All right, welcome everybody. While everybody gets settled, we'll get very efficient with the disclosures.
Okay.
We're going to read Morgan Stanley's first. For Morgan Stanley's research disclosures, please see the website at morganstanley.com/researchdisclosures. If you have any questions, please reach out to your Morgan Stanley sales representative. And then I'm going to read NetApp's Safe Harbor disclosure. Today's discussion may include forward-looking statements regarding NetApp's future performance, which are subject to risk and uncertainty. Actual results may differ materially from statements made today for a variety of reasons described in our most recent 10-K and 10-Q filed with the SEC and available at netapp.com. We disclaim any obligation to update information in any forward-looking statements for any reason. All right, I passed the reading test.
All right.
All right, well, Mike, we are delighted to have NetApp here with us today. Let me see, we have Mike Berry, CFO of NetApp, here with us. I'm Meta Marshall. I cover the storage space here at Morgan Stanley for those of us who don't know me. Obviously, it was a big week last week for you. Fiscal Q3 was another bright spot for NetApp, seeing record gross margins and operating margins and year-over-year growth for the first time in four quarters, seeing 21% growth in ARR of flash. That was 6% sequential growth. Maybe starting with the top line, just how are you seeing the current macro conditions with customers? And for everybody in the room concerned, we will get to AI questions, but I'm going to start with the quarter.
Well, so I thought I'd talk for 35 minutes and then just shrink that time.
Yeah.
Meta, thanks for having us. We're thrilled to be here.
Thank you.
Thanks for playing Fine Young Cannibals to get us ready. We appreciate it. Yeah, we had a really strong quarter last quarter, although we were guiding for growth.
Yeah.
We were super happy to get back to it. So when we started fiscal 2024, we did say, "Hey, it continues to be challenging but relatively stable.
Yeah.
That's largely how it's played out. I think now they're talking about economic growth similar to 2023, which is great. We've seen some pockets of strength, but it's still cautious. We still run into a little bit of elongated budget cycles and approvals. Hey, higher interest rates move all of those decisions up a little bit, so we've seen that. But as we look, and we talked a lot about our enterprise customers took a little bit of a step back in our fiscal 2023, we have seen some of that come through. Not all of them. There's still some cautiousness there. But in Q3, we saw some enterprise strength and some really good public sector strength as well. So we expect it to be reasonably stable. Still uncertain. There's so much going on in the world, but we feel really good about our product position.
Okay. Maybe part of that product position that you feel really comfortable about is that the C-Series has been a great kind of early success. Just what customer types or use cases are you seeing adoption of the C-Series be strongest with?
Yeah, so we are super excited about the C-Series. We had a QLC product before, but with the introduction of the new capacity flash, it's taken off a lot more than we thought, which is wonderful. I do want to make sure we hit the go-to-market changes in conjunction.
Yeah.
Because I think both of those go together. So a lot of the use cases is the great thing is it really fits all verticals in all parts of the world as well. We see not only in the U.S. but other areas. Virtualized environments are some of the areas we see. Databases, secondary storage, backup, and even some AI. I'll throw that in there. We've seen that as well. And the good part for us is we really expect capacity flash to, as a lot of that mid-range hard disk drive arrays come up for refresh, we think it's a perfect area for an upgrade to capacity flash. And we saw George talked a lot about some of our competitive wins and displacements. In the past, we haven't had products that have enabled us to go after net new workloads in our customer base, but also net new customers.
So we saw, for example, three competitive displacements across Europe and in the U.S. as well. Those were really focused around unified storage in terms of file and block and data management. We actually had a couple of displacements where we sold A-Series and C-Series together. So it fits use cases that we were not able to get to before. Either we had to kind of bring high performance down or hybrid up. I do think just really quickly on this, the introduction of capacity flash was very important, but also the go-to-market changes we made at the beginning of the year to really focus the client executives, our sales engineers around the core storage, especially flash, and then separating the cloud piece, the cloud specialist has really helped.
