exposure there and how you see that going forward.

Yes. I mean again, the characteristics of blockchain as being a way for a technology that brings a lot of different companies together to transactionally flow as a group. So blockchain is really good as a substitute for like a centralized body or an exchange, someone who's controlling all of the flows, as what blockchain does. So blockchain enables a dozen companies to interact with each other without having one of them be the nominated guardian, et cetera. That's a model of a dozen companies that are trying to interconnect in the business together, which fits very well on top of that mix. So we see in digital business, smart contracts, blockchain payments are all common attributes you're going to see in any digital business. That's not to be -- and it always gets associated with crypto and, currently, that is the implementation of blockchain as well as where it came from. But we see blockchain as a very separate entity and workload type than crypto specifically.

And maybe just while we're on that topic, do you have any significant crypto exposure on the blockchain?

We don't really have overall. We don't really have any significant crypto expansion within the business.

No. Like I said, that's why we're more interested in blockchain. Blockchain has a much longer tail across multiple industries and more use cases than just finance. Now clearly, we have a very strong finance body, and we do have customers that do use this technology, but it's not a big exposure.

Great. Maybe we could talk a little bit about some of the other solutions that you have expanded into in recent years. Maybe one of the first ones was Equinix Fabric. Maybe you can give us a kind of an update on what the reach is of that currently and maybe some of the other solutions we can get into as well, like bare metal and how the adoption has been.

Sure. Well, Fabric was one of the first technologies we implemented to -- in all cases, we're trying to speed up what people are trying to do on the platform. So a lot of the reasons why people came into Equinix was to connect to networks and clouds. So we originally created a thing called the Cloud Exchange, which was a way to connect to multiple clouds much easier, simplifying it, making -- and all the complexity was sort of pushed under the covers, made it really simple for customers. That evolves to -- well, why not just cloud. Why not connect to anybody who's in that facility. And then we connected all the data centers together on the fabric, which meant that now if you've got a connection to the Equinix Fabric, in, say, Silicon Valley, you can connect to any other participant on the fabric anywhere else in the world with an API call with 0 terms and conditions that spin up the direct peering connection. So that's become now used for other things. So it's not just cloud, it's for B2B business transactions, companies connecting to themselves and using it as their backbone. But we started offering other services that are available, like you mentioned metal and network age and other tools, but a lot of our suppliers coming to the data center offering their own services like Pure Storage, NetApp, et cetera, which can just be made available on the fabric. So with an API call, you can spin up compute capacity, network capacity and other capacity and add it to your environment. And so it's becoming a delivery vehicle really for IT services and technology.

I think when this is really first starting to scale, there was a concern within Equinix's management team as well as with -- among investors that you could be cannibalizing some of the demand that you're seeing from the fiber providers. Maybe you could address that and other solutions that are available kind of tapping into the network.

Well, it was justified in that the last thing we want to do is compete with our customers. What we're aiming for is how do we software define something that otherwise is laboriously slow and hard to do. And so we sort of took the stance of if we show network providers that this is how it can be done, and we're in a position to do it for all of them, wouldn't that help them deliver business better as well. That's what happens. So yes, we were concerned and we had to do a lot of education. We went out ahead of it and talked to them all. And it depends on who you spoke to, when we say NSPs, some sellers inside the NSPs were concerned. Others were terribly excited, right, especially enterprise sales teams. Now networks are more than half of the use case of the fabric today. So they are our biggest customer of that technology. It's because now a customer says, "I want this, this and this." And that provider might have half of that, but they can spin up the rest using the fabric and offer the service for the next day. So it's become a game-changing amplifier rather than a detractor. And we seem to be working in the same circles like is this controversial or a problem or let's discuss it. And hopefully, we can get to a point where they see the amplification factor before we do it, not after, but it's always worked out really well.

Can you talk a little bit about the competition and the software-defined networking arena? It seems like there are some other solutions that have kind of followed since Fabric was initiated. Maybe even some of that preceded like Megaport, where do they fit in the ecosystem as well?

Yes. I mean I think about it, it's a big space. And so we still encourage other providers of SDN-type technologies because we still need our customers to be able to connect to us. And so our fabric doesn't go outside of our own data center space right now, right? So if they're in another data center, how do they get to Equinix, they're going to use a provider. And we want that to be as dynamic as possible. Likewise, the amount of bandwidth and exchange of traffic is going to try and find the quickest path that might not always be us, in which case, we want the customer to find the right choice. So we seem to be either propping up a lot of those additional technologies because they're writing on top of us, or they're amplifying or augmenting the space that we weren't covering. So I think it's healthy again for everyone to have enough access to get where we need to go. I wouldn't look at it as being something that's preventing us from continuing to grow the business universally.

