Seagate Technology Holdings plc (STX) Earnings Call Transcript & Summary

Good morning. I'm Katy Huberty, and I'm really pleased to welcome Dave Mosley, CEO of Seagate. Dave took over the role back in 2017 after serving as President and COO, and before that, many roles across operations, R&D and sales at Seagate. Before we begin, let me just note that all important disclosures for Morgan Stanley can be found on our website or at the registration desk. And the way we're going to run this session is I'm going to ask Dave to just open up with some remarks, and then we'll jump into Q&A. So thanks again, Dave, for joining us.

Thanks, Kate. Good morning, everyone. Let me -- before I even start, just a quick reminder that I'll be making forward statements today. You can learn more about the risk factors associated with these statements in our filings with the SEC on our website, www.seagate.com. We'll start with the big topic that I think is on top of everybody's mind, which is coronavirus. This -- I'm really proud of how the supply chain has reacted, our employees, the suppliers that we have, the customers downstream. Everyone is putting the priority first where it belongs, on the health of themselves and the community and their families. And I think it's been -- for that reason, it's been disruptive through the month of February, but I think people are handling it very maturely, so -- which I really appreciate. From our perspective, these are supply chain issues that -- like we've talked about before. We've had some measure of supply chain disruption before. This one is more pronounced. But I think with our inventory positions and our ability to flex in the specific supply chain that are affected, primarily China, I think nobody is going to be really starved for parts out of the back. There continue to be some logistics challenge with people when it comes to moving goods or getting the right people into work. But I think we'll figure out the parts as long as everybody stays focused on the people. Then I think we're going to get through it from a supply chain perspective. And I also think that customers are not really going to be particularly starved at the end of the quarter for any specific parts that they want. I do think that downstream of us, that's -- these are all upstream comments. So I think downstream of us in some of the large integration sites, which people have been talking about, there are issues with labor and there are issues with logistics and things like that. And getting all that integration done is something that's still a challenge. And so we expect a very busy month of March. People either bought things that they couldn't kit at the time or there -- they need something else or they're still trying to figure that out. And so I think there's going to be a lot of swings in the March time frame for product that people are trying to hit the end of their quarter, and that may extend for a couple of months. But again, I'm really proud of how the end-to-end supply chain's put the right priorities on this thing. And I think we'll get through the parts situation. Supply perspective -- net-net supply perspective, I think this is manageable situation from what we've seen so far. February was a little bit slow. March, we're expecting to be really heavy. That's not uncommon in our business from a linearity perspective. And so we guided just a few short weeks ago, with a little bit different knowledge of how this situation was progressing. We guided with an opened up range, and I'm still happy with that guidance of the open up range. Again, the uncertainty of March is what it is. But from our perspective, the world has not stopped. It's going to be pretty busy to continue out the quarter. What everyone's talking about is demand, end demand. We can see subtle signs of changes in end demand in, for example, the distribution channels and some of the consumer channels. But linearity has not been that bad through the course of the quarter, especially in enterprise. And so, to some extent, we expect that there will be some impacts as the CEO and CFO go tell their COO and CIO and CPO and all the other people to maybe pull back to meet their quarter, but we think some of this is temporary. There may be longer-lasting effects, but that's the same uncertainty that everybody else is worried about relative to macro. Data products actually are -- the demand is fairly healthy because I think people need mass capacity storage to put all the data that's being generated, and you can make a temporary decision to not invest, but that decision may be at the risk of not being competitive or not coming up with the right financial models that you want going forward. So you better get back on the train. And I think some of our markets are maturing quite a bit as time goes by. To that end, that -- when we start talking about things that are different than the tactics of the coronavirus response. I think Seagate's in the middle of a very large pivot. If I go back about 5 years, roughly 20% of our drive were these mass capacities drives for the cloud, for surveillance, some of the edge applications. But the real -- I think the real big drives with lots of heads and disks in them. Most of our drives, at that point, were client server drives, notebook drives, single platter drives. So 1 disk, 2 heads and a lot of drives. So 20% back 5 years ago, is now -- we've now reached about 50% of our revenue is the mass capacity store. So it's growing, and we're to a point where we're going to -- we think it's going to continue to accelerate from here. So mass capacity drives, the cloud and the edge drives and these big pit stops for data are really in secular growth with the rest of the data products. The legacy drives, if you will, which we call that is now about 40% of our revenue. It's stabilizing somewhat. It's growing in capacity, but it's stabilizing. That's great. So if I look back over the last 5 years, we had just over $7 billion of free cash flow. The way we managed the business on the downturn of client server, was to be very frugal and make sure we watched our free cash flow. That's true of last year, I talked about this quite a bit in the last few earnings calls. The next 5 years, I think that $7 billion of free cash flow that we had in the last 5 years, I think we can easily do that or close to that. And our target is to do more, again, macro notwithstanding, just based on the secular growth that we're seeing in the mass capacity storage drives. We talk about free cash flow a lot. But I think it's to reemphasize the point that against our revenue, our free cash flow is actually quite good. So our ROIC is somewhere between 20% and 30% for the last year. And we just talked last earnings call, bottoming out at 20%, starting to go back up. We've made investments in CapEx because we're confident in our model. But I just want everybody to rehear the fact that we are very committed to shareholder returns through this period as well. And we've been -- we've had a fairly generous shareholder return policy of that free cash flow. So -- maybe a different way to slice it is I have more confidence in the EBITDA over the next 5 years than the EBITDA over the last 5 years, just based on the secular growth that we're seeing in the mass capacity storage. So that's how I would frame my comments upfront, hopefully to simulate some comment.