As I've told a lot of investors, I think one plus one equals more than two here because we did it at the same time. So great product, wonderful use case, fits a niche we weren't able to get into. We think that that will drive outsized growth versus other areas of storage. And it allows us to get to net new workloads and new customers we weren't able to.
Okay. And this hasn't played out in the results, but I just want to highlight it. People were concerned about, "Oh, would this kind of erode some of your higher-end flash market opportunity?" Clearly hasn't been the case, but you can probably give a better explanation as to what you're seeing in terms of new markets that C-Series is addressing.
Yeah. And so importantly, let's focus on two big rationales for C-Series. One is to get us into those markets we couldn't before, and we've talked about that. The other thing is, and this was on purpose, we knew we would cannibalize the hard disk business, the hybrid. And this has been the big driver. And gosh, we might talk about product margins sometime later today. It's been a big push in product margins. And that goes to as those 10K arrays come up for refresh, that especially now with the compression of the NAND market, from a total cost of ownership, it's almost the same as hard disk drives, depending on how you run your numbers.
We really think that that's going to translate into more capacity, both for us as well as us enabling to go after that huge mid-range installed base of some of our competitors.
Okay. Okay, perfect. We got to question three before I ask, so.
Okay. Thank you for that.
As you look through the AI opportunity, there has clearly been a lot of enthusiasm post the quarter about storage or just the AI opportunity with storage. Just how do you see NetApp's value proposition in the various steps of AI, and how are you seeing that play out with customers?
Yeah, so let's go through a little bit of this. And actually, we had a bunch of announcements today that were perfect timing. So first of all, we've been in AI for a long time. We have hundreds of customers that have trusted NetApp for AI. You could say that's more predictive AI, machine learning, drug discovery, oil and gas discovery, financial services, a good bit in different governments across the world. So our customers trust us on their AI journey. So that's number one. Number two is, hey, we've had a long partnership with NVIDIA. A number of our competitors have as well. We just announced today certification in their latest pod for our C-Series. We were certified with A-Series before. We're also one of the only storage providers, actually the only one certified on their OVX platform as they continue to widen their breadth as well.
We just introduced new AI reference architecture for our FlexPod, which is our go-to-market with Cisco. We've had a longstanding relationship there, and we've continued to expand on that. The other big part about Gen AI is there's a lot of work going on in customers around the world just getting their data management in order. Where is their data? Do they have the governance? Do they have the right security so that they can responsibly roll out AI? The use cases that we're starting to hear from customers and see in the base is training is obviously the biggest piece, very compute-intensive. A lot of work around data lake modernization. Inference now is starting to pick up. I think somebody talked on their earnings call, NVIDIA, about 40% of workloads are there.
We're really starting to see now the model fine-tuning and this RAG, Retrieval-Augmented Generation, I believe it is, where people are doing large language models on the internet publicly, and now they need to bring it inside and use their own data and enterprise data. That's all the model fine-tuning that we'll see. I think that will really be the driver of storage, especially as we talk about unstructured data. So we feel really good about our AI journey. We've done a lot of it so far. It's still early days for generative AI, but we feel like we're in a really good position.
Maybe just outlining, what is some of the timing? It's early days, but how do you see kind of the timing of those three different bands of kind of interest? Is it different storage needs or different pieces of the NetApp portfolio that kind of get used in those various stages, or is it just kind of increasing amounts as you go over time?
So I think that that will depend on the customer and the workload and also how much is in the cloud versus on-prem. And I think that's still to be decided in terms of the cloud is such a great place to burst data. And we do see a good number of customers doing their LLMs in the cloud because otherwise you're going to have to buy get everything implemented. So they'll burst it in the cloud. But then once the data gets to be a certain amount, and obviously there's cost that goes with that, we're starting to see it pull back. So I think some of that will sit on-prem. And that's really where I think you'll see, hey, C-Series, all performance flash. Some workloads will drive different usage.
