electronics manufacturing services companies, and we're honored to have Flex. We have Revathi Advaithi, who is the CEO of the company. Revathi has been with Flex since 2019, and I think in the last five years, she has completely transformed the company, and we're gonna talk about that. But Revathi, thank you so much for coming. Really appreciate.
Thanks. I appreciate it.
So, Revathi, I think one of the things I want to start off saying is that I think you created a lot of value for shareholders. The two things that come to mind is, one is on margins. I think you've really transformed the industrial segment, other segments, margins have really transformed. And then on Nextracker, the value that you've created by, you know, divesting that business, I think that's great. Now, let's look to the future. When you look five years out, how do you see Flex then versus what Flex is today? What is your vision for the company?
Okay, great. Let me start off by saying that it's always nice to reflect on the last five years when performance was good, but we're gonna look forward. So the way, let me tell you about what we've been working on the last few years, and we're talking a little bit more about it, is you know, as we were rebuilding Flex and thinking about the Flex of the future, one of the things we focused on was areas that we thought we could have a differentiation on outside of Core EMS. Core EMS is still important to us, and we think there's a lot of opportunity to grow and improve margins there. But we also started building adjacencies to that that were very close adjacencies to Core EMS. And what do I mean by that?
In areas like products for power, we started building a portfolio of product capability, both in automotive and in data centers, because they're very similar in terms of product capability. So we started building that, and then we started building a portfolio of services around kinda end-to-end manufacturing services, everything from staging and bringing a product to the build-out, to the aftermarket and the life cycle that comes after it. So in the end, now we have the Core EMS business, we have a products portfolio that's really focused around power, and then we have the services, value-added services portfolio that kinda wraps everything around it. So if you think about the Flex of the future, I would think about it in those terms. There is Core EMS, there is products where we can differentiate, and then there's value-added services.
Each one of them bring an incremental opportunity in terms of growth, but more importantly, in terms of, margin improvement for the company as a whole. So I'd say that's kinda how you have to think about Flex moving forward.
So I wanna touch on each of the segments, but before we get into the details, you know, one question I've gotten from clients is, you know, how are you seeing the macro? Is it fluid? Or d o you have visibility? And the question I've gotten is, why did you guide for fiscal 2025? W hat is giving you the confidence to guide for the full year?
Yeah, because I think ever since I've been in this business, Ruplu, one of the things we started doing was to provide some view of a full year guide, right? And 'Cause I feel like, at least my view, and maybe I'm from the good old-fashioned industrial kinda world, but, my view is, you have to give some kind of a full year outlook to, to investors. And so we have given a guide based on our view of what we see with our customers, and I would say we're fairly good at it, right? Outside of changing our revenue once in the last year, we've been fairly consistent in terms of kinda what we see for our end markets. And typically, we're pretty conservative also in terms of what we take into account.
We look at what is in inventory, what is the channel sitting at, all of that to give a full year guide. So we've used a fairly, I would say, robust process to give a fiscal year 25 guide, and, I feel comfortable with that. Hopefully, macros are, you know, we've taken that into account and, and it'll be consistent, so.
One thing I think that caught some investors off guard with the fiscal 1Q guide was the commentary on EBIT margins. Can you give us your thoughts on how you're seeing the different segment margins in 1Q? What is impacting those margins in the first quarter, and how should we think about progression of margins going forward?
I'd say if investors were paying attention the last 4-5 years, you know, we've had cycles where we have said, "Hey, we're gonna spend more money doing X, so our margins are gonna be muted, and then we'll make up for it during the year." And if you look at the Q1 margins, first year-over-year, it is better. So we're always kinda muted in Q1, and then we ramp up as volume ramps up. So some of the things that make up Q1 margin is, one is it's a slower volume quarter, so there's always an absorption and holding cost kind of thing. The second is, you know, we definitely have some program spend that we tend to do earlier in the year, if we can, to get ready for programs that are coming up, so we're doing that.
There's a little bit of a mix impact. But our history shows last year and the last four years, that we've had cycles like this, where in a quarter we'd say, "Hey, we'll take on more cost because we're planning for X," and then overall, through the year, we improve margin. So it's very similar to that. I would say, Q4 had kinda blowout margins, so, you know, people expected that, hey, that continues into Q1. But if you're looking at the cyclicality and the history of our business, we've never performed that way, right? There's always been a, you know, kind of an up and down in terms of overall margin profile. For the full year, we have guided margin improvement again, so and I feel very comfortable with that.
