Fortinet Inc (FTNT) 2017 Q4 法說會逐字稿

Good day, ladies and gentlemen, and welcome to the Fortinet Fourth Quarter 2017 Earnings Financial Analyst Q&A Call. (Operator Instructions) As a reminder, today's program is being recorded. And now, I'd like to introduce your host for today's program, Peter Salkowski, Vice President of Investor Relations. Please go ahead.

Thank you, Jonathan. Thank you, everybody, for dialing back in. I'd like to remind everyone that the disclosures and cautionary language that I read at the beginning of the 1:30 call -- 1:30 p.m. call today still applies on this call. We will be making forward-looking statements, and all of those are subject to risks and uncertainties, as outlined in our presentation online. This call is primarily a Q&A call, so I'll turn the call back over to the operator to moderate the Q&A session. Jonathan, you can begin.

(Operator Instructions) Our first question comes from the line of Michael Turits from Raymond James.

A couple of questions. One, you mentioned the large CRM deal but were there other deals -- were there other large deals -- I think somebody asked some things. I think there was an expectation that there was a lot in the pipeline from a large deal perspective, but is there anything that's still pending there?

Michael, it seems like you're asking if anything pushed. I think, we did a good job of closing the business we expected to close, as evidenced by the overperformance.

I'm sorry, you said like you feel like you closed what you were looking to close?

Yes. I think so. I mean, we clearly beat guidance, so we felt pretty good about what we did. I think, I mentioned earlier, when we looked at the flow of the business, we had a nice -- one of the things we pointed out, I think, we had the highest average deal size we've seen and it was pretty consistent throughout the quarter. So we saw a nice drop in DSO for instance. I'm just going at -- we saw what I would characterize as a more high-velocity quarter, with a nice flow of higher average deal sizes throughout the quarter, and that's how I would characterize the quarter.

If I -- you obviously beat expectations and -- but if I look at the sequential growth, it still seems like the seasonality was somewhat muted for a fourth quarter. Is that a change in the seasonality overall? Or how should I read that?

From Q3 to Q4?

Right.

Q3 of '16 was a very strong quarter, so I don't know if that's coming into your comparison analysis.

Q3 -- I think, if you look last year, Q3 of '16 was a miss, and so that was a tough quarter. And then, coming off of that Q4 was a little higher. But again, from our perspective, Michael, we feel like we did a good job of closing what we thought was in front of us.

Okay. And then my last one, and I'll turn it over, will just be on cash flow. As I look to next year, it was very helpful that you gave us the cash taxes into next year. It also seems like this was a particularly strong year for working capital. I know there's a deferred tax asset, but outside of that, there were other working capital issues. So should we think -- should we keep our expectations relatively muted into the growth next year, given the strength of the working capital this year?

We don't guide on free cash flow, obviously, right? We can -- we give you the tax number, we can certainly give the CapEx number if that's helpful to you. If I'm going to compare year-over-year CapEx for real estate related, about $105 million in 2017, probably $60 million to $80 million in 2018. And then, you would add to those both of those years and normalize run rate of other things, that's probably in the $30 million range.

Our next question comes from the line of Erik Suppiger from JMP.

First off, just want to make sure on the CapEx, so you're suggesting that total CapEx this year would be somewhere in the $100 million to $120 million -- or I guess $90 million to $110 million -- $120 million, is that right?

Yes.

Okay. On the deferred tax asset, what would be a normalized number there? If you did $58 million -- you had a $58 million benefit, how should we think about that in the coming year?

I think, we'll come back to you on that. I'm not sure exactly -- can you try the question differently?

It's really -- that's an accounting -- Erik, that would have just been accounting benefit. We can follow up with you and get a better answer for you, though.

All right. I think, the non-FortiGate business was 25% for this quarter, for the December quarter and the September quarter. Can you give us the percentage that, that was for some of the prior quarters, and in particular, I'd be curious about the year ago quarter, do you have that?

Well, it's growing faster than -- I don't think we've shared that, Erik, but it has grown over the last years and quarter, it's been a steady ramp up. It's growing faster than the overall growth rate of the company.

Can you give us some ballpark? Because I remember going back a while ago, it was less than 10%. Can you give us some context around when it moved up from single digits to double digits?