And because so much of Gen AI is around unstructured data, that is certainly much more storage-intensive than text files or other data. So we'll talk a lot about or more about the timing we see when we do our analyst day in June as we roll through 2025. I think it's fair to say, hey, we've done a lot of predictive AI, if I could use that phrase. Gen AI is just incremental to this piece. And we do think that that will help in 2025 when we're still working through with our customers.
Okay. And you just alluded to it, but how do you think your public cloud solutions for enterprises kind of fit into all of this? So you said, "Okay, maybe I'm using my public cloud for more bursty." Are they using your NetApp public cloud solutions? And there, we just spent a lot of time talking about C-Series, but where do we kind of bring in your cloud portfolio into that?
Yep. So just to take a step back for folks that don't know, in our Cloud Storage, we are not only the only provider with a first-party natively integrated Cloud Storage solution in one of the hyperscalers, we have all three. And we've seen some of our customers as well as some of the providers of AI solutions using, for instance, FSx for ONTAP to train their models to do that. At Google Next, we showed how we work with their Google Vertex. So because we're natively integrated, it's super easy for us to connect with our customers into those solutions. And I think that you will see, especially as they roll out the large language models and they want to burst to the cloud, them use those. Where it ends up is going to be interesting in terms of on-prem to cloud.
But we feel like we have a huge head start because keep in mind too, especially our first-party native solutions, they are focused around high-performance workloads. And if you want to run Gen AI in the cloud, there's a really good chance that our solutions are the best fit for that.
Okay. Maybe just before we kind of jump into other areas of the business, you've mentioned kind of some of these competitive displacements. What has been kind of the biggest driver of that? Is it the breadth of the solution? Just where are you kind of seeing the biggest opportunities for competitive displacement?
Yeah, great question. So I think there's a couple if we could talk about product and go-to-market. So from a product perspective, C-Series gives us the opportunity to go into those customers that we didn't have something to go sell them before. Or if we did, they didn't think of us as a capacity flash provider. So that's number one. When the block-optimized solutions roll out and they're rolling out, that's another area for us as well. I would go back to probably the biggest short-term driver are the go-to-market changes that we made, not only to focus the sales team on that, but also, hey, we've moved the majority of our renewal business out of the core sales team into the customer success team. We talked about that on the call to free up more capacity.
And hey, refreshing our base is hugely important, but net new workloads and then going to get those competitive displacements because that's a lot of capacity that maybe was on the renewal motion that's now focused on go-drive net new sales. So you add all those things up. I think go-to-market is the first piece, and then product will carry the day.
Okay. Okay. Got it. Before I step off of AI, does anybody have any other AI questions they want to ask? Okay. Speak now. All right. So moving on, your public cloud business has undergone a number of changes over the last year. Can you just talk to some of the changes in go-to-market and within that product portfolio?
Sure. So just for everyone in the audience, so if you look at the public cloud revenue, now about 65% of that is in Cloud Storage, and then the 35% is in CloudO ps. What we've talked about is our focus is in when we did the strategic review, we really want to focus on our native first-party offerings, which is, again, across all three of the hyperscalers' different products, and marketplace. And I'll talk about that in a second. I don't know how much other folks see, but this kind of vortex of the big three's marketplace to suck dollars has been pretty significant. So that's one part of Cloud Storage. The other part is what we call bring your own. This is where they'll buy CVO or our software solutions either from us, and then they'll deploy it in the public cloud.
What we're seeing is that's more of a subscription business. That is being pushed to first-party or marketplace because every big company in the world has made commits to one or more of the big three. And when everyone was spending a lot of money on cloud, that was easy. And then when we hit a little bit of the softness and they pulled back on cloud, now that gap became bigger. And them wanting to burn down their commits. So why buy it from us when you can buy the same product in the marketplace and burn down your commit? So we think that that is a trend that we're going to continue to see. But that all goes to the growth in first-party and marketplace. Those businesses grew 35% year-over-year this past quarter. So that's a big part of Cloud Storage, the strength.