Let's talk a little bit about bare metal. Maybe just give us an update on how many metros you're in now and what the adoption rate has been among customers. And maybe if we look at just in any given metro, what the initial footprint was and how that's expanded over the past couple of years.

Sure. Can you start with metros?

Yes, I'm happy to start. So today, we have Equinix Metal available in 23 markets around the world. And as we noted on our Q3 earnings call, we saw strong demand within digital infrastructure services overall. And I'll probably turn it over to Steve to talk about we sit in a unique position where coupling together metals, fabric and network edge we can go in as a trusted adviser to our customers. And enabling them and giving them the building box to architect infrastructure the way they would like to, either on a physical basis with us or on a virtual basis, leveraging that digital infrastructure services.

Yes. And we found that there was multiple benefits that come into that. One was the suppliers that are in our facility now have this way to deliver their technology through our infrastructure on demand. So again, it's kind of like Fabric. We're enabling software-defined access to their technologies, Dell, HP, Cisco. And it's also typically a lot of the same technologies that enterprises were already buying. So they're not against it, either. They think it's a great idea. And then from the provider side, IT services companies that are trying to scale up bandwidth or capacity in different regions find it much more conducive and much more normal for their business model, like the cloud consumption, bulk pricing, et cetera, and spinning architectures that way. So it's lending itself well to what our future or more advanced uses want today. Our customers who are coming along for that journey from enterprises are already shifting to that model now. And the service providers or technology or technology offerings that we have in our data center, it's a marketplace for them to sell their own infrastructure. So again, it shows an amplification of all of those different components of our platform, leveraging and selling and cross-selling into each other. So I look at it as an enabler and a channel for how do you get access to other kinds of technology.

And from Equinix's perspective, do you have a preference whether the space available within your facilities is absorbed via bare metal or via standard colocation, were you agnostic?

I'd say that we're agnostic. We're indifferent in between. We're just giving the customers the building blocks of how they want to architect between if they want to do it on a physical basis as they've always done, or leverage digital infrastructure services to do that on a virtual basis.

Well, both, right? It's what's best for the customer. In a lot of cases, it depends. And also the nature of what we're offering and more offerings come online, it changes. The nature of why the customer made the decisions in the first place changes. But we want to optimize the space for them and for us because it works at the best for both of us. In some cases where people wouldn't use or don't use it is because the technology they're deploying is far more unique and specialized in design for that specific business purpose, and that's a little different. And so where we go down that path is like for NVIDIA, AI everywhere is a very hardware AI analytic platform, not a generic compute server. But we are working out models where we can offer that service to you through the metal construct, but it's not really a generic server. So I think we're trying to find what's the right consumption model for the customer for the right technologies they need.

Great. And maybe just as a variable cost solution, do you see customers often debating between using bare metal with you or going to the cloud? Or is it usually a much wider spectrum of considerations that they're looking for?

I would say it's the opposite. Usually the decision on where it goes outweighs the technology choice than how you want to consume it. Like so -- we know it's going to need to be here and we know it's going to need to be this size. Okay. Do you want to use -- and we might want Dell. Do you want to use metal to get Dell? Do you want to use Dell directly? Do you want to use it Dell retail? How do you want to do that? So I'd like to think of it as the reasons why the workload is there is architectural and that justifies it, and then what's the best way to implement it with choice to solve for it. It's rare that we're going to have a conversation that we have the best Dell server and, therefore, you should work -- let that work. The benefit is already made by the architecture of why they're at Equinix and the latency of the access to counter-parties, et cetera, to do that. And then we're trying to be more modern in that if they're used to a cloud consumption model, they can still get that, right? And that same consumption with intents well. The choice is important.

And speaking of choice being important, let's talk a little bit about Network Edge and some of the solutions that you're making available to customers there, a little bit more on the software side, SD-WAN, et cetera. Maybe talk us through that and what those opportunities are.

Yes. So I'll tell you why we did Network Edge and I'm going to turn it back over to Katie because Network Edge was -- everyone needs to deploy network connectivity, interconnection, compute and storage. So let's start with the network. How do they get a router? In some cases, it can take 6 months to get a router deployed in, like, Brazil. And by the time they even shows up on site, it's already almost a year old. So what we said was these are all the same kinds of things that people are deploying if we had a way to just turn those up the same way cloud does with the same provider, Cisco, Juniper and everybody else. Wouldn't that be easier for people just to move faster rather than just provide the connectivity, there's something on the other end of that connectivity that's needed as well, how do we provide that. So we started Network Edge for the network and the virtualized appliances. And then if it doesn't fit in the virtual appliance model, metal picks up and takes care of the rest. So we're sort of building a spectrum out of that.

Yes. I think it's well covered.

Sure. Let's talk -- change here a little bit and talk about the channel. Maybe just take us back to the beginning of the channel initiative. What was the original scope and goal and where are we now?