Yes. No, that's really helpful. Just because coronavirus is top of mind, I want to make sure I understand the message. So obviously, February was slow as all of your customers were dealing with the labor shortages. But at this point, you think March has the potential to be strong enough so that you'd still hit guidance. Obviously, there are unknowns along the way, not perfect visibility, but that's the view of the quarter at this point.

That's well said. There are some customers who realize they can't build the kits that they want, but no one is really giving up on any competitive position because they can't. And so they're still eager, as people do come back to work as the factories do come back online to highly prioritize these products as part of the overall integration, so they can go get revenue.

Yes. And the data points, whether it's the production data or some of other -- the other components players into the cloud market are suggesting that, that area of the business remains quite robust.

That's right. And remember, we expected some seasonality from our Q2 in December quarter into this quarter, and we expect the Chinese New Year to be soft from a labor perspective. And all these things. So the fact that February has been -- they have had some channels that have been slower is not really a big surprise. I think it was...

Not that it was normal.

So stepping back and looking at the numbers. I do think that people will get back to work, and at some point, be pulling product. But it's going to be a very busy March, I think.

Yes. Okay. Great. So let's step back and talk about the long term here. You have partnered with IDC over the last couple of years to try to size the storage market opportunity particularly around some of the new applications like edge. And the conclusion is that the hard disk drive demand could actually grow faster in the next 5 years than it did the last 5 years. And so talk about how you see the industry investing in that opportunity? And is there the potential that at some point over the next couple of years, we could actually see tightness in supply if the market executes on this correctly and perhaps some better pricing dynamics?

So we've been saying for a while that we're out of the period where we make disk drives, and we're into the period where we make heads and disks, that's where our -- most of our OpEx and CapEx is actually pointed. People like to focus on the big disk drives because they have long test times, but that's a de minimis amount of capital and planning and things like that. It's really about who has the heads and disk for the next technology node and so on. I think the way we forecast it today over the next 10 years, demand for data is going to be enormous, and that's what the IDC numbers say, actually, the data itself is becoming more and more valuable as well because of all the innovation that's going on in big data. And there's more places that are -- will be needed to pit stop that data, whether it's for a minute or an hour or a day or a year or so on. And when you talk about those type of -- that type of architecture, disk drives are the -- play a central part. We're not talking about the compute centric architecture. So this is, call it, the first order, call it the cloud architecture, but it's extensible for the edge as well or on-prem data centers as well. So there are these converged solutions or hyper-converged solutions that will exist on-prem or exist in the cloud. But there's all the -- as -- all the while, when people buy that, they look to use that 24/7 or also buy it as a service, all the while, the data just keeps growing and growing and growing, and the uses of data just keep growing. And that's the fundamental driver. So our heads in media growth, our ability to go introduce new technologies or just invest in it is probably going to be behind that demand. And so I look over the course of the next 10 years, there may be times when supply is ahead of demand, but I doubt that, that will be the norm. I believe that mainly the demand for data products will be bigger than supply just because we won't be able to keep up.