I think, you may be -- Erik, I think, you may be referring to a slice of it. I think, it depends on the composition. Again, maybe that's something we can follow up on to give you better clarification, but the way I would look at it, it's been a nice, steady ramp-up, growing faster than the overall business consistently for the last couple of years. And then, we've seen it become more meaningful, really, over the last 18 months, if you will.

Okay. And then, lastly, the FortiGate 6000, is that something that would take some time to ramp because it's such a high-end product? Or when do you think that could be -- is that something that could make a meaningful difference in even the first half of '18? Or how's the timing on that?

It's a product way ahead of our competition -- competitors. There's a -- the enterprise takes a little bit long time, but on the other side, we do start in beta right now. We probably could warrant an interest on the field. We see this as very -- one of the best product. Like we say -- I say, it's the farthest one in our appliance base. And it's a very good product for the customer. It's difficult to say how big impact it will create but I do see the interest is already strong on this product.

Do you see that as equal between service provider and enterprise? Or is that more of an enterprise product?

It's more enterprise. It's more enterprise. The service provider tend to go to the chassis base, which is a 7000 series.

Erik, it's Drew again. Just coming back on the -- you had the deferred tax asset question. I just wanted to make sure you understood, that's not a cash number. That's just an accounting number. So that's just the new tax law. And I think, you've seen a lot of companies writing that down in their Q4. It has been a trend. And so that's just basically the implementation of that. But...

Our next question comes from the line of Catharine Trebnick from Dougherty.

Just to clarify on Erik, was that 25% of billings or revenue for the non-FortiGate products?

It's billings, Catharine.

Okay. Billings. Good. Because I didn't want -- all right. And just to clarify, on the new box that you released today, new platform, it's targeted for high-end enterprise and carriers?

Yes. More for the enterprise but also some carrier provider, also see it's a pretty good solution for them. But it's mainly for the big enterprise.

Okay. And then, have you been seeing -- I know you said that you had the largest area CRM build this quarter, and I hate to beat this drum on this, but were there other large data center builds also that you have seen a like stair-step climb in winning these and then the non-FortiGate products are helping you expand beyond the data center. Are you seeing that trend?

Definitely, we see the fabric promotion starting to make the deal larger. It's including some other part of cybersecurity. That's -- I think CRM is one example of that is really a fiber approach. And that also makes solution long term more sticky with the customer because it's helping them to lower the management cost at the same time. It's a multiple product solution working together is -- help them lower the management cost also.

Yes. And Catharine, what we track internally is the number of fabric deals and the number of fabric customers, and that continues to be up into the right. And so that's definitely a positive trend.

Okay. And Drew, congratulations. We'll miss working with you, and best of luck.

Thank you, Catharine.

Our next question comes from the line of Gregg Moskowitz from Cowen and Company.

Drew, congratulations as well. So this is a good quarter, certainly, although I did want to ask about your Billings in Americas, Drew, because it did decelerate a fair amount this quarter. And I think the compare was only slightly more difficult than it was in the Q3. Just wondering if there was anything that you could call out in relation to that?

Well, it was a bit of a tough compare. I think, we grew -- I'm going back, I believe the number was 20% last Q4. So yes, we did see it as a tough compare. We had -- we really -- what I mentioned earlier, Gregg, was there were some concerns we called out, caveats we called out when we guided, and those actually didn't occur in a negative way. We actually were able to close most of that business. So we took it as, really, a very good quarter. I do think we had a couple of challenges in the U.S., to be fair as well. But aside from that, it's a pretty good quarter. Again, I would go back to what I said to earlier that it was a nice flow of business throughout the quarter. Had our highest average deal size ever. It was just more characteristic of that type of a quarter.

Okay. Great. And then, a question, I guess, either for Ken or for Drew, your carrier business was about 20% of billings in 2017. How did that compare with your original expectations, if you will? And how are you thinking about 2018 with respect to that vertical?

I'd say that 2017 international carrier is starting to catch up a little bit. The U.S. -- the few U.S. majors carriers still a little bit behind. But I hope they are starting to turn around in 2018, following the national (sic) [international] trend.

Okay. Perfect. One last, just clarification for Ken if I could. Did you say earlier in answer to a question that you don't foresee a slowdown in product growth going forward in spite of the fact that longer duration deals do shift more revenue from product to services? I just want to make sure I understood that part of it.