Where we're looking to optimize is we had a couple of smaller products that we killed. We stopped. Also, within Cloud Storage, we had some products that really weren't products. They were more features. We're going to stop doing that and roll it into the core product. And then the big work is around our Cloud Insights, which is a great monitoring tool for storage. And we got a little over our skis on, hey, it does a little bit of network, and it does some security, and it does some optimization. It's a great storage monitoring tool, and we're going to refocus that to support storage on-prem and in the cloud. That Meta has been the biggest drag on the subscription that we've talked about. And then Instaclustr as well in terms of the open-source database really supporting storage. So a lot of the work was bringing that down.
It's going to likely be smaller exiting the year, but more healthy. In addition, when we talked about the go-to-market changes, as important as that was for the hybrid cloud business, the other thing that we did is when we had the cloud specialists, so we had cloud teams faced off against Azure, Amazon, and GCP. But when we rolled them under our org, they were aligned with how we go to market. Well, the other three go to market three different ways. So there was a lot of angst and concern from them with just, hey, I don't know who to call. I used to call Meta. Now I have to call Mike. What about Gloria? So when we did the go-to-market changes at the beginning of the year, we pulled those teams out, and we aligned them vertically with their hyperscaler.
That's been a great success because now we're so much more aligned because even though their sales team will sell it, it's their product. Once it gets to be rather technical, then they bring in our team to help. That's helped as well.
Okay. I mean, visibility had been an issue there over the past year. How does moving to, in my mind, first-party would give you less visibility, but it seems as if you have more visibility now? Is it because kind of the more technical experts are aligned in there, or just, how have you gained more visibility of that business over the last year?
Yeah, great question. Let's carve consumption out from subscription. We'll do subscription first. The visibility we lost was how much of that business was going to get renewed, or especially in CI where there were big plans to go to the cloud, they pulled back on those, and then they may renew, but for a lower amount. We lost visibility there. On the consumption business, the lack of visibility was when everybody optimized. Everybody was—all three hyperscalers in massive numbers—growing significantly. Then when the economy softened, they optimized, and that was the visibility. The cloud specialists now, we're so much closer to that business in terms of what that end customer is going to do. The visibility there is a lot better. Are we surprised every once in a while? Sure.
Because at the end of the day, all three of those, that's still their customer, but they're also ours as well.
Okay. Okay. Perfect. And we just spent a second on it, but cloud optimization has clearly, when we were here a year ago, was kind of a big headwind. It seems like we're past peak. But just, are you seeing encouraging signs there? What signs are you seeing there just about kind of improvement within that cloud business?
So we are encouraged by it. It seems to have, while it's still there, lightened up. You hear all three of the hyperscalers talk on their calls of, "Hey, it's still out there, but it seems to have stabilized." The nice part as well is that I think now where there was I don't want to say overdeployment, but there were other folks where they had really deployed quite a bit in the cloud, and then they realized, "Gosh, that's a lot." And they pulled back. We had talked about one particular workload that was like that. So we have not seen that growth. Therefore, we have that visibility. So we are encouraged. It is stabilized. That business all along, Meta has you get some customers that go up and some down. But from a downside perspective, it's a lot more stable.
Okay. I mean, in the past, a lot of that cloud business that you had sold was to kind of net new customers. And maybe there hadn't been kind of the highest penetration or below what some investors thought of penetration of that into your install base. With AI and just renewed moves into the cloud, are you seeing a pickup here? Do you have initiatives to kind of help with a pickup there?
Yeah. So thank you. The cloud has been our biggest driver of net new customers. Now, especially around FSx has been a really great generator of net new because those are typically customers that we wouldn't have. I think it is a little early to say Gen AI may or may not to the extent that it's focused on larger enterprises, maybe less net new. But that is the net new engine of the company. We do have significant programs to drive net new. And then really that in its cross-sell, upsell, we talked about the renewals moving to the customer success team. We also have a customer success team whose sole job is to cross-sell, upsell into that base. And if you look at the cross-sell numbers, I mean, it's interesting. You have some at the lower end, but actually our largest customers have the highest percentage of penetration.
