Good morning, everyone. My name is Joe Cardoso, and I'm the analyst here at JP Morgan, covering hardware and networking. For the first session of today, we have Insight Enterprises, and with us, we have CFO Glynis. So before I get started with my line of questioning, I'm gonna pass it over to Glynis for some prepared remarks.
Very briefly. So Insight is a Fortune 500 solutions integrator. We're global. 80% of our revenue roughly is in North America, 17%-18% in EMEA, and 2% in APAC. What does it mean to be a solutions integrator? We actually combine hardware, software, and services to provide a solution to a client that solves their business needs. And what differentiates us is that we believe there are very few companies like Insight that can actually go soup to nuts, that can go from the actual product that's gonna be required in a solution, all the way through to the actual final solution. In today's world, digital makes everything more complicated. There's a certain expertise that clients need to have in-house to leverage the variety of technology that's out there today, and that is only growing.
So if you think about basic hardware devices, and you think about what we do for clients, we actually can give them a solution, soup to nuts. We spec the equipment for them, we give them their own personal catalog. Their employees go in there and order what it is that they need from a devices perspective, in this particular example. It looks like it's theirs. It doesn't say Insight anywhere. We spec it, we burn the protocols on there. It goes through their hierarchy with regard to who needs to approve it. We send it out, plug and ready, to the teammate in the office or at their home. They leave it at a TSA station.
We wipe it clean, give them another one at the destination that they're at, and at the very end, we dispose of that device sustainably for them and give them another one, and the cycle is rinse and repeat going forward. If you think about it from a data center perspective, in terms of the infrastructure, we can do something similar. We go in, we do an assessment about what it is that the client has in their data center. Believe it or not, there are many clients that have not yet moved data to the cloud or optimized the benefit and value they can get from the cloud. So we do an assessment.
We walk them through exactly what it is they need to do in terms of on-prem, upgrading their on-prem infrastructure to be better able to take advantage of the digital technologies. We help them migrate certain workloads to the cloud or clouds. So it could be that you have a backbone that's Azure or AWS, and then you have a GCP application as well in your cloud that's around data and data analytics. It's a great place where Google plays, so we think of Google as being the and cloud. Don't tell them I said that! And then we help them optimize the spend, because when you move to the cloud, it can be expensive.
So we help them also optimize that spend, and over time, yes, the initial cost of the cloud diminishes for the clients, but as data multiplies and expands, it continues to expand. So we have certain financial goals with regard to how we're gonna, over through 2027, expand our gross margins, have cloud and our core services grow at double the pace of product. We're gonna expand our ROIC to greater than 25%. We're gonna have EBITDA margins that are in the 6.5%-7% range, and through three thirty-one, trailing twelve months through 3/31/2024, we're at 6% EBITDA margins. So we're well on that journey and are making progress against the criteria that we laid out and progress against our strategy to be the leading solutions integrator.
That's it for my comments. I know you have questions, so I'll turn it back to Joe.
No, that's great, Glynis. Thank you for that overview. So maybe just starting off, big picture question, and maybe this is a bit of a crystal ball question that CFOs love to get. Maybe just talk to us about how you're thinking about the macro going forward over the next 12 months, maybe more specifically, the product segments that you participate in. Where are you feeling more optimistic, where you're feeling a little bit more hesitant?
Mm-hmm.
in terms of demand?
It's actually a trick question. Here's what I would say: We believe that hardware is gonna recover and that, in particular, devices will recover. Why do we believe that? We believe that because in COVID, companies went from having desktops primarily to having notebooks, primarily PCs on their devices, we call them, endpoints. That means that they're now 85% of the PC market is notebooks and 15% is desktops. That's a switch from 55 desktops and 45 notebooks. It's never going back. What does that mean? It's a shorter refresh cycle typically, because of the... you're moving the mobility associated with, a laptop that you don't have with the desktop. Gets a little bit more wear and tear, so there's a refresh cycle coming associated with that. There's a refresh cycle also coming associated with Win10.
Win10 goes end of life in 2025, so there's a natural refresh cycle. Every time a Windows operating system goes end of life, it doesn't always trigger a refresh cycle, but in this particular instance, there are security and other aspects that are built into the Win11 operating system that would be beneficial. And then last, and I put it last because I think it's in 2023 and for the in 2024 and for the next 12 months, likely not as large a contributor, are the GenAI PCs that are gonna be coming out, GenAI-chipped PCs that are gonna be coming out. We're testing, beta testing for a variety of the chip manufacturers, different specifications around those laptops, and it's gonna be a question of price and value, right?