All right, maybe I'll ask you one more high-level question before we dig into the segment details. I mean, just from an overall macro environment, do you think visibility is now better versus 90 days ago? How long a forecast do your customers give you? And is this year a back-end, is fiscal 2025 a back-end loaded year, or how should we think about seasonality in the year?
I would say visibility is better, and the biggest thing I think that makes visibility better is inventory in the channel. We have seen a lot of our customers talk about inventory in the channel and the inventory digestion period. So I think the visibility of that is getting better. So I think that's one of the reasons why overall, projections are getting more clearer. I would say that I think the seasonality will go back to the traditional seasonality for the business, and COVID and supply chain crisis and all of that threw seasonality off totally, right? So we've always had a slower Q1, a pretty bullish Q4, because it's kind of the December time frame for most customers. We'll go back to what we think is the historical seasonality.
Okay. So Revathi, AI is a big theme that investors are looking at. Can you talk about what is AI? How is Flex benefiting from that? How do you participate in the AI, you know, market? I mean, what is it that you're doing on the power side? What is it that you're doing on the module side? If you can help investors understand, what is it that Flex is providing customers?
Yeah. So, you know, we talked about this in the in our recent Investor Day, and so Flex has been building a portfolio around data centers for the last, you know, so many years, right? More so in the last 4 years, particularly around power. So if you think about data centers as a whole, we think of our portfolio in two buckets. One is the compute side, and the other is the power side. When you think about the compute side, what we do is basically do all the integration for large data center customers. So we could do everything from just building the server storage boxes, we could do the wrapping the whole racks and enclosures out of it, and fully deploying it in the hyperscale or colo as needed, providing the cooling solutions around it.
The thing that differentiates Flex in that is, we will do it end-to-end, right? So we will fabricate your sheet metal, weld it, put it together, build your server storage boxes, help you redesign them if you want, build a cooling reference for you that you can build into your overall rack integration, and do it end-to-end, so you can scale it as quickly as possible. And very few people can do that end-to-end. Usually, it'll be a very fragmented supply base, like, you'll have to maybe get your racks from, you know, the people who, who bend metal. You maybe you'll have to build your server storage like traditional EMS does. You may or may not have cooling reference designs. So Flex does that end-to-end, and on top of that, we do the services afterwards.
So that means pulling out your server storage boxes , refurbishing it, and putting it back into the aftermarket if you need to. So we do that in terms of the overall compute space. On the power side, we have a very differentiated story. We do two things in power for data centers. One is we do power within the rack. So if you look at what sits within the enclosure itself, you need power to power the GPUs, right? And so that's what we call Embedded Power. And so it is power that finds a way to power the GPU and then manage all the efficiencies associated with the rack itself. There are very few companies who do that, couple, you know, two companies probably, and they're a lot smaller ones.
So Flex is one of the largest people who plays in the embedded power space. We work with the, you know, with the chip companies and the hyperscale companies to design the power envelope, right? Including the power shelf. It's our own design, and we take it to market. So long lifecycle in terms of technology and deployment. So someday, if you're gonna go to an 88-kilowatt rack or a 200-kilowatt rack, we're thinking about it and we're designing it for you, and we're looking at figuring out how do you take all the heat out from it and build a cooling profile around it. So that's what we do in terms of embedded power, power within the rack itself.
And then we have another power business, which was an acquisition we did 3, 4 years ago, which looks at power outside the rack, so that is switchgear, power distribution units, busway, all that stuff that delivers power to the rack. So which is still inside the room, so it's inside the data center. We're not doing things outside the data center, but it's delivering power within the room to the rack itself. That comes from an acquisition we did 4 years ago called Anord Mardix, or 3 years ago, and it's a traditional power distribution company, building switchgear, PDUs, all of that.