They're the ones most likely to deploy into the cloud as well.
Okay. And then just kind of rounding out kind of the product portfolio, just have you seen increased interest in Keystone or as-a-service offerings for some of the more on-premise hardware, particularly as we've gone through kind of rougher macro periods?
So we have continued to see a really nice increase in Keystone, but it's not a universal mandate. For us, the storage as-a-service ultimate is the cloud. And because we're so uniquely positioned there, that's more the conversations we have with our customers. But Keystone continues to do really well. It has every quarter set records. This past quarter, it grew by more than 100%. It's not a huge piece of the business because it is radical. But any procurement model that our customers want, that's really what we want is however you would like to deploy. We're not going to push it one way or the other. Most of the time, what we see is it's an on-ramp to the cloud. Maybe they don't want to plunk down CapEx. They want to go to the cloud, so it's an interim.
Or from an economic perspective, because of interest rates, they say that's a better economic decision, and they'll go that path. So we have seen great uptick. It's not a huge number yet, but it's not a universal mandate.
Okay. You've spent a lot of time talking about kind of your changes in go-to-market over the last year. A lot of organizations don't manage through changes to go-to-market quite so seamlessly. What is it that you feel like kind of helped this go-to-market transition, maybe more so than we've seen kind of in past go-to-market transitions?
Yeah. In some of my prior licensed software, someone I think an investor once said, "When he hears go-to-market changes, he immediately takes he or she takes a step back." So I think the change here is the whole team bought into it. I think there was from a I'm not a salesperson, but I would also say, "Hey, those folks want great products to sell. They want a clear path to hit their number. And tell me exactly what you want me to do, not a whole bunch of stuff." I think where we erred, and we'll raise our hand, we, the collective we, is we asked them to do too much. And to that extent, it dispersed their resources. So I think they all feel better, most of them. Not everybody's super happy because, "Hey, I had this cloud deal.
Are you not going to pay me on it?" There's the one-off there. But it was the right thing to do from a company perspective. And I think more importantly, it gave them very clear guidelines and directions. Here's what we would like you to do. And I give kudos to Cesar and his team. Then the comp plans aligned to that. It was, "Here's the comp plan. Here's what we'd like you to do. Here's the best products in the industry. Now go forth and hit your number.
Okay. Okay. All right. Yes. We all have PTSD about go-to-market transitions, but this one seems to have broken the mold. So just focusing on that for a second. So let's move on to profitability. Clearly has been another area where you guys have stood out over the last year. Part of this is from mix shift. Part of it's from attractive commodity pricing. Part of it's from kind of costs coming off from supply chain. Just how do you rank the factors on gross margin increases year over year?
Yeah. So can we do year-over-year, and then we'll do quarter-over-quarter?
Yep.
So if you look year-over-year, the biggest drivers have been cost. And the two have been premiums. We talked about that. And do you talk about PTSD every time someone says that? PTSD. So thankfully, those are gone. Hey, it was $40 million-$50 million a quarter. We were pretty open about that. That was a big drain on profitability and cash. And then as NAND prices really started to come down, especially in the first part of fiscal 2024, that's where you saw the bump from, call it, 50% on average in 2023 up to, again, the mid-50s. We did 55% and 55% in Q4 and Q1, and then 61%, 63%, and then we got it to 60%. The majority of the quarter-over-quarter has been mix because we said we locked the large majority of our NAND purchases.
Throughout the Q2, Q3, and Q4, costs have stayed relatively consistent. That's really been the great growth of capacity flash. In three areas there. One is instead of us selling hybrid, which is at a lower margin, now you're selling a higher margin solution. All the net new is coming in as capacity flash at a higher margin. And then higher revenue does help because there's some fixed costs in that base. That's why we have felt so good about resetting after we saw the success of C-Series, resetting the structural margins because yes, NAND will go up, and it has. We have locked a majority of our purchases for 2025. You can't do it all. We have TLC and QLC, so just forecasting gets to be fun, albeit at a higher price.