What do consumers, not consumers, we don't deal with consumers. What do companies perceive the value to be? And depending on that answer, they may do a big push into GenAI chip PCs, or likely, I think this is, like, the Insight point of view, it's likely to be a 2025 and beyond scenario. There's a second part of your question?
Maybe on the infrastructure side.
On the infrastructure side, it's a little bit more muted. So we had heavy backlog in 2023, and we flushed that backlog through Q3 of 2023. We believe our clients are absorbing that backlog now, similar to what Cisco said in their, in their earnings release. Clients are absorbing that backlog now. We're starting to see some clients coming back into the market in terms of infrastructure, but not enough that I would say it's gonna recover in 2024. I think it's definitely a 2025 recovery. It is not dead. I've had this question before: Is infrastructure dead? It's not dead. It's just gonna take a little while to recover 'cause their supply chain was so long and extended. It's literally only in 2023 that most clients got the products they ordered a long time ago.
And maybe can you talk about, and maybe this is a weird way to characterize it, but the health of IT spending from a customer behavior standpoint, meaning we've seen over the past couple of quarters, OEMs, channel partners discuss, you know, conservative nature coming from customers.
Mm-hmm.
What are you seeing across your customers? Is there any change across different verticals? How has that been trending?
I would characterize in general that spending is conservative. So, you know, Q1 is always down relative to Q4, traditionally, no matter what the economic cycle is. Coincidentally, we saw devices increase sequentially in Q1 versus Q4. But clients are a little bit cautious. I think everybody maybe was waiting to see what happened with inflation as an indicator of potentially what would happen with the Fed, and I think maybe clients are a little bit more comfortable now with regard to the Fed potentially easing this year, once or twice, potentially. But I think that is having a little bit of a hangover on the market in terms of what's not the market, 'cause that's really going gangbusters, but in terms of corporates' willingness to spend, I think that's having a little bit of hangover on corporates.
No, and that makes sense. Maybe the other thing you can touch on, that we get questions all the time, is about AI and the potential cannibalization of spend that's going on as enterprise customers are trying to debate what they should be spending on-
Mm. Mm-hmm
... particularly as it relates to AI.
Yes. So I would say that today, the majority of the spend around AI is through the hyperscalers and OEMs direct to hyperscalers. We're now working with a couple of the OEMs on specific server technology that would be AI-enabled in terms of bringing that to the corporate space. Originally, chips were in short supply. The AI chips that the infrastructure partners needed were in short supply, so the hyperscalers got the first demand. I think they've maybe solved that supply chain issue, so right now, chips are a little bit more available, and so they've opened it up to more of the corporate/large enterprise infrastructure. I think that's gonna take off, but I think that many companies are still working through: How do we leverage AI?
Not just AI, 'cause it's AI's been around for a while, but how do we leverage generative AI and the capabilities that are possible with generative AI, the governance issues around generative AI? I think companies are still working through those elements, and we don't have many fast movers right now. That, that's not a negative comment. I think it's just that people are trying to figure out how it is they're gonna leverage and use that technology, and there are, like, three key use cases, but they're really trying to figure out, how do they scale it? You can always do a use case, a quick proof of concept, but how do you scale it in your business to really drive value?
So maybe you can talk about that from an Insight perspective. We've been asking all the companies that we're hosting at the conference how they're leveraging AI internally, and so can you talk to some of the internal initiatives that you guys have under the umbrella?
Sure. So internally, we've actually used it, probably the first use case was around our internal benefits infrastructure and a simple chatbot that is now hypercharged with the use of generative AI in terms of the answers that it can give back to teammates. It can't do calculations, but it can certainly go across the spectrum of all the different policies, and procedures, and documentation that's out there and summarize pretty quickly the right answers to a question that a teammate would have about a type of benefit, about our 401(k) plan, about our ESPP, as an example, about our medical benefits, and it summarizes the information across a variety of sources, so it's a complete answer that the teammate gets. Our chatbot before wasn't that all-encompassing in terms of the answers that they could generate.
We're also leveraging it in our technical area with our technicians, or software developers, or our application developers, in terms of getting to a quicker, faster, more agile first phase of a potential new app that we may be leveraging for a client. That's actually proving to be pretty beneficial overall from a team, a technical teammate productivity perspective in terms of how much more a teammate can do because they have this quick start. That enables us to kind of leverage the same base of resources across more clients and across greater, more solutions that we can bring to bear for clients. We've also internally developed a couple of I guess we'd call them platforms, where we leverage the technology that's out there with regard to making it quicker.