I think the cool part about where we are positioned in this, and we haven't talked much about it because we were building it out the last few years, and we knew our thesis always was power is gonna become a thing for, for data centers, and it's gonna become harder and harder. And then in the last year, of course, this conversation exploded, right? It's all about power and availability of power. But what's unique about what Flex can do is we at can focus on your power within the rack, develop and deploy and figure out how do you want to scale up to 100, 200 kW rack. And eventually, the integration of the power coming into the rack is going to become a technology problem to solve.
We are one of the few companies who can build that power architecture and think about how we solve for that moving forward. So we're in a great place in terms of power itself. We said we do $1 billion of power products today, around a couple of billion dollars of compute, so around $3 billion kind of overall data center space. We said it grows around 20% because that seemed like a good number and, to, you know, to throw out there in terms of growth, the opportunity is big. And so that's kind of our focus around AI as overall. So really very, very interesting and unique value proposition, I would say, both in terms of compute capability and integration and all of that, and then power integrated with cooling and putting it all together.
It's a very unique capability that we're building out.
Revathi, can you clarify, do you have one large hyperscale customer? How broad is Flex's reach in AI-related revenue? I mean, is it just one customer, or do you have more than one customer?
First is, you know, the question around hyperscalers is interesting. There's very few. You can count all of them in, like, one hand, right? So, there aren't many, but, we work with multiple hyperscaler customers, not just one. And so we're diversified, I would say. On the power side, we work with multiple hyperscalers, but multiple colos also, right? So that is very diversified. So I would say, yeah, fairly diversified in terms of our overall kind of data center portfolio.
And maybe if you can help clarify, you know, in terms of the recent hyperscale win that you've talked about in over the last six months, was that a competitive win against another EMS company, or how did you get that business? What was the outcome of that?
I'd say there's so much growth to be had. We don't have to, like, win versus somebody else. There's enough growth to go around for everyone. How we win is the unique differentiation that I talked about, right? We're not just selling an integration solution for customers. We're selling everything from, I will stage it for you, integrate it into your data center, I will build cooling solutions, I'll integrate the power shelf for you, and I will manage the end-to-end life cycle that comes after that. So I think we win because of that value proposition, and it's not a zero-sum game. There is enough growth to be had for everyone in this space.
So, Revathi, I want to ask you a question that, you know, you talked about liquid cooling. That's one example of the type of things that EMS companies are doing. How is Flex unique in this? Like, what is the competitive advantage that Flex has in this?
S o doing liquid cooling itself, having a design for it, you know, whether it's single phase or three phase or someday maybe immersive cooling, that's not kind of a big deal, by the way. Like we can build reference designs, and we can customize it for the data center customer, and you can do that. I would say for us, if you think about our cooling design, w e're in the middle of designing this power envelope for customers, so we can see where power is going to go, and we have to build cooling solutions thinking around that power, that growth of power that's gonna happen. So, we're in a very differentiated, you know, position, I would say. The EMS companies aren't doing that, right?
That's more the industrial power companies that are focused around that. So when we think of cooling solutions, it's about integrating, thinking about the future of power, right? Which is going to be very different than where things are today. So I'd say that's how I would say we are differentiated. Just building, like, single-phase, three-phase power solutions and cooling solutions is good, but the future is all about what type of cooling are you going to need when your power envelope is gonna change so significantly, and who can think about that and design that? And I would say that is a very differentiated value proposition.
Just staying on the AI theme here, one concern that investors have is this is a very competitive market. So thinking about margins in both the module stuff as well as the power stuff, do you think that because of competition, this is dilutive to your two-segment margins? Or, what guardrails do you have, and how are you thinking about margins in both of these segments?
First, is not at all dilutive. In the power side of it, envelope, we've said kind of mid-teens, you know, margin, profile, and you just have to look at industrial power companies, and you can see where their margins are at, and that's kind of where you should be. So we've set a very conservative view of mid-teens, and the reason being because there is technology investments, there is commercial, application engineering investment, so it behaves as a, like a product company should be. In terms of the services side, which is, how do I stage products for you to integrate? How do I pull your aftermarket, your, used products out and bring that to, the aftermarket? That should be kind of in the high single- digit, low double digit range because it's services envelope as a whole, so that's how that should behave.