It's really that mix change, and it's a material driver, especially when the net new comes in at a higher margin and you're selling less hybrid. That's why we felt so good about resetting it. Year-over-year, it's mostly cost. This year, quarter-over-quarter, it's mostly mix.
Okay. And so just to refresh for everybody, the last week's call or fiscal Q3's call, you noted that you now think high 50s-60% can be the product gross margins from mid-50s% traditionally. Just maybe level setting, what would get you to kind of the high 50s, and what gets you to 60 just as you think about factors over the next year?
Yeah. Great question. So assuming supply chain stays relatively constant, let's pull that variable out. Why we guided lower in Q4 is it's our Q4. And there's two big drivers for interquarter. One is pricing. And in terms of not only competitive, but from our team as well. Q4 is always the quarter where we push to hit our number and comp plans end. The other piece is the mix within the product family. And so if we sell more high capacity than low within flash, you'll see that move around. Those are the two variables in a given quarter, assuming costs stay relatively consistent. Those are really market factors. What do our customers buy? And then how is the pricing environment?
Okay. Okay. But I guess the question there is, and I know some of this we'll get into more at the analyst day, but it's just C-Series continues to be a huge success. Is it that you get to can we push above 60 again? Or I guess is it the mix of C-Series versus hybrid, or is it kind of the mix within flash that's kind of most influential to gross margins?
I think at this point, given the strength of all flash, it's probably the mix within the product family. It's pretty clear there will be some workloads where hybrid still works, which is great. Outside of some people think disk is going away. I'm not sure that's the case. So that is an enduring growth. I think we'll see that no matter what. So for us to be able to tip over 60 really depends on the mix within the products as well as just the general pricing environment. And I always want to. I'll just be blunt. Especially in Q4, I always want to leave room. If there's a good deal that comes in, I don't want to have to worry about, "Oh gosh, I told Meta a bigger number.
We're going to do the right thing for the business." That may mean lower margin percentages but higher margin dollars. We'll take that every day.
Okay. You also were going to utilize your balance sheet in fiscal Q4 for a lot of NAND purchases into fiscal 2025. You just mentioned you expect to kind of there might be some instances of modeling. But so you expect now with Q4 purchases to have taken care of the vast majority of fiscal 2025, or just how do you think about the purchases you're making this quarter?
Yeah. So we have done, as we call them, prebuys or price locks for not the large majority, but the majority of fiscal 2025 NAND demand. So a couple of things. One is it's a year out because we have to forecast QLC and TLC, and that gets to be a little dicey. Plus, all the manufacturers have severely limited supply. Some of the pundits are talking about really big increases. We're going to let some of that play out. So we feel good about what we've hedged. But there's some variability there, and we want some flexibility. And so you expect to see inventories grow in Q4. And then we will also most likely write those checks. And that's all baked into our new cash flow guidance.
But yeah, to the extent we can use our balance sheet, I'd much rather do the right thing and pay 90 days before to drive the margins. Even with a little bit higher interest rates, it's still a better economic trade-off.
Okay. You've been quite conservative on OpEx over the last year, particularly headcounts were cut, macro, some of the cloud businesses underwent changes. Just how do you think about OpEx growth and where you would invest incremental dollars, particularly just given this last page of questions, we're all about all of your growth opportunities?
Yeah. So keep in mind in Q4 last year, hey, we did a pretty significant RIF. And it was, I think, 1,200 headcount. And management doesn't take that lightly. I think when you have to do that, it's a failure of management to plan, me in particular. So as we've gone through the year, we want to make sure and invest in the right areas. But hey, we know there's some uncertainty there. So our goal is we want to invest where we know we can drive growth. And there's really two areas. One is, as we've talked about all the AI, is there's some roadmap things within, especially ONTAP because we are a software company around AI, some of those use cases, making sure the certifications that we're addressing where we think that market's going. There's also some things we want to do in Cloud Storage.
