Fortinet, Inc. (FTNT) Earnings Call Transcript & Summary

Okay, everyone, thank you very much for joining us today. My name is Andy Nowinski, software analyst at Wells Fargo. And it's my pleasure to introduce you to Keith Jensen, CFO of Fortinet. I think we have a full room, and we're webcasting this session. So I know Keith has a safe harbor statement he'd like to read first before we move on.

Looking forward to it. Yes, thank you. I'd like to remind everyone that we may make forward-looking statements during today's fireside chat. These forward-looking statements are subject to risks and uncertainties that could cause actual results to differ materially from those projected in these statements. Please refer to our SEC filings, in particular, the risk factors in our most recent Form 10-K and Form 10-Q and to other reports that we may file from time to time with the SEC for additional information on factors that may cause actual results to differ materially from our current expectations. All forward-looking statements reflect our opinions only as of the date of this presentation, and we undertake no obligation and specifically disclaim any obligation to update forward-looking statements. Thank you very much.

All right. Thanks, Keith. So if anyone -- just before we start, if I anyone has a question that they'd like to get in, just feel free to raise your hand. I'll just repeat it for those that are listening in on the webcast, and we can go from there. So thanks for joining us today.

Thank you for having us.

You're coming up -- I want to talk about maybe some of the high-level question that I've been getting since your last earnings call. A lot of people have kind of compared some of the trends you're seeing now to prior spending cycles, like going back to 2013, where, if you look at back then, and it was maybe like one quarter before the big Target and Home Depot breaches, where it -- your growth dipped and then it rebounded right away. And then you look at another cycle, like back in 2017, where we went through a big firewall refresh cycle, your growth dipped, and it took maybe 4 or 5 quarters for it to rebound. So I think a lot of questions I've been getting since the Q3 earnings call were, where do you feel like we're at with right now? Is it more similar to a 2013, where it was a one-and-done and popped back up? Or is it more like a 2017?

Yes. Everyone wants to know how long...

How long? Yes.

How long is the question. And just for context, I think you can see it in some of our investor slides in the earnings call, you saw a dip, that Andy is referring to, in 2009, '13, '17. And truthfully, we probably would have expected one in 2021 because of the 4-year cycle nature of it, but we didn't get that dip. In fact, we got quite the opposite. With the supply chain and interest rates being very, very low, we saw outpaced growth. And I think that probably is important to keep in mind because maybe what we're saying right now is that every cycle is a little bit different. We had the fast recovery that we referred to in one case and the longer-term recovery in another. This has certainly shown to be coming off of a higher peak and probably moved a little bit quicker in Q2 and Q3 than I would have expected it to when you look at things. I think the other comment I would offer is in terms of a starting point, really, I think Q4 of 2022 was probably the apex in terms of bookings for us. And billings was probably more around Q1 of 2023. So if you're looking at a starting frame, I think around the beginning of the year is probably the right place to kind of start the clock in terms of where you're seeing the shift in the market.

And you announced last quarter a pretty big shift in your own business, really pivoting towards the SASE market and which I -- it is certainly a high priority and maybe the highest priority in a security budget, and it's really not much of a shift either. It's very complementary to your network security business, the network technology. But I'd love to just kind of know, what change -- like why did you wait until Q3 to sort of announce the pivot to SASE versus Q2? And you noted like bookings started to slow in Q1 of 2022 and billings in Q1 of 2023, and then you announced the shift to SASE a few quarters after that. So maybe just what triggered the -- what was the...