I believe what I said is really, there's a few percentage come from -- I mean, go to long-term deferral, or whatever, service revenue because the service periods are longer compared to 1 year ago from 22 months to 26 months. I cannot give anything going forward how product revenue may grow, but I do believe we have pretty strong product, including the new one we just announced today. And also, like I said last time, this year, we probably have some other new FortiASIC [SPO] come up, which also will help in drive the performance, lower the cost on the product side. That's also helping on the displacement of some other refreshment opportunities. But is -- I cannot really comment on the future product growth. But it's -- we have a pretty strong offering right now.

Gregg, this is Keith. I would just supplement that by offering just pure unit shipments, we were very pleased with the growth in unit shipments, both quarter-to-quarter sequentially and year-over-year. So kind of regardless of where that ends up in the income statement or on the balance sheet, we're very pleased with the product shipping out.

Our next question comes from the line of Fatima Boolani from UBS.

Drew, best of luck to you on your new adventure. Keith, maybe a question for you on the free cash flow. I just wanted to drill into that a little bit on how the year-over-year increase in margins and some of the tax impact and the higher cash taxes is going to flow through into your cash flow and just a question I'm getting from investors. So I just wanted to make sure I was very clear on that around your cash conversion given that all of these dynamics.

Well, we certainly had a pretty good free cash flow year in 2017. I think, the team following what Drew was asking for, did a very good job in executing on the conversion cycle, whether it was in DSO, whether it was on VTO and managing the inventory. Obviously, the turns were very a good in Q4, and I think the forecasting in Q4 obviously pushed that along very, very dramatically. We provided the cash out for taxes in the year, $40 million to $44 million, up a little bit from the prior year for the CapEx number. I think the opportunity for us as we continue to manage the capital cycle throughout 2018 and looking for some additional areas to continue to create opportunities for us.

Okay. That's fair. And just on the OpEx cadence through the year, it seems like it's very front half-loaded. Is that the right way to think about it? And how should we think about the leverage kind of coming into the second half?

Well, kind of in reverse order, I think, obviously, as the recent sales hires that we bring online in Q4 and Q1, giving them their ramp time, we would expect to see greater productivity out of that group of people in Q3 and Q4 on the top line. On the cost side, obviously, you're going to have that cost coming in from the people that we hired during Q4 and all of Q1, and then the incremental hires coming in, in Q1 as well, moving out throughout the rest of the year. It's a little bit unknown what kind of attrition we might have with that group. Year-end, it's kind of a time for a lot of movement but we'll continue to watch that.

And then, last one for me, if I could. Just in terms of the model shifting to more ratable and more visible, and even some of Drew's comments and your comments around the very consistent transaction velocity you saw, can we expect that you'll, going forward, see just better linearity -- quarter-to-quarter linearity in the business because more of the revenue is coming off the balance sheet? Is that a fair characterization?

From a pure GAAP revenue viewpoint, obviously, that build up in deferred revenue certainly lends itself to a high degree of predictability going into each quarter, right? When we look at it, we look at really 3 areas of GAAP revenue. One is coming off the balance sheet from deferred revenue. The other is the run rate from the SMB business. And then, we get the more challenging aspect of it, of course, is the enterprise. Enterprise deals often come towards end of the quarter, it's less predictable, et cetera.

Our next question comes from the line of Rob Owens from KeyBanc Capital Markets.

Could you characterize your business, either through the unit shipments that you just mentioned you were very happy with, or billings and/or revenue? Just what's coming from new versus what's coming from renewal? What's your customer acquisition look like versus kind of renewing the installed base at this point? And even if you can't point to a direct number, maybe how that's been trending over the last few quarters?

I think, if you look at our product example using FortiGate 60 model, which has been the #1 best-selling globally by any account, we tend to price the same, even new generation like from the very beginning, FortiGate 60, 60A, B, C, D. Now it's the E version, it tends to be priced the same on the hardware. But we do offer some additional service bundle. Early days, it's more on the UTM bundle. And then, we announced we call the enterprise bundle, which had additional service, and also, it's a little bit higher priced service bundle. So that's probably when some customer renew, we do see the trend, they more go through the -- like an enterprise bundle but also some kind of service. We do have 8x5 support, but now trend might go 24/7. That's also making the service percentage a little bit -- I mean the dollar-wise, whatever, the way they account for revenue more and more into the service part. So that's maybe -- like I said, maybe have a little bit few percent impact on the product revenue. For the refreshment like I said, every 4, 5 years, the customer tends to refresh. But also, we see the opportunity, especially on the fabric. So as up-sell, cross-sell. And the same time, the new product, like the 6000, that's the new series, is a huge advantage, especially addressed for the enterprise to try to deal with cloud access and also the internal segmentation. So this is the product really target some like new ground and also -- is also about changing of the landscape of the space, which more go to the internal segmentation instead of just the border protection. And also, even within the data center, security inside the data center and also secure segment inside the enterprise.