And this is why you invested even during the cloud downturn last calendar year?

That's right. Right. So this last -- we've learned this drill many times, Seagate has, don't push too hard of something that's on the old product, rather get your factories cleaned up and get ready for the next peak. And then invest, and that's what we did. If you look at our CapEx, and we spent a lot of time shaping our product portfolio for the next peak. That's not a 1 quarter, 2 quarter decision, that's really -- we wanted to get to the next product. That's -- that product, as we've talked about publicly is going to be in the factories for years now.

Yes. Yes. You broke out -- you said -- you mentioned this in your intro remarks that you're breaking out new segmentation, the mass capacity market, which, as you said, is now 50% of the business. Those are applications like cloud and surveillance, where there's really no NAND cannibalization risk and the end markets are growing very fast. And in the mature markets, which are PCs and gaming and mission-critical enterprise, where there is this conversation around NAND cannibalization. You talked about this 50-50 mix. Where were we 3 or 5 years ago and where do you think the mix for the good part of business mass capacity will be 5 years out?

Yes. That's good. So we're not really a 50-50. We're at 50-40-10. So we have 10% other business. So legacy is waning as a percentage of our total revenue, but it's actually stabilizing quite a bit. In some of those segments that you've mentioned, the exabytes, for example, continue to grow. Even consumer products that -- exabyte was pretty healthy year-over-year. So 12 exabytes, the last quarter, we talked about that on our earnings call. Mission-critical, for example, exabytes continue to grow. But the mass capacity is the secular growth that we see going forward. From -- where are we in that cycle? We're about at this tipping point right now. We could have called it 2 quarters ago or 2 quarters from now, but we're about at the tipping point. I'm actually -- the reason we broke it out the way, we just -- it's not helpful to talk about notebook and mission-critical anymore. It's -- because we're not making investments in OpEx or CapEx really for that. We're making investments in the mass capacity storage, the big drives going forward. Some of the drives may only be 4 terabytes or 8 terabytes. You said NAND cannibalization. I don't think that's a helpful way to think about things anymore. Maybe it was when we talked about NAND coming into iPods or notebooks or something. But going forward, NAND is a healthy part of the ecosystem, even in the cloud, it satisfies this part of the workload. The hard drive satisfies this part of the workload. People know what that architectural expectations are, and there's probably not a lot of substitution going on. There's more just development of all that's going on.

Yes. And it's actually an argument that, that mature business going away faster would help provide capacity for you to chase the cloud and surveillance demand when it's all about the heads and disks.

That's true. And actually, if you think about the last 5 years, that was the story. We weren't really adding a lot of heads and disk capacity. As a matter of fact, most of the CapEx that we were spending in the last few years were really more around factory consolidation, build new factory buildings so that you could move and get a better footprint. But the future will be about installing heads and disk capacity. So there was spare capacity, if you will, coming down off of client server that we could use in the mass capacity, that spare capacity is kind of out. I don't -- the drives that are left in the legacy business have very few heads and disks. So even though as we talk about revenue being 50% mass capacity and growing, the heads and disks is much, much more than that already.

Yes. Yes. But the point is, with significant data growth and now 50% of your business tied to these growth markets, there's a real argument that Seagate should be able to grow the top line more sustainably going forward than it has in recent history.

Yes. And that's why we've -- like I said, we've taken the OpEx out of that and put it into that. We've taken the CapEx and pivoted over. So we want to -- and we've really pushed these new technology points, the HAMR transition, for example, we've been talking about for a long time. The platform that gets stability there. And we're happy with the customer relationships. Some of the customer relationships in the legacy are -- we don't run away from them. We continue to maintain them, but we help them do the pivot as well. And that -- I think all of that positions us very well in the predictable future.

Yes. So cloud is the biggest driver of that mass capacity segment. Last quarter, you highlighted that the nearline demand, which is the drives going into cloud, saw more diverse customers and applications, including Tier 2 clouds. Is this -- do you think this is a new trend, and that end market is diversifying?