Tell you about the life journey of SASE for us, and we got there. I think it's a great question, and certainly, a common one I've heard today. Yes, 2 things. One, I would say, to start it off, one, we're a company that doesn't announce products until they're ready and until they're released. We don't tend to announce. That's just our philosophy. We don't necessarily announce the vision as much as we do the products. Secondly, I think the journey for us actually started with the OPAQ acquisition, a very small acquisition, 2 or 3 years ago, a $9 million acquisition. But it was where you could see that we had an interest in SASE. And I think that's a good starting point. We brought that product to market and pretty quickly realized that it was an exception of our build-versus-buy philosophy, where we want to put everything in our own operating system and run products that way. In this instance or in that instance, we brought somebody else's operating system to market, realized it wasn't the right fit for us. Pulled it from the market and spent the next 2 years really quietly building out SASE features and functions. If you fast forward a little bit, I think if you look at the -- more about the journey, you've heard us first talk about a year ago about the strategy of delivering SASE through the service provider network. We have historically been very faithful to the service providers. We usually give them the first opportunity with new products, as we did with SD-WAN. And I think what we saw in the SASE space was the service providers need to go through their own engineering rework and design, if you will. When you give them a product, it was taking too long to bring to market. And so the next thing in the middle of the year you really heard us talk about was, maybe we should actually be buying more of our own PoPs and data centers and bring them to market that way. There's certain cost advantages to those 2 structures, but there's a time disadvantage to both of those structures. I think the next real milestone, and you heard us talking about this, the evolution of the strategy about how we are going to deliver it, was we had a management team meeting in early July, part of our sales -- President's Club annual event. And all of our managers were there, and we were talking to sales about what do you need as we go forward? What do you see in 2024 and beyond? And they made a very compelling presentation that SASE was something that they wanted us to bring to market and complete the suite of products as we're -- and bring it to market quickly. We had a pretty good idea then that the Gartner Magic Quadrant for Single-Vendor SASE would be out in the near future and that we were going to do well in that particular market. And that's -- in that Magic Quadrant. And I think that when you have that, a third party providing affirmation of the product, then we get comfortable with our own product announcements. And we have a customer or a sales group that is asking for the product. So really, the decision point in that meeting was getting to market faster. And we've opted for the Google PoP strategy. We've certainly talked to other cloud providers, and we will continue to talk to other cloud providers. But we did that because one of the buying objection we heard most frequently from customers through sales was the number of PoPs. There was this litmus test of how many PoPs do you have? And we would probably announce 20. And they would say, until you have something like 100, you're probably not really ready. And so they solved that problem very, very quickly. So I think what you saw was a long series of events that were destined to be talked about as part of the Q3 earnings announcement. Unfortunately, it was also a quarter in which the firewall underperformed expectations.

That makes sense, just a coincidental timing of the 2 coming together.

Well, I think it's such a -- it's a great point, but I think it's such a shift in conversation for us. We still very much believe in the firewall market as we talked at your first question, the cyclical nature that we see in the market, and absolutely confirm when we talk to customers that it's a hybrid world, whether that means on-prem solutions, cloud, SaaS, SASE solutions, CIOs in SASE and CISOs are very much committed to this hybrid world of solutions. Firewalls will still be a large part of it.

Okay. I want to continue talking on SASE here. We learned a little bit more this morning in some prior meetings about when we discuss the underlying architecture of SASE, whether it's using Google. Like you guys announced and brought your solution basically up to speed overnight with that announcement and leveraging. I think Google has about 187 PoPs, or that's what they refer to them as. But I guess maybe the downside of leveraging Google would be that Google potentially prioritizes their own traffic over your traffic. Do you see that as a risk? And I know Fortinet still has a propensity to own a lot of your own data centers and footprint, which takes time to build out. But just curious how you think about the trade-offs of using a Google versus building your own?

Great. I mean I think there's trade-offs both in terms of my own income statement, my own balance sheet, in terms of how I deploy capital. So that would be a consideration. But you also made reference to the point that while we started down the path of acquiring data centers and building them out and also, more importantly, building out PoPs, we haven't stopped that. We will continue to do that. So I would expect that in order to mitigate not only prioritization of traffic concerns that could evolve, I've not seen them to this point, but also locations. It's not that Google has a PoP in every location and every place. And so we'll still want that flexibility. Three, we think that a source of price advantage over the much longer term, right, not this quarter, not 2024, is owning some of your own locations, I think, actually helps you in terms of managing your cost structure and, in turn, managing your price points.