Yes, happy to take that. When we launched our channel partners, it's really a great way to amplify the reach of our direct sales force. So we have about 600 quota-bearing heads that there's always so many hours made that they can be out prospecting. So the channel is a great way to amplify the direct reach. And we have a few different types of channel partners. We have our network resellers, our strategic alliance partners. In the Middle East, we do some brokers. But really, we're putting our effort behind the strategic line partners and the network resellers. And they're allowing us to go out and amplify our reach to customers, who may not directly make their way to Equinix. And it's been a great way for us to amplify the reach as well as acquire new logos. So in Q3, the channel actually contributed about 60% of our new logos in the quarter and roughly about 37% of our bookings in the quarter came from the channel partners.

The cool thing about that is they're not channel partners in that they're only selling Equinix. They're selling their own services on the platform, which happens to pull through us as well. So as our number of providers goes up and they're making money and amplifying their own business by bringing in more customers to the platform who want those services and the value to them is going up. We're watching the value of the consumers go up. We're watching the value of the producers go up. And we're amplifying how we get this to go with a channel to help all that run at software speed. That's why it's kind of interesting because now it's kind of like 60% of new logos. Our partners are growing the platform faster than we were, and that's bringing in more partners. So I think we just need to sit back and just amplify all of this to see where it goes.

Yes, that's interesting. It's kind of feeding on itself. How involved do you need to be? Or just do you want to stand back? Or do you want to be more active in cultivating this marketplace?

Yes. I mean it's mixed. In a true platform business model, you externalize the engineering and the growth, right? And where we want to spend the time is what can we add value in that's actually amplifying for both of them. So we're not offering anything that's going to compete with them right directly. That's not the point. It's got to make things faster. So we could offer more like directory and marketplace-type services so they can find each other. We can do more matchmaker kind of algorithms to help them discover opportunities that might not have been seen. And we can leverage those same kinds of models that other marketplaces do just to facilitate and accelerate what this business should look like. What I also do in my team and as a company is we don't go out and look at these different participants on the platform that's just verticals that consumers produces. And there's a value of the platform to them, which we've now realized is much greater than even Equinix is as a company. So if some of those partners are doing deals that are deploying infrastructure to our consumers, and we might get $1 out of that, they're getting between $10 and $200 like for that. So if we said our opportunity is this big, it's 10x bigger for the producers. So the value of the platform is actually quite large. And if we can show them that, advertise that, facilitate that, meet and greet and match-make around that and make it software-defined in the middle, that's where we fit. And so that's kind of where I want to see us go. And right now, all of the strategies are lined up towards that.

And would you try to monetize that just through your SpaceX Power interconnect and the other opportunities that we've discussed already, Bare Metal, Network Edge, et cetera? Or would you try to monetize that in an incremental fashion?

Yes. I mean we understand the economics that market -- other marketplaces charge a percentage on transactions and things. Quite frankly, we don't see the need to do that. If anything, our goal is to try and get as many consumers and providers for choice and as many market size of access to consumers on the platform. Because that in itself is where the value comes from. And if you implement an arbitrage or a cost on top of that, that's detracting from business growth. You're starting to compete with yourself. The fact that they're still buying our stuff and they're still deploying our data since using our Fabric, it was enough. That's plenty for us.

Sure. In terms of verticals. You've had a lot of success with financials. You've had great success with networking. Maybe you could talk about some of the sectors that you're maybe under-penetrated in or have a good opportunity for you going forward that you're not fully having to do just yet.

Yes. I would look at it as being there are certainly industries that had to be very digital very quickly in order to compete. And other industries are only sort of switching towards becoming more digital now. But as of like -- we do a report on this every year, there isn't any industry that's not actively pursuing digital transformation and at a growth rate of -- and some of the slower ones are now around 60%, 70% compound annual growth as a shift to digital because they were behind, right? So because they're coming from a smaller base, the growth rate is much higher. The health care, manufacturing, transportation, logistics, travel, all of those now are significantly higher. Even public sector, government and education is actually pretty high. So it would be the follower industries that wasn't their primary revenue source with digital. So now they feel like it's something they have to add to participate in the economy, are the ones trailing. And obviously, the Streaming Media's payments and financial services. Like my background as Wall Street, we digitized trading a long time ago in the most marketplaces. The models do the trading, not the traders, it's all software. We did that 10 years ago. I think you'll find that even a healthcare diagnosis is going to be done through a lot of AI, a lot of activity and the doctors are going to manage the models, "So okay, this is what you have wrong with you." So it's all that technology is coming to the forefront across all industries, health records exchange, telemedicine, MRI scans and real-time responsive AI on that. All that stuff is coming to bear. Down to something as simple as Alexa in the hotel -- in the hospital room. So patients -- I'll just back up, the patient, 9 out of 10 requests for a nurse could be handled by Alexa. And so now we have different companies like Sentara Health that put Alexa in all the hospital rooms, and for COVID reasons, the nurses didn't need to go in the room anywhere as much. It's the technology implications across the board are huge. So I haven't seen an industry now that says, "No, we don't need to do digital. We're not focused on that. That's not a problem for us." It's everybody. And it's also global. There isn't a region that's more or less than the others now. It's the U.S. EMEA was about a year behind and APAC was a bit slower still, not anymore. They're all equally growing at the same pace across all industries.