We have a thing called Lens for GenAI, which makes it quicker for our teammates to actually go in and look across Lens, look across your platform, and see what you have across your infrastructure, sorry, and see what you have that we actually can then create a solution for you in a faster timeframe. The thing about GenAI is that everybody expects speed and resilience, and we have to remind them of the structure and the data and the governance stuff that they need to have around it. We're also using it with our on our website in terms of our client engagement, customer engagement.
No, that's great. And maybe flipping this in terms of offerings from Insight, when you think about AI and the portfolio that Insight has, I think most investors will think about it as, there's AI exposure on the PC or the devices side, AI exposure on the infrastructure side, and then you have AI exposure on the services side, right?
Mm-hmm
... with the cloud business. So when you're thinking about AI parceled out between those three vectors, how are you thinking about the magnitude of the opportunity across each one of those for Insight? And then I have a follow-up on that.
I think that the magnitude is less around the devices and more around the infrastructure and the services that you wrap around new infrastructure that a client may be adopting or new ways that the client can work. I think AI-generated PCs, we have to figure out how best to sell them. I'm not disputing, I'm not diminishing the value of that, but I really think that it's cloud and services that are gonna be the winners in GenAI, in terms of leveraging that technology and capability, in terms of how we can help our clients. And because it's cloud-enabled, the infrastructure needs to be upgraded. There needs to be a good data state in order to get at the right data to make the right decisions. So I think that the value that we would see is gonna come for us in the cloud and services perspective.
There will be some around the edge, as you think about how a client may wanna make a decision closer to, to where the actual information is stored, where the information's generated, but that was there before generative AI. That's not necessarily new.
No, makes sense. And then, you know, you've mentioned a couple times around the elusiveness of AI PCs and maybe the value prop that's presenting to customers. I think one of the other reasons, and correct me if I'm wrong, a lot of enterprises are still trying to kick the tires, particularly, I think even Microsoft announced new PCs today, or Microsoft and Dell announced new PCs with Qualcomm chipsets in them.
Mm-hmm. Mm-hmm.
There's still some waiting even for chipsets to come out before even, you know, making the next move.
Yes, yes.
But I guess, can you just talk about where you are seeing the AI PC value prop resonate with enterprise customers, and where you're theoretically seeing maybe enterprise customers saying like: "Hey, this isn't for us today?
Well, I think if the enterprise client was not a beta test for GenAI PC, they don't really know what they're getting just yet. So we've actually been a beta site with Microsoft and Qualcomm around the product that they announced today. And I think that there's speed and security as being two really key components about what it does for you. I'm sure there's more, ultimately, but those are two key components that matter from an overall client perspective. And I think what clients will have to evaluate is, what is the price point relative to the incremental value that they think they will get out of a GenAI-chipped PC? Some people will get it just because they wanna be state-of-the-art, and they will learn as we go forward.
At Insight, our client base, that is heavy, large enterprise, are not typically early adopters. They adopt later. So our commercial clients would be the ones that would be buying these PCs, they'd be adopting early. But our large enterprise clients have a, you know, they have an IT infrastructure, they have rules and protocols that they go through, and they would make a, a longer-term decision about how they're gonna transition to these AI-chipped PCs. And then maybe they do it, you know, they have a refresh cycle that could be, you know, three years or four years. They'll transition over that point in time. One of the other potential headwinds is that there are new chips coming out. Different different chip manufacturers have announced chips coming out June, July, August. It kinda like continues on.
At what point do you make the decision to pull the trigger? I think that's another thing large enterprises would be thinking about.
No, that's great. So maybe changing gears a little bit and focusing on some of the near-term trends that we're seeing, particularly on the infrastructure side. Maybe can you just like, dive a little deeper in terms of, particularly in the infrastructure side, if we had to parcel that out between the major product categories there, server, storage, networking, what you're seeing, is there any variances between those different product segments? And then the second part would be, you guys are embedding a recovery in the infrastructure side going into the back half of this year. Like, what's driving that...
Devices.
Devices. Yeah, but your growth for the back half, right?
Q4, maybe.
Okay, so I guess, what's driving that conviction that there is a recovery going towards tracking into 2025, I suppose?