Makes sense. I -- just out of curiosity, so Fortinet, obviously, has been known for a long time for your custom ASIC and your top -- and your performance that you leverage from owning that piece of the hardware technology. Are you able to put your own firewalls into the Google data centers to continue leveraging the value proposition of your ASIC?

Yes. I think, again I'll go back a little bit further. I think there's been an evolution of this conversation with Fortinet about the ASIC, and then we'll come back to how it plays. Very early in my career in Fortinet, it was very much about speeds and feeds, speeds and feeds. And that's certainly very true when you're talking about the ASIC and the performance advantage. And you saw us be very successful where that was key, for example, low latency trading environments and telcos early on. And then you think you saw the evolution from us in understanding the complementary nature of the ASIC and the operating system, where the ASIC enabled more and more capacity for the operating system without a degradation in performance. We've talked in the past about things like SSL encryption and de-encryption. We talked about SD-WAN, right, as being part of the operating system. And now SASE is part of the operating system. I think the new generation now is really, in terms of our conversation, is talking about the operating system and where -- how that is form factor agnostic, whether it's a firewall or whether it's a virtual machine or it's a SaaS offering or what have you. And it's not just the firewall technology, which includes proxy and SD-WAN, but it's all the other security offerings that are part of that operating system as well. And so as we look forward, we want to focus on the operating system as being important to it. To your question in terms of where might you see a firewall in the use case, I don't necessarily expect that I'm going to see Google's cloud data centers with our firewalls. But I do expect that there's going to be instances where they won't have firewalls or will be using our own PoPs or data centers or own locations, and those will obviously be our own firewalls. And also in some of the larger organizations that we've been having conversations with, their preference at this point is with a private SASE offering, meaning it would be on-prem. And there, too, there would be an opportunity for the Fortinet firewalls.

Okay, that makes sense. A number of your competitors will talk about how SASE eliminates the need for a firewall. And it sounds like -- I think you may have answered this question already when you said FortiSASE is part of the operating system that you offer, part of the software platform, so you can deploy it on any form factor. So I guess, are you selling, like if a customer wants to deploy FortiSASE, could it just be a software-only sale that...

Absolutely. Yes, absolutely. Yes.

Do you view that as sort of cannibalizing your own hardware business? Or is it complementary to it?

I think that as you continue -- I think it's complementary and perhaps even expansive in that if you look at the history of cybersecurity, it's one of constant growth in terms of the use cases and the technologies that are being offered. I have the chance to talk to a lot of customers. By a lot, I mean, something in the order of 30 or 40 customers each quarter. I've never encountered a customer that says, "I'm going off firewalls and I'm going to SASE." That's complementary. There are a number of instances that I'm involved with, with customers who are working on proposals and RFPs that are 2 different projects that they're going to both fulfill, one is a SASE offering and one is a firewall offering. The longer term they see both of those applications. I think what it does provide is a higher-margin service revenue opportunity for us to supplement the firewall market.

And what about the cash flow impact of this pivot? Obviously, firewalls have been a very cash-rich business for Fortinet, and everyone values your stock on a free cash flow basis. But does this pivot to SASE and more software-oriented sales have any impact on your sustainability of your free cash flow?

Yes. Great question I think when you look at the SASE solutions, what I've read is that these are typically billed annually in advance. So call them 1-year contracts. If you compare and contrast that to my average contract term that Fortinet probably ranges over the last several years from 25 months to 29 months, they get a little bit larger if you're doing large, say, SD-WAN deployments or something like that. But they've pretty much been in that range, fairly stable at about 27, 27.5 months for the last several quarters. Keep in mind, my business model, I have probably 1/3 of my business, if not more, is on SMB. And generally speaking, the SMB are already on 1-year contract. So there's not much of a difference there. I think at the other end of the spectrum, there are some players in the space that are very much focused on large enterprises, and their average contract duration, not terribly specific about it, but I think we can kind of read between the lines and think that they've probably been 35 to 40 months or something like that. So I would imagine in that environment, you'd probably see more of a shift in terms of the average term. I think at the end of the day, given my customer mix, my product mix, I'm not anticipating that I'm going to have a significant change in my cash flow on my contract duration in 2024. But if I do because of tremendous success from SASE, I'd probably be okay with that.

