Iron Mountain Incorporated (IRM) Earnings Call Transcript & Summary

Okay. Good morning, and welcome. I'm very pleased to be joined by Barry Hytinen, CFO of Iron Mountain. Barry, thank you for being here with us.

Thanks for having us, George.

Of course. So Iron Mountain has transformed its business over the years to build out several high-growth businesses, including data centers, asset life cycle management, digital solutions. Can you talk about the strategy more broadly and how these businesses are synergistic with your legacy Records Information Management services?

Yes. Thanks for that. So fundamentally, we -- as you alluded to at the end of the question, we have a client base that's 240,000 clients, and most of them have been doing business with us for years, if not decades. And our customer retention rate is very, very high, I think, like 99-plus percent. And so years ago, the team started investing in areas where we could plant seeds for future growth. And those were all the things you just mentioned and a few others, that the concept was how do we cross-sell from that large client base where we've got a consistent revenue and cash generating business that's based on trust and continuous strong service and a variety of other dimensions that clients have come to rely on us and how do we logically extend ourselves. And so whether it be in the form of digital solutions, that's the nearest in, I would say, in terms of the cross-sell. Because we have literally millions and millions of boxes of material that clients would like to have access to and be able to do a level of analysis on. But in most cases, it's, today, dark data for them. It's just sitting on a shelf. And for them to have the ability to do a level of analysis on it requires a consistent process. And that's where our digital solutions business comes in, and it frequently starts with a level of digitization and then can -- now is frequently leading to use cases around our DXP platform, which is a proprietary software-as-a-service offering that we've created to help clients both manage their inventory, do the sorts of analysis that I'm describing. We can meta tag and use AI agents to help clients do a whole variety of use cases. Some of those are in the financial services area, some of those are in health care, et cetera. But digitization for us -- and our digital solutions business is now at a run rate of over 500 -- nearly $550 million annually. It's a business that has been growing at 20% CAGR over many years, and it cross-sells very easily off of our core. In asset life cycle management, that is also a business that's growing very rapidly for us, to dimensionalize that. In 2021, I believe we did $38 million worth of revenue in asset life cycle management. This year, we're on a run rate to do something like $575 million or more. And we've been growing both organically and inorganically in that business. Last quarter, we grew 40% organic, 70% total growth. And it also cross-sells very well because we are going out to clients and on the enterprise side, and we're saying to them, look, what do you do with your end-of-life used IT gear where there can be confidential information on it, there can be concerns around sustainability, how things are properly recycled or reused. And we offer a service offering whereby we can come to their facilities on a routine basis and pick up that year, not unlike we do with their physical inventory that they're having a store. And then we have the processing capability that we continue to build out around the world where the chain of custody of that gear never leaves our facilities. And we properly wipe it, then we might, in some cases, recycle it or reuse it. In other cases, it would be appropriately end of lifing the gear. And that's a portion of the economy that is secularly growing. It is a very, very large TAM. It cross-sells, we think, very, very well off of our core because we already have relationships with all the clients that actually need this. And for -- currently, the market for it is very, very fragmented. So there's a tremendous number of small vendors that are doing this work for very large corporates. And we think the market is ripe for consolidation, and that's what we're attempting to do. We're already the largest player in that space. And that's the majority of the TAM there, George. But the other part of that business cross-sells quite well with our data center business, which is where we do data center hyperscale decommissioning. So think about the largest cloud hyperscalers, they've been building out their fleets of data centers for decades now -- for a couple of decades now. And generally speaking, most players in that space refresh the gear inside the data center on about an every 5-year basis. And so they need a partner, and we're one of a few that can handle that gear that comes and takes the gear out of the data center while they put in new servers, and they're making that change at that time generally for better compute and/or better energy efficiency, better power draw. And as a result, the gear that's coming out, the servers that are coming out still has value. They're not taking it to obsolescence, like is what's common on the other part of the market I was mentioning. And so we have relationships through our data center business, which I'll talk about in a moment, whereby there's a great degree of cross-selling and consistent collaboration with those clients on doing their decommissioning as well. That is more of a revenue share model. It's a highly technical portion of our business where we are wiping the servers. In some cases, depending upon the client, anything that's been written to, we would destroy after wiping. In other cases, we disassemble the servers and then sell off the various components, the CPUs, the drives, what have you, and in that way, share that revenue. So that cross-sells very well with our data center business. Our data center business this year will be approaching $800 million in revenue. We've been seeing very significant margin improvements as we've been commencing more and more of our more recent lease signings over the last few years. And the vast majority of our growth in data center over the last several years has been coming from major hyperscale clients. And like those that I was just speaking about on the decommissioning side. And we think when you look at what we're operating in data center, we have 450 megawatts of capacity that we're currently operating in our portfolio. That's up materially over the last 4 or 5 years. And we're about 98% leased in that portfolio. We're under construction of about 200 additional megawatts, of which where the majority of that is already pre-leased. So all we have to do is finish the construction, and we'll start leasing. And then we have a portfolio of about another 700 megawatts, round numbers, that we have not begun construction on that we can construct and sell over time. Obviously, we continue to look for additional land. So data center for us is another one like digital and like ALM which we think can secularly grow as a market for a long period of time at a relatively high rate and we can take share. And that's what we've generally been doing.