I do. I think if I go back 5 years, for example, and Seagate put out a slide on our Investor Relations website that really highlights kind of our journey on this, you see these periods of peak and digestion. But those were largely determined by the behavior of 1 or 2 big cloud service providers at the time. And I think now we're out into a world where we're seeing such a bigger customer set. Some of those customers that you may have never heard of are growing quite quickly, and they're growing data center footprints quickly. And then the more mature it gets, the more you have to cycle through your existing infrastructure. So if you think about a data center, do you build yet another data center or do you replace an upgrade what's -- in that existing data center. As we -- as that whole market builds inertia and some of the smaller customers build inertia, we hope that we get into a different cyclicality. I do think it will happen. Of course, everybody says it's Tuesday today and what's going to happen on Wednesday. But over the long haul, we believe that some of these replacement cycles and some of these investment cycles will become much more predictable, and therefore, spread those peaks.

So just as you've said, when investors are looking into the peak of the cloud CapEx cycle, they're starting to ask, how do you deal with digestion when it comes. Is there anything that you've learned from past cycles that you can do to manage the financial impacts of going through 3 or 4 quarters of digestion?

Well, I think we just did. Actually, if you look at the last 4 quarters, we upgraded all of our factories, continued to invest in CapEx. We held the line and didn't ship a bunch of stuff in that was older stuff that we didn't need to. We stayed really closely connected to our customers, and we qualified the highest capacity disk drives that's in the industry and got on to that platform that we wanted to. And all the while, those last 4 quarters, we generated $1.2 billion of free cash flow. Had we not invested as much in CapEx, we could have said there was a bigger number, but we invested ourselves and we upped our shareholder return as well. So I think that was a down cycle. We're waiting for the next big up cycle. And of course, everybody wants to be able to predict the market perfectly like that, but we think we're into a period of long-term secular growth now that this is becoming a such a big part of our business. It's -- data is growing and people need places to put it. It's becoming more valuable. It's all you know...

The TCO for the highest capacity nearline drive is very attractive. So a lot of the cloud providers like to buy the biggest drive they can, and Seagate has a 16-terabyte, which ramped significantly in December, more so this quarter. Talk about your expectations around the ability to take share with the 16-terabyte and followed up with 18-terabyte later this year? And what are you seeing from a competition pricing perspective as your competitors having to compete with 14-terabyte in the market?

Well, I think nobody is really out there at 16 or 18. And from my perspective, market share is really an outcome of driving your product portfolio. So we have good conversations with our customers about how many 16 do you need? And when do you -- what do you need to get it qualified, and how many 18s will you need when they come out, how much 20s will you need and so on and so forth. The nice thing is the 16, 18, 20 is, this discussion is all on the same platform. There are some slight technology changes inside. But we've been changing platforms so much that, that was disruptive from a cost perspective and our readiness and a ramp perspective. I think we're through some of that now. So I think we're positioned quite well. To your earlier point, when we make the drives at 20 or 30 terabytes, I -- the TCO is huge. Would you rather build another data center to house that much capacity, or would you rather upgrade. And so I think people would say, I'd rather go to the next part, especially the drive that you maybe upgraded maybe a 4-terabyte or an 8-terabyte that's about 5 years ago.

And it's an important point that your cost structure on 16, 18-terabyte is more attractive than your cost structure at 12 and 14-terabyte. And so as you scale, the margins like-for-like should be better.

We made a big point of making this platform the one we wanted rather than just acing for an extra 2-terabyte, no. An extra 2 terabytes against a TCO proposition is big when you're going from 4 to 6 or 6 to 8. But when you're going from 14 to 16 to 18, it's a lot less so. So it's better that you have a good stable platform that you don't have to reintroduce parts and pay all the upfront tooling costs and get all the scrap charges and everything else as you're learning about those new platforms. So fundamentally, it's a better platform for us to drive cost.

Yes. There's a longer-term debt that you've made on HAMR, which is a different technology that at least at this point, your competitors are betting on. Talk about the advantages of HAMR versus the alternative technology MAMR.