And then, Rob, Drew, just a couple other ways to think about it is, we just step back and look at the billings of revenue. Again, we're taking share, clearly, when you look at the rate of growth of the market. And as we mentioned, the units are, I think, a great indicator of new business. And it's true, some of that is probably going to our installed base. And -- but again, they're growing -- outpacing the revenue -- the product revenue growth, certainly. And the other fact that we're seeing is just the broader fabric resonating. And so as you kind of create these layers and levers to the revenue that I've talked about before, which is kind of a combination of customers and units getting out there. And then, creating broader economics, opportunity through upselling higher-value subscriptions and services, the enterprise bundle is taking off very nicely. We continue to do that. We're seeing a higher average deal size. All those things tend to indicate a rising tide of opportunity, if you will, and our success at making that happen.

And you provided me a good segue relative to the non-FortiGate success that you're seeing there. And just can you maybe characterize what the typical customer looks like on that front? Is that more around your entry-level, distributed enterprise SMB? Or are you actually seeing that push into the enterprise customers?

Yes, both, actually. But also, the enterprise is more concerned about how to lower the management cost because they tend to have like, easily, 20 to 30 different product solutions from different vendors. So the fabric help them to lower the number of vendors they need to manage and also making all the product working together, so that's where we design products from the very beginning and so really helping customers to address the management cost. At the same time, it's making it more secure because you can all have different parts working together to create a security fabric.

Our next question comes from the line of Anne Meisner from Susquehanna Financial.

Drew, sorry to see you leave but congrats on the new role.

Thank you, Anne.

So just quick question on the sales organization. We're hearing about some modifications to the buyer program that you've been doing, and I'm sure there's always tweaks happening there, but I also understand you made some changes in sales management. Can you discuss any broader sort of strategy shifts you're making on the go-to-market side as you continue to move upmarket to target the larger enterprise deals?

Yes. This is Ken. I think, we are keeping our hands in the sales regulation. Right now, like Mike Bossert, he's with the company for 3 years. He's managed the merchant sales. And -- but also, before he joined Fortinet about 3 years ago, he is the VP of Sales for FishNet, one of the largest system integrator in the security space in U.S. So we don't see much changing. It's really keeping our hands in, make the team stronger. And also the same thing for channel program, we want to make sure our channel partner and also all the distributor also can benefit from the strong product we have, and also kind of put additional investment into the marketing. Maybe some of you already see, whether on the airport or some other things, you see the Fortinet brand starting appearing there. But it's really the investment we made to help the channel, help the partner grow business together.

Okay. Great. And a quick follow-up for Drew. I know you don't guide deferred revenue, but in general, you continue to move upmarket and see those sort of longer-deal durations, would you expect to see the long-term deferred continue to grow much faster than short term? Because the mix definitely kind of came in this quarter a little bit differently than we had modeled it, and from what I can tell, what the Street had modeled as well, so I just want to make sure we have that right going into '18.

This is Keith, and I would probably just respond again, it's really driven by enterprise mix shift and deal term or deal length. So to the extent you see a similar mix of deal terms in Q1 as you do in Q4, you're going to get similar results and that will continue on through the year. To the extent that we're more successful in the enterprise, we expect those customers to continue to push us out of 1-year and 2-year contracts into 3-, 4- and 5-year contracts, and that obviously is going to push the long-term deferred revenue.

And are those longer deals, are they -- the deal terms, do they tend to be more customer-driven based on sort of, I guess, coming out of CapEx budgets and wanting to actually drive sort of the longer-term subscription and maintenance deals because of that? Or is it partly sales-driven?

Look, I think, the traditional enterprise customer wants -- the answer to that question is yes. They want a longer-term contract. They more likely are going to renew with us. They like to use that for negotiating leverage obviously as they go through the process with us. Are there other opportunities for salespeople there as well? I'm sure there are.

Yes. Also, it takes them like 3 to 5 year to refresh the hardware. Sometimes, it's just, okay, let's buy a server [is the hope here]. And that's a little different than the service providers. Service providers in the past tend to be like based on how much customer they kind of have every year. But when you go to the big enterprise, they're already kind of keeping the hardware for a few years before they do the refresh.