That's a great overview. Thank you for that, Barry. If you look at all of those growth businesses, so data centers and digital solutions and asset life cycle management collectively, how big is this growth portfolio and how much -- how fast is it growing altogether?

Yes. So it's currently this year going to be 25%, maybe even 28% of our total revenue, let's say, high 20s percent of revenue. And to put that in perspective, 6 or so years ago, it was about 8% or 9% of the company's revenue. What's important about that is the other part that is our core, the physical business has continued to grow at like a mid- to high single-digit percentage over that entire period of time. So we've gone from being -- pre-COVID, we were about a $4 billion revenue business. We're rounding home to nearly $7 billion this year. And the growth portfolio continues to grow. Collectively, those 3 businesses are growing north of 20%, and we expect that to continue. In fact, we gave some forward guidance on our last call that for our data center business, in particular, we would expect it to grow, round numbers, 25% next year on a revenue basis without even any additional lease signings, that's just based on what we already have in the backlog. And then we have another couple of hundred million round numbers of additional backlog that will turn into revenue generation in 2027 and beyond. The ALM business also has some structural benefit of -- it's kind of a booked business. You win the business, and then you continue to expand with the clients. That's one of the reasons why our organic growth has been so strong in the last several quarters. The digital business, I think, is a business that is strength to strength as we continue to educate clients on what we can do for them by helping them take dark data and make it something that they can do analysis on, it creates that much more value add.

Makes sense. You had previously put out medium-term targets, revenue CAGR and EBITDA CAGR of 10% spanning 2021 through 2026. You're now surpassing that target for both revenue and EBITDA. So when you think about the drivers of what caused that outperformance, how much of that upside is coming from the growth portfolio versus the traditional legacy Records Information Management business?

Yes. And I'll just give a little bit more context there. So we said -- at our prior Investor Day, we said, measure us from '21 to '26, 10% CAGR revenue, 10% CAGR EBITDA. And I gave kind of a scenario analysis on where -- how that would unfold. We suggested that our core business might grow mid-single and data center would grow a little north of 20, et cetera. Look, we've been running ahead of that for -- and by the way, those were all based on then FX rates. So the kind of revenue growth that we've been putting up in despite the fact that the dollar has been stronger over that period of time is just a little bit there on the fact that we've been outperforming these numbers. I'm going to give you even more. But we've been outperforming those numbers by about 300 basis points for the last several years. This year, we have recently guided the midpoint to be 12% on revenue, 14%, I believe, is the number on EBITDA. And so we've been seeing margin expansion. The outperformance has really been coming, generally speaking, George, across the portfolio. Our classic core business on the physical side has been growing more like high single -- mid- to high singles, not 5% like we had originally suggested. And I see the ability for us to continue to drive revenue management and pricing activity together with kind of consistent volume, and that's an algorithm that's working very well for us in that core business. Our digital business has obviously been strength to strength, as I was just speaking to and growing faster than what we had originally projected. Data center has been accelerating over the last few years and outperforming, particularly so also on the margin side. And our data center EBITDA margins are up around 700-plus basis points in the last few quarters and kind of at a new level now. And then our ALM business, which had some pricing issues in component pricing during the second half, if you will, of the supply chain, the supply chain crisis kind of unwinding, we're now growing very fast, much faster than what we were originally anticipating. And so we're catching up there. So it's really outperformance across the portfolio, George.

Makes sense. Let's touch on the legacy RIM business for a moment. Storage volumes have been relatively stable, in fact, growing positively fractionally on an organic basis for years now. What would you say is driving this slight positive organic growth in the legacy storage business?