Yes, a positive, higher-margin business, sustainable. Okay. Yes, that's great. Why don't we -- actually, one last question on your SASE business. So how important is SD-WAN to that? Because you -- there's a lot of vendors that have it, some that don't. Gartner seems to think everyone needs it, it's got to be part of the solution. You have, obviously, that's probably one of your strong seats of your operating system is the -- is SD-WAN. So...

Yes. We were having a conversation with our Board, and it was up at the whiteboard saying SASE is really made up of SSE and SD-WAN. And explaining that, and to have a SASE solution, you need the SD-WAN offering. We've provided some -- I would expect that as we go through 2024, and we've put more wood behind the arrow with SASE in the go-to-market approach, that part of that will be greater success with our existing SD-WAN customers. We probably have -- well, we have tens of thousands of SD-WAN customers, and that to me would be a logical place that I would expect us to see significant traction with SASE solutions. We know the customer. They know us. It's a great opportunity. I was a little bit surprised when I went back and looked at, say, the prior 6 months, and I saw several hundred smaller SASE deals from our sales team. I expected that the majority of those -- vast majority would be SD-WAN customers. And in fact, it was something on the order of, I can't remember, 30% or 35%. So without a lot of direction, maybe we got a little more of a balanced mix from our sales team. I would expect in 2024, when I talk about that pie chart a year from now, it's probably going to be a different percentage. The other surprising part of that or perhaps educational was there was another 30% of that pie chart that were firewall customers for other use cases, right? It could be the data center, it could be micro segmentation. That I get. We have a relationship with the customer. And then that other last 25% or so was pure white space. So we'll see how this plays out. But I think I took all of that as being good indicators of the success that we may have as we go forward.

So 70% of those customers that you sold FortiSASE to were not deploying your SD-WAN yet?

Correct.

That's pretty -- it's surprising. I would have thought of...

It is. But it was also over 45% of them are SMBs, which is kind of where I would expect our natural success is going to be immediately, is in the SMB space, and particularly on the international part of the stage as well. So...

What does the competition look like in SMB? It seems like all the big SD-WAN or SASE vendors are targeting that large enterprises...

Yes, I think you've got 2 guys that are going -- 2 companies that are going toe to toe in the U.S. enterprise space, right, yes.

Yes. It's a pretty captive market that you already got with SMB that you could sell to...

I don' think you need to tell them that.

Okay. Well, that's great. Why don't we talk a little bit about maybe more -- since we're lucky enough to have CFO on stage here, why don't we talk a little bit about the financials. So billings in Q4, based on your guidance, it looks like it was maybe -- might be down on a year-over-year basis. And certainly, a comp is -- a difficult comp is one of the biggest factors there. And you obviously benefited from your backlog over the last 18-plus months. But I guess, what really -- was there any other factors that drove that year-over-year decline in billings that we're expecting here in Q4?

Yes, I think the color on Q4 is a little bit of what we saw in Q2 and Q3. One is service providers have continued to be weak for the second half of this year. I think that's consistent with what you've read from or hear from other suppliers that are in the service provider space, some softness there. Retail is bit soft as well, and those are 2 of our largest verticals. We don't have -- we have a large international government presence, but not a large U.S. government presence. And some of the outperformance you've read about in the industry has been in the U.S. government segment. So on a vertical basis, maybe a bit of a disadvantage as we look at the fourth quarter. I think, obviously, coming off of the results -- reported results in Q2 and Q3, and as Ken, our CEO, pointed out, some macro headwinds there, some digestion of inventory that has to happen, but also some work that needed to be done around improving sales execution and marketing execution and efficiency. I think we've been somewhat transparent about that. We can also put the macro in there. But I want to be careful that we don't take the easy way out and just point to the macro and say there's nothing that we can do better. And so with that, I think there's a fair amount of focus in terms of how we evaluate our supply -- our pipeline. And while that showed some signs of improvement in the third quarter, I think we'd like to see more improvement in the fourth quarter as part of turning the business and the performance around.