I would say the whole integration business itself, which is more of the traditional EMS business, is margin accretive to the overall margin profile of the company. That's because of the complexity of the solution you're deploying, right? To be able to kinda cut and fabricate all this metal at scale and bring it together and integrate everything in with your cooling solutions is not an easy thing to do. So when people say it's margin challenge, I don't agree with that at all, because to deploy these things at capacity with the complexity that it's required, very few companies can do it. So if you're doing your job well, you should command a premium for it.
Got it. Another thing about AI is, how is Flex using AI internally? I mean, is it something that can help you save cost? And how do you think manufacturing over time benefits from having AI?
Yeah, I'd say that's our productivity story. I feel like the world of manufacturing, whether it's for us or any other company out there, has really not benefited enough from what we have seen in terms of AI or any of those kind of data improvement tools out there. And the reason, because manufacturing companies have very fragmented software systems deployed across their shop floors, and so to harness the power of scale and really use that information in a smart way hasn't been possible. So for us, I'm super excited in terms of a productivity improvement, which is going to be because of the combination of automation and AI, which I don't think manufacturing companies have harnessed the power of that. And we are seeing really good improvement in terms of deployment of that, but I'd say we're still in the early stages.
There's lots of room, you know, for us to do this differently, and I would say it's kinda just the starting point of where we have seen things. For me, the productivity story around that should be significant, you know, for us and everyone else, actually. So I'm quite bullish about that.
So we spent a lot of time talking about AI. Let's move into some of the other segments as well. Let's start with reliability and automotive. Can you talk, give us a quick overview of what you're doing with the internal combustion engine versus EVs? What is your dollar content per car, and who do you work with? Are you dealing directly with the OEMs or with the Tier 1s?
Yeah, so our automotive business, overall is a little low, is over $3.5 billion, as we have talked about. We talked more about that in the Investor Day, is what's beautiful about our automotive business, it's very diversified. So diversified around compute, power, and kinda all other stuff. Diversified around geography, North America, Europe, and China OEMs. And so very, very diversified portfolio. Within that, again, it's the compute and the power story. Around power electronics as a whole, we have our own portfolio of products, so we have our own DC-DC inverters integrated with PDUs and battery management systems. So we make the platform, design the hardware platform, and provide it to OEMs. They can customize it and stick their firmware software on it, so that's our power electronics portfolio.
We compete with, like, Tier 1s on that, and that's the fastest-growing part of the business. In terms of compute and infotainment and all of that as the largest portion of the business, it's agnostic to platform. So whether it's, you know, hybrid or ICE, it's all of it. You know, every car has a compute platform today, so we're growing in that pretty significantly. The nice part about this is, you know, doesn't matter where SAAR is going with automotive companies, like, I think we have said it's, like, couple percent, and we've said 10% growth in automotive just because of how the portfolio is positioned. Because if we think about compute and power in car, everybody wants compute and power in every platform, right? Whether it's ICE or hybrid or EV.
So automotive is a great story for us, similar to data centers. 'Cause again, we built our own power electronics platform a few years ago, and we saw the disruption of Tier 1s happening, and it's been great for us. We've been able to take full advantage of it. And then, Ruplu, one more thing I'd add is, one information we shared, you know, in the Investor Day is if you look at compute and power for automotive and compute and power for the AI data center space, by the way, a lot of similarities on each of them, right? Everybody needs more power, everybody needs cooling, everybody needs a compute capability. That is gonna be 40% of the company overall moving forward at, you know, through our three-year plan cycle.
Pretty significant portion of our overall business. Automotive is doing really well, and that's again the combination of compute and power products and being platform agnostic.
Great. Let's move to medical. I mean, Flex has always been one of the number one or number two in the medical device manufacturing, in the medical EMS space. Are there segments that you wanna focus on more than others? Like, is medical device manufacturing more beneficial than life sciences? Just give us your overall thoughts on this space.
I'd say, you know, we, we're both in the device space and the equipment space. Equipment is like a history and heritage of building large-scale, complex equipment. I would say less kind of focused around med equipment, more focused on med devices. 'Cause anywhere you need scale and complexity of of manufacturing is where we do well. And, med devices is all about scaling, right? There's millions of billions of little devices that people are using, and complexity is significant because anything that interfaces with the human body has complexity associated with it in terms of manufacturing capability. So we're more focused on the device space, than the equipment space, and that's where we feel like we have the best kind of potential to differentiate ourselves as a manufacturer.