Yes. It's just inherently a really sticky business. The average box that we bring in from our clients stays with us for 15 years. And it's kind of been of that order for a long time, like decades. That's not to suggest there isn't a level of dispersion across that. It averages 15. So we have some, like the boxes that come in, it's a smaller pool of our inventory that's regulatory oriented. They skew to being destroyed much faster, like 3 or 4 years. And then we have some stuff that comes into us that never gets destroyed. But the average life cycle of a box is 15 years. The way we've been growing is a few things. One, we've continued to consolidate share with our existing client base. Nearly all of our clients still store some portion of their physical inventory themselves. And we can help them as they continue to work through their commercial office real estate and other changes. In light of what's going on in the economic cycle, we can ingest that volume for them and bring that in. So that helps. Number two is we're generally growing in markets where we have long-term secular opportunity to continue to grow. Like take a market like India. It's one of the largest ones. It's probably the largest of the form that I'm about to describe, which is that market is only relatively recently beginning to outsource physical storage. And so we and several other players that have a level of scale there are growing faster in terms of volume in India, and that's probably going to continue to occur for years to come. There are several other markets that have similar dynamics, they're just smaller in nature than India. And then thirdly is we continue to actively go after new business. We still win new accounts. Despite the fact that we do business with 95% of the Fortune 1000, there's still more accounts out there to win. The other thing is all of our clients have generally gone through some level of digital transformation, George. And when that happens, if they're using relatively less paper, we go up against some comps based on what did they send us 15 years ago. But many of the industries we've worked through have gone through a level of digital transformation multiple times, and we've kind of anniversaried that. So now we're growing with the broader business cycle. So what does that all spell? We expect our physical volume to continue to be, as you described, slightly up year-to-year. By slightly up, I mean something like 20, 30, 40 basis points. And that's a model that works really well for powering a tremendous amount of cash flow because not only are we getting the revenue management or pricing elements that I mentioned, with every bit of little incremental volume, it gets us a little bit more utilization. We continue to drive productivity. And that portion of our business that you're asking about requires very limited capital to grow. So it's a very strong cash-generating machine.

So let's talk a little bit more about the revenue management program that you just mentioned. That's been put in place now several years. We've seen mid- to high single-digit growth from pricing increases because of this revenue management program, and almost all of the growth that we're seeing in legacy storage revenues are coming from pricing because volumes are up 20 to 40 bps. So how sustainable would you say mid- to high single-digit growth in pricing is? And what has customer feedback been like to this revenue management program?

Yes. So a little bit of context. It's been about 9 years now that the company has been really, in earnest, begun the revenue management program. And for the first few years, there was kind of like tipping the -- putting the foot in the pool kind of thing and just testing it out and doing a lot of research. Because there for -- quite a while there, the company really wasn't even keeping up with inflation, George, on revenue management. They're getting a lot of incremental volume, and they kind of -- it was very much an operationally focused kind of portion of the business. And as the company tested revenue management, what they found is, candidly, we were way underpriced for value. And over the next few years, we proliferated our revenue management program across our business both from a standpoint of client-type verticals as well as markets. And at this point, we have the revenue management program firmly in place basically across the entire business for the last several years, and we're comping obviously quite positively, to your point, something like mid- to high single-digit kind of percentages, and we think that's quite sustainable. How do we test? Well, the interesting thing about our business, I've been involved in a couple of other companies where pricing was really important. In this business, you get to see, I think, enough data where you have a very good line of sight to elasticity because all of our clients are continuing to send us new physical boxes on a very regular basis. That might be weekly, in some cases, that might be monthly, et cetera. And so we have the ability to see, based on when we take a pricing action, what was the incoming volume from the client 3 months before, 6 months before, a year before, 2 years before. And what is it 3 months later, 6 months later, 12 months later. Generally speaking, we don't see much in the way of elasticity. And it's important, sometimes investors ask me, well, aren't they required to send you volume or something? Actually, no. None of our contracts require any client to send us volume. So if they don't like our service, if we've -- occasionally, we make a mistake in customer satisfaction or if they don't like the terms that we're doing business, if they don't like the pricing action, they can just stop. And so it's a relatively straightforward answer is we've got to continue to deliver enough value to justify the pricing. And what I think what you find is the offering that we have, especially from medium to larger-sized clients, George, is so strong that we're delivering plenty of value for them. We continue to offer them new services that make our business that much stickier. And why do I point out medium- to larger-sized companies in particular? It's because we are really one of the only providers -- we're really the only provider that can carry a client across the globe. We do business in over 60 countries around the world. There's no other player in this space that does anything more than like 1 or 2 markets. And in many cases, if you think about it as a large corporate, this is all about consistency, chain of custody, making sure that the box is there when you need it in the future. And we've got a track record of being able to do that. In that way, they standardize with us. And they know we have a consistent way of doing things in every one of our markets. We operate 1,500 -- nearly 1,500 warehouses around the world. And they're going to get that consistent view. Also with us, they can get the digitization on demand. They can get the digital solutions offerings I mentioned, they can get the asset life cycle management. So I think we've built a fairly sticky business. Naturally, we do see some level of elasticity that generally skews to the smallest clients with us. And I think that's where we generally -- the offering is -- there's less value added, if you're dealing in just 1 location, for example, as opposed to a multisite client. And so we're going to continue to test and learn and see where we are. But in terms of looking at our revenue management program, I think you should be continuing to expect it to be delivering at the levels it's been delivering at.