And macro certainly is a piece of that as well, as you mentioned. But is there anything you can do on the discounting side that might help offset some of the macro pressures?

Well, I think the -- we have taken certain pricing actions, both in terms of the price list, but also discounting and incentives to the channel to as part of the correction that you've kind of made reference to. It certainly has been apparent, say, for the last 2 quarters, that if you give a customer a choice right now between extended payment terms or discounting, they'll opt for the payment terms. I'm very open in 2023, they're offering either discounting or payment terms. And for whatever reason, payment terms came back with being much more appealing to people. Now maybe that's just because interest rates moved so much so high so quickly that they're really -- that, that -- there's been more pain around that. Now keep in mind, I'm a 2-tier distribution model, right? I'm selling to distributors, to resellers and to users. And so when I provide extended payment terms, it's really to the distributors. And what I'm doing there is I'm enabling their financing function. The distributors will provide financing, but they're using my extended payment terms as a source of capital. And I'm completely fine with subsidizing them that way. That seems to be the thing that has the most interest right now as I look through the channel.

Did you have -- I guess, as you were drawing that, you saw the backlog going down. Did you see that also showing up on -- as your distributor -- in the distributor channel?

No. No, and I say that with conviction because I'm an old accountant auditor, so I'm not going to be stuffing the channel. We pretty much -- we limit distributor inventory levels at 6 weeks of sales, right? So there's -- sometimes it will go over that and will reverse those sales out, so it shouldn't be that way. And I think that the question that, that kind of leads to is inventory digestion. Where is that inventory and who is that inventory with? Did I -- keep in mind, again, the business model, 1/3 of mine is SMB. I really don't think the SMBs are out there putting inventory on their shelves at the end user level. I don't think that happened. We talked about my channel. Did I have some resellers maybe that we're putting that I don't have a lot of visibility to, putting something on their shelves? Probably. I did see it, I think, in enterprises. And an example of that would be, say, an enterprise that has a large SD-WAN deployment that they were kicking off, and they wanted to make sure that they weren't going to run out of materials or components. And that deployment may take 12 or 18 months. So I can point to examples of U.S. retailers that did that. But it's not -- was it a headwind? Yes. But was it the sole cause of what we experienced? No, I don't think so.

Okay. I want to maybe shift gears to the services revenue piece of your business. That's always been very -- had really sustainable growth in that 30%-plus range. And a lot of it is tied to the maintenance contracts that go with a firewall. So it does give you some pretty sustainable business, but there's a lag, I think. Once you see product revenue start to decelerate, service revenue eventually will lower as well too, but it's not right away. It's typically at least 12-plus months. But in your guidance for Q4, and it shows a pretty sharp drop in the services revenue growth, too, is like what else -- what are we maybe misunderstanding on the services side? Why did that drop off so quickly, too?

Yes, it's a great point. And I would -- keep in mind that we offer 2 traditional service offerings. One is our traditional support, 24x7 call support that everybody else does called FortiCare. And that's the one that we've seen the slowdown in. FortiGuard, which is the sale of additional security subscriptions or a bundle of security subscriptions that are sold at the time of sale as I think the growth was still over 30% or 31%, if I'm not mistaken. So I really think that what you're seeing there is the support related to the product billing. And given that we saw the deceleration in product billings come into play starting in, I think, the first quarter of this year, maybe the fourth quarter of last year, there is, as you point out, there's that lagging effect, right? Billings and product revenue are good leading indicators of the business. Service revenue is much more of a lagging indicator. So you are seeing that. And as you look at 2024, we encourage people that are doing modeling, whether it's analysts or others, to look at short-term deferred revenue and that growth, first and foremost, right, to give you a sense of what we would expect in the next 12 months. But also look at that trend as you're doing your modeling as well and making sure you're capturing the trend as part of your own modeling.