Well, I would say that when you talk about HAMR, you really need to be talking about transitions that will take you into big jumps, not small jumps. So SMR, for example. I think we introduced our first SMR drive to the cloud market in 2014. I mean we've been talking about SMR for a long time. It's a trick that will get you a couple of extra terabytes or a few percent, if you will, and people would argue about what that is. But it's really just to an extent just small adder to the S-curve that we've been on for perpendicular. I view a lot of other technologies as nice to have that will get you a little bit, but they won't give you the big jumps. I mean HAMR enables fundamental new media technology that allows you to have smaller grain sizes, better signal and noise, which is the way we need to drive this market. And it's -- when there are 50-terabyte drives someday or 60-terabyte drives someday, there will be HAMR technology. They won't be the legacy technology perpendicular, for example. So we're at the cusp of that transition. We've been investing in it pretty heavily over the last few years, there were times when there were doubters, and I'm one of them, but no longer at this time. The parts are really coming true. We feel really good about the technology development that we've been making in the labs. And we have the bag of parts. So we talked about having a 20-terabyte drive by the end of this year. Actually that 20 terabytes in 2020, I said that in an investor conference in 2013. So the net-net, the industry has been fairly predictable on this front. And that's because we -- it's all underpinned by science parts, not charts, right? We think that by 2025, we can be in the 50-terabyte-plus range as well. And we have the technology and demonstrations in the lab to be able to do that.

Is there the potential that you make the right bet and you execute on it, that this could be a real disruptor in terms of industry structure and market share?

It's always a potential, but I would also say that I really respect the competition, the suppliers to the competitors that some of whom are my suppliers as well. If I go back maybe 15 years ago, I learned a deep lesson to say we have all the technology you don't. So therefore, you just go away. Those suppliers turned around and said, oh, now I have spare cycles to go do experiments and look what they came out with at the time. So I think a much more mature way to look at the industry is the industry will move forward. We have to assume that everybody is going to stay competitive, and that's certainly what we assume.

And what does the -- how much are you investing in HAMR and what does that investment cycle look like over the next couple of years?

It all fits inside of our model. I think the investment has been big over the last 20 years in experimentation and over the last 5 years in parts, certainly, building parts that you can learn, building drives that you can learn about because we have to run drives literally for years to be able to make sure that we have the requisite reliability points. The average drive -- mass capacity drive is going to live 5 years in a data center, 7 years in the data center, something like that. And it's going to be expected to work 24/7 with no pit stops. So -- and this is not -- it's not a slow seeking device fluttering faster than a Hummingbird's wings. These are the kind of reliability specs we have to be able to meet in order to intercept with the next drive. So we have to do a lot of experimentation to do that. But I think we're getting through that period so that we understand most of the experiments from what we have to do from here on out. It's just as hard as it would be to make a 25 or a 50-terabyte hard drive. I mean that -- those are -- all the subsystems have to come together.

Okay. So shifting to surveillance, which is another big category within mass capacity. This is a market that is not only about China, but a lot of the demand has come out of China. And so how dependent is surveillance on the economy in China? And how do you see this business tracking everything going on with Coronavirus?

Yes. There has been some seasonality to surveillance over the years. If you looked at the curves that, like we talk about the cloud digestion phase, you might see similar types of buying patterns. And there are big customers that are influential in that. It is starting to change a little bit, but we were expecting some seasonality coming out of Q2 and going into Q3 anyway. I think it would -- my answer to your question would be it's a little too early to tell. The customers are strong. I will say that there's also secular growth in surveillance. There are certain applications that have come out of surveillance companies that are now smart city applications, they're not really surveillance anymore. And as you see that around the world, there are traffic and efficiency metrics and for people in how they live their daily lives, you're seeing municipalities starting to want to adopt those things outside of China as well. And so I think this market is going to continue to grow the tactics of seeing through this particular coronavirus issue, notwithstanding, I think it's going to continue to grow. And some of those applications are driving such efficiencies inside cities that we think that -- and that's an edge application for cash, right? We think that there will be a lot of growth there.

Intel also said yesterday that edge -- they've always viewed edge is an exciting market, but they're really seeing the use cases start to take off. What are some of the other interesting use cases? You introduced the Lyve Drive at CES for storage at the edge. How are you addressing those use cases?