Got it. The third segment within the reliability segment is industrial. We've talked about power. Now that you don't have Nextracker, is solar still a relevant area for Flex, or this industrial is so broad? Give us your thoughts on where your target segments are.
I'd say industrial, core industrial, which is outside of power, and has several kind of end markets that we're interested in. Industrial has tremendous opportunity in terms of outsourcing, and it is the space that is most capacity constrained that could benefit the most in terms of outsourcing. So solar and, you know, the whole energy space is definitely one that's of interest to us. Yes, it has a little bit of a pause because of interest rates impacting residential solar. IRA has some tailwind for it that could be helpful, but we are still very bullish on solar overall, right? So just because there's a little bit of interest rate pause, we don't think that's an, that's an issue long term. You know, we make everything from kinda charging stations in our industrial business.
But then the base of the core industrial business is just simple things like breakers and meters and things like that, that industrial companies wanna outsource because of capacity and capability standpoint. Robotics is super important for us there, warehouse automation and robotics. So those are the end markets that are important to core industrial. It's more kinda mid-sized volume, but more complex products, and that I would say is our target market for industrial.
Got it. Let's move to Agility. I mean, the CEC segment is still your largest segment. And we talked about AI, we talked about what you're doing for hyperscalers and colos. But are there segments within CEC that are lower margin? Like, I'm talking about the net-core networking, enterprise type, you know, IT products. Does it still make sense for Flex to focus on those segments?
Ruplu, the one thing I would say is, you know, five years ago, when people were talking about EMS being a 3% business, we said we'll shift the mix because the, you know, the available markets are so big, you should always pick and choose the end markets that you're, you know, you participate in. I would say the same applies. Look at CEC margins today, right? I mean, they've doubled in the time that we've seen that business in the last five years, and it's all about picking the end markets that you want to participate in. So you'll have choices to make, right? If you're so focused on AI and data centers and growth around compute and cooling and all of that, then you will have to refocus on, on that versus other things. Y ou'll naturally make those choices.
Our teams understand really well that life is about choices. You have to look at your mix and move your mix all the time. So in CEC, will there be a continued and natural shift of margins? Absolutely. You know, that's what our salespeople are paid to do, and we'll continue to do that. And, you know, I'm sure there'll be areas like, you know, telecom and, you know, parts of that, that, you know, we'll de-emphasize over time. So I think that's a natural progression.
Okay. Talk to us about lifestyle and consumer devices. I mean, consumer devices, you've kinda de-emphasized. Are there still things in there that make sense for Flex to focus on?
You know, the things we do today, we only do because we feel like there's a value that our customers pay us for. If it isn't the case, then we've already de-emphasized it. Our lifestyle business and our consumer device business, by the way, is the place we started testing the thesis of this end-to-end services. So it's not about I just manufacture X for you, like, I don't manufacture, like, a vacuum cleaner for you, but I will do everything end to end, which means that I will take it back when it ships back from the consumer. So talk about the e-commerce platform you need for that. I will remanufacture it if you need it and ship it back, or I will take it to waste, right? So, so for us, when you look at lifestyle and consumer device, it's not about just manufacturing X.
We focus only on high-value brands in this place where there is complexity of manufacturing, and then we wrap services around it, which makes it one of the most profitable businesses we have today. So we've shifted the thesis on that.
Got it. We have about a minute left. Can you give us your thoughts on uses of cash? Like, where are your investments going to be? Are you inorganic growth, internal investments? How do you see that piece?
Yeah, very simple. We'll continue CapEx around the 2%. We have enough room in that to make the investments we need. We feel like our share has a lot of room for growth. We feel we are undervalued, so buybacks will be a very important thesis for us. We spent $1.3 billion last year. We'll continue to spend at that rate. And lastly is that we will target M&A around power and around areas like power that integrate more of the kinda AI story. We have a very good portfolio, but we could add a few things in if properties are available at the valuation we want. So those will be the three areas.
Great. I think we covered a lot of different things.
Great.
Revathi, thank you so much for joining us today.
Thanks for having me.
Thanks for all the details.