We have great relationships with all the three hyperscalers, but they're three different companies that have different roadmaps. So there's some work to do there. So we will spend some more on engineering there. On sales and marketing, it's more of there's certain areas where we need the capacity. Hey, we freed up a lot of capacity by moving the renewals to the customer success team, and we've funded that underneath the covers over the last couple of years. So those are the areas, especially where we can go drive net new, we'll do that. And it was before you joined us. I mean, when I first joined, we added 200 headcount. It was kind of a boom. Hey, we don't think we have a capacity mismatch. We watch this all the time. But there are areas that we want to fund.
In general, George and I's goal is, hey, we want to invest in growth, but we will invest in OpEx at a lower rate than revenue because our goal is to continue to drive gross and operating margin dollars and EPS. We will fund investments where we need to, but we have knockdown, dragout fights about, "Well, can't you reallocate? Is that really growing? Can you move those dollars?" And the whole NetApp team does an awesome job doing that. And you've seen it. If you take out the variable comp year-over-year, we're actually down OpEx. But yet, we are growing 10% product revenue this past quarter.
Okay. Perfect. Another part of your story has been kind of the significant amount of cash you've returned to shareholders over the last year and a couple of years. Just how are you balancing investments and these fights you're having internally versus shareholder return, particularly with just given how the stock has meaningfully kind of rerated over the past year?
As a CFO, I'll say, "Yes, it's rerated, but not high enough.
You're obligated to say that. So check.
Well, I truly believe it. When our P/E multiple is higher than the S&P 500, then we can have a glass of wine and have this conversation. So what we talked about two years ago at analyst day is that so the dividend is $2 a share. That's about $400 million a year. When interest rates were zero, that was a meaningful yield. Meta, we're not going to chase the yield curve up. It's a great foundation, but we're not going to be a dividend-yielding stock. So don't expect that to go up meaningfully. Then we said we will always offset dilution depending on stock price. That's between $400 million or $500 million a year in buybacks. The rest of it, we will toggle between acquisitions and more buybacks. Given the work we've done in public cloud, we've taken a step back in acquisitions.
We'll still look at them, but it won't be a meaningful driver likely at this point. That means that we will over-index on share buybacks, and that's what we've done. The last two fiscal years, 2023, we knocked down share count by 4% and this year by 4%. Our goal is if we don't do acquisitions, we will spend about 100% of free cash flow on shareholder returns. We still have between $500 million and $600 million of net cash. Importantly, hey, that investment grade rating is really important to us. We never want to have a conversation with a customer about our financial condition. We want to keep investment grade. We can run the business for less than zero net cash if we need to, but we like to have a small cushion.
Okay. And then maybe just last question for me. You've mentioned M&A. You guys have been acquisitive in the past around the CloudO ps business. Just how are you judging where inorganic investments might make sense, or where are you kind of evaluating those opportunities?
Yeah. So if you look across the M&A spectrum, we're mostly focused on what we call tech and talent, where it's more of a roadmap accelerator, build versus buy, versus, hey, when we did the CloudO ps deals to three deals, those were really adding adjacent businesses that had net new TAM. And we've now looked at that and said, "Okay, that probably doesn't make as much sense as the smaller acquisitions where we've bought technology and people. We've then wrapped that into our product, and it's enabled us to increase the value of our product." That's mostly focused around storage, on-prem, and in the cloud, and the services that surround it. So that's where the focus is now. We still have a great team.
There's some interesting stuff out there in that kind of tech and talent, but that's more like $20-$30 million type of deal versus $200-$300 million. Where you're buying a big business, you buy 300 people and then all the fun that comes with that.
Okay. All right. Well, congratulations on all the success over the last year. And look forward to hearing more at the June Analyst Day. But thanks so much for being here, Mike.
Thank you for having us. Appreciate it. Thank you.