Our next question comes from the line of Sterling Auty from JPMorgan.

Drew, Ken, hate to put you on the spot but we all want to try to better understand the change. Drew, congratulations on the new role, but what kind of motivated the change from either side to move on to this next opportunity?

You're asking me -- well, look, I really probably shouldn't be discussing my next opportunity on the call. It was clearly something I decided to do. I would just offer that I do have some family on the East Coast. It's a great opportunity by the way. I'm looking at a great opportunity, I should start with that. And I -- but I think we should probably leave it there.

Yes. Drew is doing a great job. But also the same team -- including building a very strong team, and that's also -- at Fortinet, we also have a pretty good strong team and also a succession plan in all -- pretty much all function areas. So that's where we feel -- yes, I think it's -- we will keep on building the team and make it stronger.

No, that's fair. Just everybody wanted to know and since it's already been announced where you're going as to your -- this is the best way to have one answer that everybody can talk to. But now back to the...

Yes, great opportunity, it's a big opportunity, and it works really well for me personally.

No, it's good, that's good. I do want to revisit 2 items we talked a lot about. The first is the free cash flow. I think where a lot of us is struggling is we just want to make sure that we're using the right baseline. So if we look at slide, I think #10, and add up the adjusted free cash flow, it's $566 million, is that the baseline that we should think about as your free cash flow ex real estate that you should be growing from here? Because if we think about next year, margin expansion, we would expect that your free cash flow growth should be better than what your actual billings growth would be if your margins are expanding. Or I think the questions around working capital, I think there was a jump in accounts payable, was there something that maybe gave a little bit of a boost in the fourth quarter that we [shouldn't] be extrapolating for 2018?

Yes, I think the boost in the fourth quarter did indeed come from accounts payable you pointed out. And I think I also mentioned that the inventory numbers were down, which is obviously related to accounts payable as well to some extent. So I do think you saw us execute extremely well on forecasting our inventory needs for the fourth quarter, and that's reflected in the turns. That's an abnormally high turn for us recently. That's going to the payables number.

But again, is that something we extrapolate going forward so we grow off of that $566 million?

Yes. I'm not going to go too much deeper I think than what I just commented on other than those 2 components.

Okay. And then the other question is around the margins. You've given us the first quarter margin guide, the full year. But in just kind of doing some quick work here between the calls, it would seem like, sequentially, some of the expenses might actually have to decline to be able to hit the new numbers. Am I thinking about that right? Or what should we think about as the seasonality or pattern of the expenses throughout the year to hit the full year margin guide?

I don't know what specific expense categories is going to drop during the year, but you would certainly expect us to manage it during the year as we go through our -- as we go to hit our targets.

Yes. Also, like, to invest in growth take a little bit of long time compared to control the cost. That's also we would have been proving multiple times if we want to reach sort of margins, it can be sooner and quicker compared to investing to growth to see some results. But we do believe, long term, the company will be better positioned to keep on growing, at the same time, balance the amount of growth and profitability. So we'll have a target for both instead of just one of it. We also can discuss some detail in the Analyst Day, and we also want to collect feedback from the shareholders, from the partners, from the customer on what would be the better balance among the 2, of growth and profitability. So we'll keep on improving ourselves.

Our next question comes from the line of Hendi Susanto from Gabelli & Company.

My first one, some of your competitors are still working on its sales force productivity and reorg. How does that affect your competition and what you are seeing in the North American market? Did you see some benefit of that in Q4?

I think we do keeping improving ourself also but, at the same time, we have much better products, more broad offerings compared to competitors. I think we're a little bit behind on our hiring in last year, starting to catch up in Q4. Like what Keith said, we hope we'll be contributing to the growth later this year. I don't see the competitive landscape has much changed. We keep on gaining market share. And also if you look on the billing in the last few quarters, last 4 quarters, we're probably the fastest-growing on the billing compared to all the major competitors we're facing right now. So we keep on gaining more market share compared to other competitors. So that's where we do feel we're pretty strong, pretty good about the potential going forward, especially the new offering, both on the network security side and also the fabric, which also give us additional growth opportunity. I think none of our competitors has such a broad offering as we are. And it's really more about how to train the team, how to train the partner to sell the additional solution together with network security.