So, for infrastructure in particular, we had very strong Q1, Q2, and Q3. It declined significantly in Q4, so we have an easier comp in Q4 with regard to what it is we expect and anticipate that we're gonna be seeing. But we are starting now to see clients talking about what they wanna do from an infrastructure perspective. They're starting the conversation. They're not pulling the trigger, but they are starting the conversation about what they'd like to do from an infrastructure perspective. To the first part of your question about, are we seeing any differences across networking, server, and storage? Not really. They're all down. We're not really seeing much difference across either one of those. And as networking picks up, then server and storage, I think, will also pick up.
Some OEMs may be better than others as it relates to that. But in general, I would say as goes networking, as goes server and storage, in our population anyway.
No, it makes sense. And then, and you kind of touched on this in some of your prepared remarks, but just wanted to take a step back and maybe not necessarily focus on the near-term trends in the infrastructure side. But maybe can you discuss what the value prop is different for Insight in terms of what you guys deliver versus some of your peers as you kind of tackle the hardware segments? What are you guys able to do that's better than some of your peers? Why do customers go to you for it? I think that's one of the areas where investors who are new to channel partners are having a hard time in terms of understanding what the differentiation is across different players. So maybe you can talk about that.
Sure. I think the differentiation, I think, is greatest when you think about infrastructure, the cloud, and the digital transformation that companies need to be going through right now. So we can actually help you from the product side, product meaning hardware and software that's embedded in the products that you're using in your data center that makes it more nimble and agile. I'm assuming that maybe the client's already consolidated data centers, and instead of having 10, they may have two, one for redundancy, et cetera. But when you think about what is available now for clients in the data center, how they determine which cloud to use, how they leverage Google and the analytical capability that Google has, I think that we help them navigate through that environment, and then can manage it for them at the end of that process.
Now, there are other people that can navigate through the services side of it, so I'm not, I'm not disputing that. I think the difference with Insight is that we can navigate all the way through the cycle, from soup to nuts, in terms of from the speccing and upgrading of the existing data center. Most of our clients have existing data centers. They're not nascent cloud companies. They have an existing data center. How do they best go about upgrading that data center? Which workloads do they move to the cloud? Which additional workloads, as the cloud has developed and matured, should they also be moving to the cloud? And then how do they best manage and organize around that on a go-forward basis?
I think it's the ability to do all of that, all the way through, that makes us a, makes us a differentiator. Sometimes clients start out with us on the product side, then they bring somebody in to do the solution side, and then they recognize that there's some value in terms of just using one partner to do the whole thing, kind of like one throat to choke.
Yeah. No, it makes sense. Sorry, I actually have to turn it back to AI in the- on the infrastructure side.
Sure.
One of them that I forgot to ask you was just... And you mentioned this, that a lot of that business is being done directly by the OEMs. One of the areas where investors are curious about is: how does that proliferate to the channel, and when does it proliferate to the channel? So maybe, can you talk about that? Like, what are you guys seeing today?
Mm.
Are you seeing any AI infrastructure going through your business today? What's the catalyst that drives a more material uplift?
We are starting to see AI go through infrastructure, AI-related, come through the channel, and we're working with several of our partners in that regard. I will admit that we didn't see that before. It was really at the hyperscaler level, and it was because of the chip shortage or other reasons in terms of a decision that the OEMs, our OEM partners, made with regard to servicing the hyperscalers first. But it's starting to come out now into large enterprise. Not seeing it yet on the commercial side, on the smaller side, but in large enterprise, we're getting access to new technology, chip-enabled infrastructure products that we can leverage across the platform, and they're looking to us to help them push that out through our network.
So I think that's a change that is starting, and I don't see them pulling back from that necessarily, so long as the demand is there. And I think that most large enterprise clients have to do something. The question is: What do they do? So I think that the demand is there from a large enterprise perspective for these AI-enabled server storage networking to upgrade their entire infrastructure and to be able to handle the volume of data that's out there right now. And GenAI just increases the volume of that data.
Oh, that's great. Before I change gears, let me just pause here, see if there's any questions from anybody in the audience. If you have a question, please raise your hand. Just wait for the mic, please. Sorry.
Thank you. Good morning. Tomas Kubica from Marshall Wace. Can you please talk about the impact on your business from the license maintenance transition to cloud, and how that kind of flows through the PNL and your margin and kind of the net, net impact?
Oh, okay.