Yes, let me -- thanks for asking. Let me talk about it a little bit differently because I hear of a lot of people talk about compute and edge compute. And they'll talk AI and what is that and it's software on compute and that is not a -- those are not solutions, those are kind of endpoint or edge problem solvers, but they're not solutions. The solution only comes when you talk about the data. The data that's going in or the data that's coming out. And I completely believe that all of that is going to grow. But when you talk about compute at the edge, and there will need to be more compute at the edge, okay, that means there's going to be more data at the edge being aggregated, moved from the edge back to the core and vice versa. And what we're finding is very interesting, is not all that data movement, not all that traffic is urgent. And some of that traffic actually requires physical motion. Now this is something that's -- counter the narrative that everybody else has out there in the world, but you can go down to various stores that are not too far away from here and buy consumer products from us, that are data transport products and are being used more and more for that. We believe there's an enterprise parallel to that, we have introduced this live data transport system. And it's really getting us into some great conversation that's about what is that data that you need at the edge? How much of it do you need? Is it all -- does it all -- right next to the computer or can it be more of a -- of an offline or nearline application, just the same way the cloud architecture would be? And then how does it start moving around the world? So a fascinating discussion. I would say the other thing is, and we all know this, is that the traditional ecosystem sometimes have trapped data and trapped data, I don't think it's going to be a popular way to think about data going forward. I think people want their data to be more free from one particular application, one particular compute -- they'll want to be able to move it around, physically move it around or else just be able to have it moved around. And I think that, that is the predominant trend in mass capacity for the next 10 years is how do you answer all of those calls.

Yes. I almost don't have to ask you a PC question in future years. But I'll ask one now, given there's a lot going on, but it's just not that significant to your business anymore. Obviously, we have Intel CPU shortages that the PC industry has been dealing with. We talked a little bit about coronavirus. How big of a headwind are those in the near term? And when do you see these headwinds starting to fade?

Well, the coronavirus is just -- I'll put that aside with my earlier comments. It's created a lot of uncertainty, but there's still business that's plowing through it with the right prioritization. The PC business, what I would say from our perspective is that the drives inside of PCs have actually stabilized. So people always want to interpret that as a notebook comment, but it's more of a desktop tower PC comment. They tend to be in high-end systems, sometimes with SSDs nearby, I mean, big gaming rigs are great examples of this where you use high performance storage, and then you use high capacity storage as well and you trade-off between the 2. I would say the interesting thing about the distribution channel and some of the other kind of smaller scale, but interesting markets, is that they're disrupted for various reasons now. The 2 are the ones that you mentioned. But they've also become more industrial than the past. It's not just white box PCs to everyone. It's more very specific applications that VARs are going after. So this is part of the legacy market stabilization. That's why I don't think it's a good narrative to say, well 2 things: I don't think it's a good narrative to say that legacy is going to 0; and secondly, I don't think it's a good narrative to say that you run away from those customers in legacy. They will pivot their businesses as well. We want to be there for them to -- as a solutions provider, and then we'll help them go through the pivot to wherever they're going.

Yes. So you talked about mix of the business, 50, 40, 10. The 10 is the non-HDD business. It's a segment that has generally declined in recent history, the margins are below corporate average. How strategic is this business to you, given the less attractive financial profile?

I think it's interesting because I don't really think about it as revenue growth, although I think there's some opportunity for revenue growth inside there, for example, the systems business. I think as we come off of client server, there's a -- everyone knows this, there's a panoply of different products that people can offer. So if I think about it as boxes to put hard drives in, there's all kinds of boxes to put hard drives in that there were 5 years or 10 years ago. The fewer number of boxes means the less integration that I have, the less that friction creates time-to-market problems. And so the more we can aggregate that, especially on the go-forward boxes, the better. And I think that's how we've really shaped the systems businesses. So it's good. It's not bad. It's actually enabling channels, either via some of our traditional customers or new customers for large capacity, mass capacity drives on edge, on-prem, so on. The SSD business is more one that I think about is that we have customers who value our relationships across HDD and other products. We continue to find ways to serve them and look out for places where we can stay inside of our model long term, where our brand has a shadow effect, and it's all about those customer relationships where we can continue to drive revenue and do it as -- with positive free cash flow, and that's the way we're really focused on those things. So I think those provide assets to us to be flexible into the future, especially as against a world where storage may actually be more compossible rather than more hyper-converged, if you will, I hate to point it out that way. But I don't think everybody buys just converged solutions in the future. I think data becomes a lot more fungible and more compossible. And I think there are opportunities inside of those assets to be able to answer that call faster.