Got it. And one more question, Fortinet did a significant share buyback in Q4. How should you think about your share buyback approaching 2018?

I think we said going into Q4 that we expect to be aggressive on the buyback during Q4. We entered the year -- or closed last year with $443 million available on the $1 billion authorization from the board. I just teed that up to note the comment that I would expect us to remain aggressive in the Q1 on the buyback.

Our next question comes from the line of Gabriela Borges from Goldman Sachs.

My follow-up is on bundling and pricing. We talked a little bit about earlier about, as your customers come up for refresh, you can put the enterprise funnel in front of them that has higher pricing, a little bit more functionality. I'm curious if you think, over time, is there a scenario where you can roll out more bundles like that, whether it's enterprise pro or enterprise plus additional functionality, that either goes to market formally with your non-FortiGate products or with other network security functionality. Just curious about how you think about rolling out those types of bundles over time.

Thank you. I think we keep on adding additional function and -- into the current bundle. At the same time, the fabric, non-FortiGate, we do see some additional potential to offer the new service. So that's where -- but we also are sensitive about the pricing. And so from our point of view, customer will keep on getting better and better both on the product and service going forward. And at the same time, we also will keep the healthy margin ourselves.

Our next question comes from the line of Ken Talanian from Evercore ISI.

So first, I was wondering if you could frame the size and the timing of your service provider opportunity in the pipeline. And I ask that because I know, in prior years, it's been a much stronger area for you than it is -- than it was this year.

Like I said, the international service providers are doing better compared to the North America. I hope that North America will be starting improving later this year, follow the international trend. Because security is a bit complicated, it's difficult for some of the customers to do the security. That's where we need to -- the service provider to add value and -- but there's also -- there's the transition, there's a change going on, whether to the cloud or to some other internal segmentation and for -- like we said in the fabric. The fabric approach definitely are accepted by a lot of service providers, not just the enterprise customers, but service providers also like the fabric approach. So that's where we're hoping the North American service providers can start improving later this year.

Okay. And also, Drew, it's been great working with you, but I was wondering if we could have an idea of what the time line is around the CFO search, if it's typically the 4 to 6 months and whether you're considering internal, external candidates, if you'd hire someone to help with that, et cetera.

This is Ken. First, we do have a very strong team in Fortinet, and especially Keith has a very strong background being CAO and even CFO in a public company before, so we'll keep on searching. But also Keith is also the one we -- we'll consider all this together. So we will search. We have a lot prodigious duty to do the process. But at the same time, we do have very strong internal team and can do the job. So that's the comment I have.

Yes. I think we're -- it's a great team, Ken, and I'm sure will work out for the best, however it does work out.

Our next question comes from the line of Melissa Franchi from Morgan Stanley.

So I just wanted to follow up on the prior question on share repurchases. I'm just wondering if I could ask for sort of an overview on capital allocation in general and how you feel about allocating your use of cash between those share repurchases versus potentially M&A. And then, in terms of M&A, are there any particular areas that you feel like you need to round out the portfolio of non-FortiGate solutions and potentially would look to round that out through M&A?

I think we always want to have the best capital structure, both pay it back to the investor, shareholder and also supporting the company growth. We also have a very strong engineering team. And we have only one build, all these like broad sort of solutions, the fabric internally, making and working together and also have the resource to build like a FortiASIC and SPU for the long-term performance and cost advantage. But on the other side, we also want to hear the feedback from shareholders, from all the partners to how to support and balance the amount of growth and also the profitability. For the merger, acquisition, we do look in and what can support the company growth into the next stage. The fabric, like we did some acquisitions in the last 2, 3 years is a little bit more related to supporting the infrastructure side. But on the other side, we are open and do want to look in a lot more broad areas in the space. And I do believe we're starting to go to the third generation of network security, which is the fabric approach, the infrastructure security, compared to the first generation when the company with NetScreen is [routing] the connection for our VPN. And then Fortinet started 17 years ago, 18 years ago. It's about content application security. Now the infrastructure fabric probably evolved in a little bit more broad than just network security so we do see a lot of potential, both building something internally and also keep on looking more broadly for the infrastructure side.

Our next question comes from the line of Keith Bachman from Bank of Montréal. (Operator Instructions)

This is Keith Bachman, I have a few questions. You indicated that real estate was going to be $60 million to $80 million in CY '18. Is there real estate expenditures anticipated beyond in CY '19?