So, cloud is reported net. So what does that mean? That means that on-prem software has revenue, cost of goods sold, and gross profit. Cloud, SaaS, infrastructure, platform as a service, any one of those, is reported net. That means that gross revenue equals gross profit. That's a GAAP, U.S. GAAP determination. Ultimately, the rules came out in 2018 that drove that, that type of, accounting. So internally, we track gross revenue, but we don't talk about it externally. What happens when you look at Insight's PNL is you see that software, on-prem software, is converting over to the cloud, has been, and will continue to convert over to the cloud.
If you think back to Adobe many, many, like a decade ago by now, when Adobe used to be an on-prem application, it is now a fully the first fully cloud application once the cloud started. We have this similar relationship with Adobe. Probably, it has expanded over the years. So when you think about our Insight as a channel or the channel in general, we have the ability to service to provide the service to our partners, whether it's on-prem or in the cloud. And the cloud opens up more opportunities, potentially, to work more closely with the client on digital solutions that leverage the technology that's available, and the fact that it's ubiquitous, sorry, in terms of how we can use it in our clients' infrastructure. Did that answer your question, Thomas?
Yeah. No, that, that's helpful. I was keen to understand whether company vendors are more likely to go direct as well, let's say, Adobe, do they sell some of the products out directly, because easier or others so that-
Well, all our vendors have a portion. We call them partners, but all our partners have a portion of their business that is direct, and then they have a portion of the business that they leverage the channel for. Usually, vendors use to go direct for the larger clients. And over the past 5+ years, and also more maybe accelerated during COVID, many partners made the decision, OEM partners made the decision to, or cloud, or software partners, made the decision to focus on the higher end of the market, reduce their sales force, and leverage the channel. For many of our partners that are born in the cloud partners, think about HashiCorp, before they just sold, they leveraged Insight because we had a channel for them.
We could actually take their products and reach more people than they could reach with their sales force. Reach more clients, actually, not people, it's companies, than they could reach with their own sales force. So we, we used to get this question all the time when the cloud first started: "You're gonna be disintermediated by the cloud." We are now one of the top three partners for Azure with Microsoft. We're the fastest growing partner with Microsoft around Azure, and I think that the publishers, the cloud providers, have made a commitment to the channel. AWS was never in the channel, it's now in the channel. Google was never in the channel, it's now in the channel. So I think that it is a, it's a motion to get to end clients that is here to stay.
I don't feel we're gonna be disintermediated in that regard.
Any other questions? We have one over here.
Hi, Alvin Pak from Crestline. Could we maybe share the conversations you're having with large enterprise OEMs? I mean, a few years back, there was this whole debate of TCO between on-premise versus the cloud for traditional workloads, but how is that conversation going along? Any details that you might have, since you're so engaged-
Mm
... in the forefront of the on-premise versus the cloud debate and TCO, but for cloud workloads, given that it's so nascent and developing? Thank you.
Sure. Okay. It's actually a good question. TCO, yes. I think I mentioned that the cloud is not necessarily the most cost-effective product, solution, rather. But what it brings you is usually better security, a better infrastructure that you can leverage to take advantage of most of the digital changes that are happening out there in the IT space, in terms of how you move forward. So I think that today, we're not asked to do TCOs about on-prem versus in the cloud. When we go out and we meet with a client, there are certain high intensity, heavy compute usage workloads that we may recommend, or core workloads to a client that we recommend they keep on-prem in an upgraded data center, and other workloads that they move, that they migrate to the cloud.
But in general, the value of the cloud to most clients is around the inherent security in the infrastructure of the cloud, as well as some of the other API features, et cetera, that make it easier for them to create a platform that's plug and play, in terms of how they can actually get to a solution faster. So they can actually drive a business outcome faster, leveraging the cloud and digital technology, than they can drive on-prem, and that's changed the conversation from TCO to what is the inherent value in the cloud.
They do have a little bit of sticker shock when they go to the cloud, and that has developed a whole new service offering around how you help them optimize, know where their spend is, some changes that they can make in terms of actually refining and minimizing that spend. Over time, the spend continues to grow just because data is growing. Thank you.
Any other questions in the room? So maybe just piggybacking off of some of the past questions, maybe can you talk about what inning we are relative to the cloud's leverage of the channel, right?
Mm.
I think a lot of us kind of understand where the Dells and HPs of the world are, very mature channel users. Where are we in relative to the cloud customers or the cloud partners leveraging the channel, and is there variances between each one?