Great. Maybe we could talk a little bit about the Edge. I don't think we can discuss digital transformation without an Edge conversation as well. I've heard Equinix management in the past said that they saw themselves as an aggregator at the core in the properties that you own now. Maybe you can talk about a little bit how your strategy has evolved over the past few years.

Sure. Yes, absolutely. And we do think that, ultimately, almost everybody as a customer of ours is thinking through and deploying an edge-to-cloud architecture. And when we say edge, we think of that as the local metro city. So San Francisco, for example, is the edge. And there's an aggregation point there because all of the mobile devices and phones and WiFi and telco and cellular and everything else, all those different broadband and access connections aggregate typically with us, because that's where the providers are. And then they hit mainstream connectivity right back to cloud. And so there's one aggregation point in the Metro City, which is the reason why people want to be there is in proximity to people, for factories or cars or logistics and transmission. But the regional core hubs, which we did start off at, were cloud or network density is really high. Those regional clouds, of course, your proximity was to cloud or your proximity is network density. And so some of the application workloads want to be proximate to cloud and some proximate to like football stadiums or something else. So that's how we differentiate. So one of them is a digital presence. The only reason why you're going there is access to cloud or networks. And the other one is because of physical presence, you still want to interact with the physical world. Cloud has obviously been the big thing for 2015 to '18. Charles is right. But since then, once they already had cloud established, they're like, "Well, I have to improve my user access, my security perimeter and smart industry. I want to do things like telemedicine, I've got to miss some computer after the edge." So the growth rate at the edge is now much higher but the big base that starts from is the core. And I kind of use it, not a good analogy, but if you had like one of those traditional swimming pools with the deep end and the shallow end, yes, if you're filling up the pool, the deep end fills up first. But then stuff spread out, right? And now we're sort of starting to see it across the board. It's probably about 70% core, 30% edge but the edge is so distributed and where distributed in a lot of these locations. It's an edge-to-cloud architecture.

Great. I only have a few minutes left. I want to hit on one of the key topics that's been debated around Equinix over the past few months, and that's the power cost inflation. Maybe, Katie, you could give us an update on where you guys stand at this point?

Yes. Thanks, Brendan. No update from our Q3 call, but I'd just reiterate, we're taking a step back. We operate in both regulated and deregulated power markets. On the regulated side of the equation, those are really taking pay environments. The rates are set by the local municipalities or utilities than us, and everyone in the market are exposed to the regulated rates. On the deregulated side of the equation, we do have the ability to go out and hedge our exposures. And the way we think about our hedging strategy is very similar to what we do on the currency side of the house. So we've progressively feather into our hedges kind of on a rolling 8-quarter basis. And our goal is not to pick exactly where we think the spot market is going to be 12 months or 2 years from today. We're just trying to mitigate the volatility for our customers. That way, when they open up their Equinix Village fund, they know exactly how much it's going to be. And so as we look to 2023, we're over 90% hedged globally, and we're 97% hedged in EMEA. And so having those hedges in place allowed us throughout -- and communicate to customers in advance the expected power price increases. And so we really developed a comprehensive communication plan to our customers, where there's 3 customer communications. The first one was to give them a general heads-up on utility costs. The second one was to range vendor for customers to give them a sensitivity of what the expected power price increase would look like. And the third one was to give them their final notification. And those will all take place, effect on January 1 of 2023. And our expectation is to pass on 100% of the higher power cost to customers as we look to 2023. And I would add, though, from a customer standpoint, we view that, since we've layered into these hedges on a feathered basis, that we think the rate we can offer customers is attractive to both what they might be seeing in the broader spot market as well as what they might be seeing from a competitive standpoint as well.

I'd only add that we're making it easier for them as best we can in the industry, and we're also working with a company that's leading in sustainability. So the only way out of this is to change the way we consume energy, and we're 95% renewable energy now. So the goal is to work with our customers and our partners, including clouds. We're all in this together. We're setting standards for the industry on how do we move to a better model to solve for energy permanently.

Great. Well, we're just about out of time. I want to thank you both. Thank you, Steve. Thank you, Katie. We appreciate it.

Thanks for having us.

Thanks for having us.