Makes sense. Let's dive more into the data center business. So in 2Q, the data center business grew 26% organically, and you're guiding to nearly 30% organic growth in the second half of the year. You mentioned 25% plus growth in 2026. What gives you that visibility? You mentioned you don't need any more lease signings. It's all based on prior signings, so you can get there with high visibility. So can you talk a little bit more about that? What's already baked? And how long can your backlog run you through if you don't have any more signings?

Yes. So it's -- data center is a phenomenal business, and it is one that is very much -- for us a -- we book a deal generally, and then we're going to be in the process of constructing it and getting the power in. And then once we complete the construction, we can commence. And so the visibility -- for years now, I've been saying, hey, look, our visibility on revenue generation and profit generation on data center is very, very high because we have signed so many leases over the last 3 years. The last 3 years, we've averaged about 125 megawatts a year of additional new leases that we would take between 12 to 24, in some cases, 36 months to deliver for the client. So when we look at what we're generating revenue on incrementally this year in the back half, the vast majority of that is stuff that we booked a year, 2, 3 years ago, and we've been in the process of constructing. And that's kind of the formula, George. And so as we move into next year, the reason I could say we see a couple of hundred million of additional revenue generation, 25% growth, to use your number, for next year is because we know what we're in process constructing. We know when it's going to turn on. And that level of growth requires no additional leasing. Now there are other elements that naturally go into the P&L model like things like we have a colo business that generally renews about annually, and we -- every quarter, we renew leases. And the churn there has been very, very low. I think we're running at like sub 1% for the first half, total churn. And pricing, mark-to-market pricing on our colo book, for example, has been running in the teens to 20% for quite a while now. So we're getting incremental revenue growth as well as profit growth from that pool. And yes, it's a fundamentally a very strong business. And as it relates to you asked what about beyond next year. So first of all, we have leasing guidance out there for this year. So it's -- the scenario I was describing a moment ago where we can grow $200 million next year without any additional leasing, that's not the plan, right? We still expect to lease incremental. And while I'm not going to roll forward our comments from the last call, we'll give you updates on the next two calls as it relates to how we did. That has relatively limited impact on revenue generation next year, but it really turns into some revenue next year as well as into '27 and '28 and beyond. And so just in terms of what we've already booked, we can generate another $200 million of revenue in '27 and beyond, '27, '28, et cetera. And so high, high visibility business, one that can grow for a considerable amount of time. And structurally, as we have more and more capacity for power generation coming online over the next 6, 12, 18 months, it just gives us that much more leasing capability, which will power incremental revenue growth for the -- all the successive years.

Right. So commencements have been very strong. But signings and leases have been a little bit softer, and you reduced the guide for data center signings this year. Can you talk about what's happening? What's causing that pullback? And it sounds like you're expecting a little bit of a turnaround in the second half of the year as demand for data center switches more towards inference compared to model training?

Yes. So I think -- and that -- therein lies, I think, a key portion of this is we have not historically made our market in the AI training, a really large deployment, I think, like 0.5 gigawatt type of deployment. And many of the hyperscalers over the last -- ballparking a year -- have been very focused on getting that kind of capacity, doing those sorts of deployments. And while there's still plenty of demand for cloud and inference deployments, particularly in Tier 1 markets of the sort that we're exposed to, to get the hyperscalers', if you will, attention has been on a leasing window that might be in the 12- to 18-month horizon. And if you look back at what we've leased over the last few years, it's kind of like a high-class problem. We've leased everything that we had that we could -- much of what we had that could be lit up in that kind of leasing window. And so as we've worked through this year, we have -- we are getting that much closer to being able to go into a lease with a hyperscaler that would initiated in 12 to 18 months. And so as we said on the last call, we have seen a considerable uptick in our activity in our pipeline. We have seen things move through our pipeline into stages like technical due diligence, which is generally one of the last stages prior to contract signature. And while I'll keep my comments to what we said on our last call, I'll just note that our confidence on the leasing guidance we gave was high. And in terms of why we reduced, it's really just because, frankly, it's a little bit of a lumpy business and just looking at what we could convert in terms of high confidence within the back half, it was of that order. But if you look into next year, I think we're going to be having another very good year for leasing as we look at what we have coming available because we've got incremental capacity to sell in places like Northern Virginia, in Richmond, in Madrid, in Chicago and in Amsterdam, just to name a few. And all of those markets are places where we know major hyperscalers who are already our clients have need for cloud and inference deployments.