So if I think about the hyperscalers, I would say that Microsoft has been a channel player for decades, long time, and that the cloud is just a continuation of the channel play that they, that they have had all along. I think AWS, when they first launched, was not interested in the channel at all. They have since come around and are now a pretty good channel partner. GCP was probably in the same place, and they've come around, and they are also a good channel partner. So I would say Microsoft is the most mature only 'cause they started out there. All of them have a direct sales force. You know, they all have a direct sales force. So, we're not kidding ourselves and thinking that we're it for them.
But we have a relationship that we've carved out with Microsoft because we've demonstrated over the years our ability to scale and change, such that we're driving what is important to Microsoft today, which is very different from what used to be important to Microsoft, say, 10 years ago. And we've demonstrated that we can make that change, and they would say that we're probably one of their most agile channel partners with regard to making sure that we migrate to where they are focused in terms of how we interact with clients, and as a result of that, we get a lot of leads from Microsoft in that regard. AWS is going down a continuum of determining how to leverage the channel.
They've come a long way, in a very short period of time, actually, in terms of leveraging the channel, but not at the same level of maturity as Microsoft. GCP is, you know, maybe somewhere in the middle, actually, between those two.
I just wanted to touch on another question we get from investors a lot, or concern sometimes, is what you guys are delivering from the cloud, right? Sometimes the interpretation is that all of this is, like, Office 365 licenses.
Maybe can you discuss kind of what you guys deliver from the cloud, and maybe go into details around, like, what most of the offerings you guys are delivering or where you're seeing most of the value you guys are deriving in terms of d-
Mm.
What you guys deliver to enterprise customers?
Yes. So, you know, Microsoft has a whole plethora of cloud-related products, and yes, there's SaaS tools out there. You know, Teams is a SaaS tool, as an example, O 365, which used to be Office 365 SaaS tool that's out there. So those, those are out there, and we deliver those on behalf of our clients and sometimes help, on, on behalf of Microsoft, and sometimes help the client in terms of how they get set up. But when you think about Azure and consumption and usage of Azure, for us, and for Microsoft, it's not just around selling, selling Azure to a client and walking away and, like, clipping coupons on a go-forward basis. It's much more related to how are we helping the client use Azure? How we're helping the client with that Azure consumption.
How are we helping the client make it more integrated in their data center, so that as they grow and as they leverage different digital tools that are out there, they're actually gonna be increasing their usage around that CPU? So it gets back to an initial assessment that we've done with a client with regard to: What is it that you have? Where do you wanna go? Let's figure out the roadmap in terms of how you get there. We don't necessarily enact the entire roadmap for them all upfront, but we do it in stages. We do the assessment, we figure out what how to upgrade their existing on-prem infrastructure that is not Azure related. We then figure out what workloads to migrate. We migrate the workloads.
We work with them with regard to how to optimize those workloads, and then we actually determine now, lots of conversations around: How do we layer on the incremental tools that a Microsoft has, for example, with regard to GenAI, and how you can leverage that across your infrastructure to make you more efficient? So a lot of what we do is enabling clients to leverage tools that are out there for them to be more efficient, and sometimes those clients have the technical talent that can take that over and operate it going forward, and sometimes we have the opportunity to do a managed service for them going forward.
No, that's great. We're almost up on time, so maybe last question that I wanted to end on. Investor day targets from 2022, how would you grade your progress on meeting them? We're kind of hitting the middle of that forecast horizon.
So how are you thinking about your progress thus far? Particularly, you've made some great acquisitions over the past year. Just curious your thoughts, high level.
Yeah. So I would say from an EBITDA perspective or trailing 12-month EBITDA through September through March 31st is 6%. That is up 250 basis points from 2022, something, something in that range, 200+ basis points from 2022. So I think from that metric, we're doing really, really well. I would remind you that 2013 was a horrible year, okay? We declined 13% at the revenue line. However, we grew 2% on GP, and we grew 18% in terms of EPS. So I think that we've demonstrated our ability to manage through a downturn. It was not a recession, but managed through a downturn, managed through hardware really underperforming, really not being a contributor to the overall portfolio, yet still maintaining and increasing and expanding our profitability.
So I think in terms of the tenets about what is it we wanted to accomplish, we wanted to grow services and cloud at double the pace of technology. We've more than done that, but we expect hardware to recover. And we wanted to expand our growth and our EBITDA margins, and we have demonstrated that we can do that. So I'd say we're well on our way.
That's great. Thank you, everyone. Thank you, Glynis, for joining us today.