One of the investor debates or questions coming out of last quarter was about margins. And we can talk in a second about where we have a focus on operating versus gross margins. But investors have sort of looked at the gross line for many years and are working through that transition. But the reason that it was a focus as you guided roughly flat gross margins, yet, you're shifting towards 16-terabyte, which as we talked about, is a more profitable drive. Just remind everybody what are the factors that are offsetting the 16-terabyte mix and landing you at a flattish margin guidance for this quarter.

Well, in particular, this quarter, we've got seasonality that some of the legacy markets are changing very quickly. Now, like we said, there's uncertainty into March. So we have to continue to factor that into the final numbers. But that seasonality was something that we saw. So there's the secular growth that's going on, 16-terabytes, if you want to use that as a proxy. There's a lot of other drives, I think because we can make a 16 or 20-terabyte, we can also make a 12-terabyte cheaper and 8-terabyte cheaper for other markets, right? So that secular growth of the mass capacity storage is offset by this legacy, and the legacy still does suffer from some of the seasonality, if you will.

Yes. So I mentioned that the company is really focusing investors more on operating margin, which is closer to the bottom line EPS and free cash flow that ultimately should drive valuation rather than gross margin. Just talk about what drove that change. And why you think operating margin is the best metric to look at?

Well, I do think it's better to look at the top line, we want to grow revenue. Look at the bottom line, and we're committed to the 3% to 5% growth that we said. And then if we can expand that to 2% to 6% or even higher than that, we'll take advantage of that. I don't really care whether that $10 building in front of me is 25% and dilutive to my gross margin when I take it. I'll take the cash at the bottom line, too. So I'd like to focus on the top line and the bottom line rather than one of these middle lines instead. Now it is an outcome of the model, the gross margin, of course. And to the extent that, like we were talking about much earlier, that we make investments in CapEx and supply becomes behind demand considerably. And then I would expect gross margin percentage to grow. But how are we managing free cash flow, for example, which is what I talked about, more CapEx investment against that. It's all a balancing act. We are managing for that operating income range. We were at the top end of the operating -- 13% to 16%. We're at the top of the operating income range. And I think those are the things that we want to drive that revenue growth to 6%, even higher, right? And then we want to try to manage the operating income range. And if we can manage the operating income range, that's a better outcome.

Yes. I guess, I'll ask one more and then we'll see if there are any in the audience. Just carrying on this margin discussion at the Analyst Day, you guided to long-term or medium-term range of 13% to 16%. That's similar to what you've guided to in the past, despite the fact that you are investing in a way that you haven't in many years. The bets that you've made are -- seem to be paying off. There's accelerated data growth and potentially this tightness in capacity. So there's a lot of reasons to believe that maybe the margins could or should be higher than that. So just talk about what you think the upside and downside risks are to that range and whether it might look different in the long term.

Yes. I think we want to make sure that whatever we have is sustainable. So the capital that we're buying, we believe we're going to be able to use that capital well over the next few years. But when we start to talk about shifting the ranges, we want to make sure that, that's sustainable because what's the risk? The risk is somebody comes out and says, well, you're outside the range again just because of something that's going on in the macro. And it's still a time when mass capacity storage is only 50%. It's going to grow to 60% or 70% over time, but it's only 50% right now. There still is a little bit of volatility. So we have to watch that. I think maybe that's -- you call that conservative, and I'm not looking 3 to 5 years out, but I am from a cash flow perspective, looking 3 to 5 years out and saying, we should be able to at least do what we did in the last 5 years. And our target is to grow even more of the...

More. Yes. Okay. Good. Let me stop there and see if there are any in the audience. There's one here.

There were some reports last month around some of the major games and apps in China, experiencing freezing or login problems and some expedited orders on servers to support higher peak users. So I'm curious if you saw anything like that in terms of expedited orders from some of the hyperscale vendors overseas?

Yes. I'm not talking about anybody specific, but I would say that -- no, it's a good way to think about us in the cloud is the high performance front end is probably not us as much. And then that data is being spun out of that high performance front end into mass capacity and then back in when it needs it. So I don't know that we're necessarily tied to that particular behavior. I think that this -- there's an investment profile around this, there's an investment profile around this. When you get into -- when you get behind the investment or you have a hiccup, I don't know what this particular issue was, but that's when the end customer sees the problem. They probably don't see the fact that you're out of mass capacity storage. That may be handled a different way.