Right. So we -- so it is fairly accurate, about a 12-month lag on the FortiCare piece of the business?

Yes, I don't -- that 12 months sounds like an awfully precise number, but...

Yes, roughly.

Yes. There you go. Okay.

And then the FortiGuard side, it is still really strong. And what are the kind of the key drivers of the FortiGuard business that's keeping that growth so nice?

Yes. So one of the things that kind of ties into the big question we get about, did you change comp plans in the third quarter as part of the SASE offering? And I would say that, no, we haven't. But it does lead to -- what we did do is, earlier in the year, maybe at the very beginning of the year, is for a subset of our salespeople, we view them as what I would call subject matter experts, and we gave them a second quota for the SecOps products, [ SIEM ], SOAR, EDR and things like that. And that seems to have provided a positive result. And you see the result, is when we sell those SaaS offerings, that revenue then appears in the FortiGuard line. So I think what you're seeing is the incentives that we offer to that group of people, we will then keep offering those incentives obviously in 2024 to them as well.

Do you think the recent acquisition of Splunk, which would compete against your [ SIEM ] and some of your SecOps offerings, I guess, does that have a tailwind? But I know your SMB customers are probably not buying Splunk, but your non-SMB customers are probably -- that was one of the products they were potentially using.

Yes. I mean when there's -- sometimes, when there's consolidation in the industry, VMware, for example, Broadcom, it's disruptive, and there can be some fallout from that as well. And it's also, I think, a question of the acquiring company, is what are they acquiring? Are they acquiring the technology? Are they acquiring the customer base? And how is -- as I think there's -- in one case, I think it's pretty well established that there'll be some separation of some customers post-acquisition. So it certainly does create some opportunities, yes.

It's not factored in anything yet, but that's something that could be potentially...

Yes. I would not get that detailed in, yes.

Yes. In an earlier discussion today, we started talking a lot about the SEC regulation that's going into effect. That also plays into that -- I think the [ SIEM ] market and SecOps space, where you have to disclose a material breach within 4 days. Do you feel like that could be a potential tailwind for FortiGuard?

Yes, absolutely. I think the role of people you aren't familiar with, within 4 days of knowing you have a material breach or believing you have a material breach, so I don't know if that's from -- don't read that to mean 4 days from when you discover a breach, you've got to go through that decision point. Nonetheless, people are going to find that to be a very stressful time, one, in terms of -- it always is when you have some sort of a security event, whether it's real or it's a false positive, that investigation activity of gathering logs, as you're alluding to, and other sources of evidence and trying to understand whether or not somebody is actually in your systems. And if so, where they're at in your system is a high-energy exercise, put it that way. To the extent that you have more automation, more consolidation in your network, that's going to be easier than it would otherwise have been. I can tell you, from experience, even investigating false positives, when you start looking at all the disparate technologies that exist inside of a cybersecurity landscape in a very, again, very tense environment where you're trying to understand something, anytime you have some sort of consolidation where you have information that's moving between the same solutions of a security adviser -- provider, it's easier than it would otherwise be. Not easy, but easier.

Maybe just to wrap up our discussion on FortiGuard and SASE. When you sell your SASE solution, would 100% of that revenue be recognized in the FortiGuard line? Or would there be components that are in product revenue, too?

Yes. Well, there could be. The majority of it, I would expect, is going to be in the FortiGuard line. You could conceivably get some, depending upon the other things that may be ordered at the same time, you may get something in the FortiCare line, and you can also if it's a private SASE offering with on-prem, you're going to see things on the product as well.

Okay, yes. If they bought some on-premise firewall solution, okay. Maybe shifting gears then on to the competitive landscape, everyone talks generically about SASE and that's where the market is moving and that's the highest priority. But I think it sounds like the way you're looking at like some of the main competitors like Palo Alto and Zscaler that are targeting the large enterprise, you wouldn't necessarily compete against because you're maybe catering to the lower end of the market or in the mid-market. But do you think this does change the competitive dynamic for Fortinet and kind of puts you more into competition with some of those vendors that you haven't seen in the past?