Right. Makes sense. Let's talk a little bit more about the ALM asset life cycle management business. So in the second quarter, the revenue growth was very significant, up 42% year-over-year organically. Can you talk about how much of that's coming from volumes versus prices? And what you're seeing with semiconductor component prices, do you expect that to be a tailwind and how you expect the volume price/mix to evolve going forward?

Yes. So it was 42% organic and 70% plus in total. And about round numbers, 3/4 of that was volume-driven, George, maybe even a little bit more than that. Pricing has been up some on -- in total. However, more so in memory than in other elements of the components. I would say my guidance has been for pricing to be pretty consistent going forward. And if anything, I would say that is -- what we've been seeing recently is kind of consistent with, if not a little bit better than that. Of course, as always, a function of what's your mix in a given quarter, how much memory are we doing, which is traditionally around half of what we do, but it can fluctuate anywhere from 60% to 40%, something of that order. So depending upon mix adjusted, the pricing has been something like mid- to high single digits up, something of that order. And I think that's not a bad place to be planning it going forward. We think that in light of some of what's going on in the new gear space where certain models have been -- original manufacturers have said they're not going to continue to produce, that creates incremental demand for the secondary because clients that have a need for that gear, they want to make sure that they have access to it, and the place that they can most readily get it is in the secondary market, which is generally what we're exclusively selling. So I think secularly, there's a lot of volume in ALM continuing to build. We are winning share, we are winning clients, and we're a beneficiary of pricing. Yes.

Makes sense. On the digital solutions side, you're waiting to hear back from the U.S. Treasury department on a 5-year contract. Can you give us an update on where that sits?

So really not too much more to say on that one versus what we said on our last call. We -- I'll just reiterate that we did submit our bid for the 5-year award that you mentioned. The government has not yet declared what they're going to do with that. They haven't announced an award at this stage. We are operating on kind of a month-to-month agreement on that, and we continue to do that, that kind of -- those at the -- around the 24th of each month. And so we're doing the work. We feel very confident, and our team continues to execute extremely well against the work, albeit that is a fairly seasonal book of business. So the business that we're doing today is relatively quite small compared to what would be in like our fiscal first and second quarter, just in light of the seasonality of that business. And we're hopeful to continue to do the business. Our team has obviously spent a lot of time developing language models and training our models as it relates to being able to ingest and digitize that content. And we think we're well positioned, but we'll await the government's decision on that.

Makes sense.

The other thing I'll just add, though, to that, George, is while that's a very large opportunity, there are many other smaller opportunities that we continue to go into the government and pitch and seek the opportunity to help save the government money through digitization and digital solutions. And so that's an element that I think is largely a tailwind to our business over the next few years as things like DOGE and other government efficiency efforts provide opportunities.

Makes sense. And then lastly, you're spending around $2 billion in CapEx this year, and the vast majority of that CapEx is being used to support your data center growth initiatives. How do you expect your CapEx trends to evolve over time? Is it going to continue to scale with data centers? Is it going to stabilize and plateau around current levels? How do you see that playing out?

Yes. If you look at how we got here, we have ramped our growth capital quite appreciably over the last 5 years. I remember the first year I was here, I think we were only doing a few hundred million dollars worth of data center growth capital. But -- and the reason that's ramped so much is because of all the leasing we've done. And we've been writing really very good return projects, I think like 10% plus cash-on-cash unlevered returns with very high-quality tenants on leases that are 10 to 15 years in duration. And so we have had to increase our capital deployment to fund the construction of those sites on behalf of those client contracts. And if you look at the last 3 years, as I mentioned earlier, we've done about an average of 125 megawatts a year. You go back 5 years ago, we were doing like 10 megawatts a year. And so there's been a huge ramp there. Now if you look at what we have to be able to lease and deliver over the next few years against what we're already commencing, I think something kind of of this order to slightly up is probably the way to be planning as we continue to lease. And then could it ramp higher? Well, we'd have to lease more to do that because we generally are not a speculative builder. What we're generally doing is we're getting contracts and then building the sites to the contract with the high-quality clients.

Yes. Makes sense. Well, thank you, Barry, for the great discussion. Please join me in thanking Barry.