Yes. Any other questions? One thing that we -- area of the business, we didn't talk about is mission-critical enterprise, which is still important to many of your customers and your biggest competitor has essentially exited that market, NAND prices potentially rise this year. Is that a segment that could actually be flattish or grow?

It could. I'll tell you the capacity is still growing. Exabytes are still growing in that segment. Units -- the funny dynamic because a couple of years ago, there were people that were still using 300-gigabyte boot drives, if you will, in that segment. And then so units went down because people said, well, I'll just put an SSD in front of that. What's happened in units now is most of the systems are either legacy installed -- and think about it, there's probably at least 100 million slots out there for 2.5-inch 15-millimeter high form factor drive. And those are systems that are running out in the field that are aging. Some of them may be 5 years older or longer. And the -- I tend to think that the IT professionals are really astute. I don't think that they're just waking up and haven't heard about the magic of the new product you're launching. I think they know these systems really well as well as any other kids. And they know what performance is -- what performance specs are needed in there, what performance specs work. They may look and say, I'll replace this hard drive with an SSD, they may. They may work and look and say, you know what, I put an SSD in there, it cost me more, more than 2x than a 10K drive, and it doesn't really add any performance, so why am I doing it. They may actually look at -- it's a to you 12 or 24. They may look at it and say, I've already got 10 hard drives and 5 SSDs in this thing, I don't want to really change it. So the legacy stuff and the experimentation, that tail is going to be very long. We're seeing the capacity points growing to 1.2 and -- 1.2, 1.8 and 2.4, which is what we have in 10K and fairly stable. I think if flash raises prices, exactly to your point, people will continue to make that trade-off and that tail's going to be quite a long tail, I think.

Okay. Any last questions? Yes. In the front.

Intel, yesterday, talked about when cloud customers start buying, and you see it from multiple quarters, and they gave 3 to 4 quarters is how long they've seen cloud purchases, kind of, continue. When we start to think about the nearline coming out with 14, 16, 18 and you're having so many SKUs in a relatively short period of time, does that provide an opportunity to elongate some of these cycles?

Yes. And I do think that the hard drive, the mass capacity back end into the cloud is very different than network and compute, what tends to be more towards the front end. So I think you have to be a little bit careful. But I do think that product complexity affects everybody's cycles, and it affects their qualification cycles of the customers and how they look at things. It may affect their investment cycles as well, whether they're repurposing off those 4 terabytes or 8 terabytes that they had, whether they're building new, do I wait for the new qualification or maturity of the new product line. I think it all plays in. I would say that in the mass capacity world, there's more customers. Some of them are not waiting because they're doing that massive install that may be on the still early adoption growth curve, and they need that to go out and compete in the market. And so it's more about what's the best TCO value that I can get right now, how fast can you ramp it, get me enough of what I need against the build platform.

A lot of investors, we talked about this before, are already asking, when's the peak? Is June the peak? It feels like we're really only maybe 1 quarter into a recovery in cloud demand.

Peaks and values are getting widened, I think, over time because of all these other secular discussions that we're having. And the growth is what we're really focusing on. We'll manage the tactics.

Yes. We do have 1 minute left. I just wanted to -- you brought up capital return in the beginning. But I think it's an important point because it has been significant. Just talk about the capital shareholder return framework that you're applying.

We've been committed to shareholder return for quite a number of years. We have a healthy dividend. And also, we've been buying back stock. I think we -- if you look over the last year, it's when you have $1.2 billion of free cash flow after already investing as much as you think you need for CapEx, given that you have kind of leading technology and best -- therefore, best visibility into how the customers are going to be buying next year. I think from our perspective, there's no reason to change anything in the shareholder return. As a matter of fact, the confidence that free cash flow raised the dividend last year, we expect to get on a steady pace of relooking at that as well. Going forward, we're very happy running the business that we have. We have a lot of flexibility inside that in OpEx and CapEx to reinvest not only in ourselves, but in other experimental markets, and we don't really see any need to change anything in the model.

Great. Thank you so much for your time.

Thanks, everyone.