Yes, I want to be. I want to be invited to that contest, absolutely. I think we're fortunate with some of our existing large enterprise customers, both in the U.S. and Europe, where we've spoken to them about our SASE strategy. And with understanding that we haven't -- have not been selling it primarily in the market for 5 years as other people have. We needed the benefit of their insights and perhaps -- and they've agreed to some co-development projects with us in terms of making sure that we understand the feature set and the functionality that's needed as we have those opportunities go toe to toe in those large enterprises.

Well, I've always been curious, like why is Fortinet maybe had less market share in the large enterprise space, but you dominate the even more ultra-demanding service provider market that even has a more complex environment than a large enterprise and higher traffic throughput? Why couldn't you leverage that leadership position into the larger enterprises...

You're spending a lot of time talking to my CEO, haven't you?

We play golf a little.

Yes, there you go. That must be it, yes. Look, I think that where Ken points out is when it's a technical sale. And by that, he means when the customer is making a very technical evaluation, as telcos do and as big banks do, that the Fortinet product does very, very well. When it becomes more about marketing and schmoozing and things of that nature, that really hasn't been the sweet spot of the company, and particularly as you got into the large mid-enterprise space, a little more difficult. And I think that when you look at the geographic differences, actually, it's a different story. What you described is very much prevalent in the U.S. But when you look at it internationally, you oftentimes will find that Fortinet's the #1 firewall vendor in many, many countries, in Europe, in Latin America and in parts of Asia. And I think there, what you may be seeing is how the channel interacts with the customer differently, that the channels internationally probably had the ear of the customer a little bit more so than some of the channel partners, by channel here, I mean, distributors, particularly, than they do in the U.S. So you get things that are successful internationally, maybe not so much more in the U.S. Things in the U.S., a little bit less successful internationally.

All right. Well, we're running out of time here, but I would love to maybe just spend the last 2 minutes here talking about next year and kind of how do you feel like -- or what message do you want to leave with investors as we think about next year and how well positioned Fortinet is entering some new markets with SASE or maybe a model that's been reset?

Yes. I'm happy that we've spent, say, for the last year, in the earnings calls, we've talked about the impact of cloud provider fees and data centers on our services gross margin, not because service gross margin was suffering, but just to kind of preview for the audience that Fortinet was moving more to a software -- or adding more software solutions to the market. Now to see those things really become something that we talk about in every meeting, those software solutions, those opportunities that supplement the firewall market that again is not going away, I'm actually very excited about that to see how those 2 things work together as we move into 2024. And the recognition that I'm hoping that we're going to get with our offering as more and more people become exposed to it.

Recognition of -- by third parties like...

Customers.

[ Customers' loyalty ]?

Yes.

Okay.

I enjoy going to customer meetings when they get to watch the presentation from our product marketing people about, one, you have all these products, but two, actually a little bit more mature than people think it is.

And do you think it's the top priority next year? Is that -- that seems to be resonating with a lot of -- or message we've heard so far, either Zero Trust as the -- being one of the top spending priorities next year?

I'm probably not the best one to ask that question, to be honest with you. I think that my temptation is to respond back to -- I think that the AI journey and ChatGPT and the LLMs, I think that's going to get very interesting very, very quickly. It's not uncommon to see that a new technology that comes forward in our space, you sometimes see it being used by the threat actors, which then causes companies to respond and the industry to respond. It's not necessarily that the industry is the first one out there with a solution against something because you really don't know what the attack vector is going to be. So maybe for no other reason other than just curiosity about how that plays out now in 2024.

Well, we'll watch for Fortinet to become more of a software-oriented vendor and a SASE...

Still very faithful to firewalls.

Still faithful to firewall. All right. Well, thank you very much for your time then, Keith.

Okay. Thank you.

We appreciate it.

All right.