The answer to that is yes. And particularly, or specifically, we're in the process of beginning construction of a second building here in our campus. That construction is in the very early stages as in still on the whiteboard. But we do -- we will spend the cash on that building this year and into next year, we plan to move into it in the fall to the end of 2019.

So am I to read that, that real estate could even extend beyond CY '19? Or do you think it closes up in CY '19?

I think what -- Keith, I think what Keith is saying is that it's consistent with what we guided last year. And the way to think about it is we've acquired the land around the perimeter of our headquarter property today. And for those of you who have been here, we need to build, one, a bigger building, and I think improve the quality of the overall structure as it is, so we needed to do that. And as we've said before, it's ultimately accretive to margin and EPS. It's near-term cash flow hit. But from a headquarters perspective, I think the bulk of that has been spread out over '17, '18, maybe a little bit into '19. But the numbers we offered a year ago I don't think have changed. If you go back and look at that number, I think we've said, yes, we guided roughly 60 and 60, I think, for a couple of years, and I don't think that's changing, Keith, right?

No.

Yes. So it's consistent with what it's been in the past, Keith.

Okay. I hope it doesn't extend beyond '19. On the revenue guide for the March quarter, if I take the revenue and the billings guide on a seasonal basis or, in other words, Q-on-Q, it's actually a little less conservative, if you will, than last year's March quarter from a guide perspective. And I just didn't know, it's only a few percent, say, on both the rev and the billings, but is there a reason for that? And you gave some description previously on the benefits of subscription versus more impact, if you will, from enterprise, but I didn't know if there was any additional comments you could make on the March guide in addition to that.

I think we're very happy with the velocity that we're seeing coming out of Q4 and into Q1. We have analytics as we go through the roll-off. We have a pretty good visibility to what we're seeing throughout the quarter.

So maybe just pipeline?

Yes.

Okay, okay. My last question is for Ken. Ken, effectively, if I look at March '18, if you take out 606, you're actually guiding operating margins to be lower year-over-year. And is the message there -- you mentioned moving into the fabric for the third generation. Is the message there as you move into a greater non-FortiGate percent of revenue that, that's less profitable than the overall portfolio? Because I'm just trying to understand, with such a solid revenue guidance of 14%, it's surprising to see implicitly you're guiding margins lower on a year-over-year basis, absent 606.

Keith, I don't think -- that's not correct. They're guiding margins up.

No, no. But if you take out the 150 basis points from 606?

No, no, no. That's not correct.

No. Add the 150 basis points for 606, and we're off of [that]. And this is a good segue to questions that have come up elsewhere which is whether or not we think it's appropriate for you to include the impact of 606 in your models, right? So the comment for the year, for example, we've guided to 17.7% to 18% non-GAAP operating margin. Additive to that number is at least 150 basis points attributed to 606, taking you 19.2% to 19.5%.

Yes, my mistake then. But Ken, could you -- thank you for clarifying that. Ken, could you talk a little bit about as you move either on an enterprise reference and/or to the fabric reference, how are you seeing that, over a longer duration of time, impacting the margin profile, particularly on the fabric? And I would think as you're adding a greater portfolio or share of wallet customers, it would actually be helpful to the margin structure over a longer period of time.

I don't think we'll have a -- at this stage, we don't see we'll have a margin impact because the fabric really are helping to lower the customer cost. Also, they see a lot of benefit of it. So we don't see the -- because there are some mix of -- some mix like management software, some other product has pretty high margin, some product a little bit low margin. But overall mix, we see still pretty healthy and also pretty much similar as we saw in the network security only.

Our next question comes from the line of Saket Kalia from Barclays Capital.

Most of mine have been answered but, Keith, I actually want to fully develop that last point that you made on 606. Just to make sure that we're all level set. So the question is, can you just talk about how much that 100 -- and we can all calculate but just to hear the range. Can you talk about how that 150 bps, how much that could add to non-GAAP EPS in fiscal '18 because I believe the guidance of $1.30 to $1.32 is before 606 adoption. So it'd be helpful if you could just maybe give us some guardrails for how much more that can add so we can level set?

There's a quick math. So we're looking at $0.12 for the year. And yes, to emphasize the point, again, the guidance on EPS is pre the benefit for 606. And then I don't think I covered off on the quarter, which is a similar discussion, we'll probably get it now. We guided for the quarter, for Q1, of 12% to 13% before the benefit from 606. After the benefit of 606, we'd be at 13.5% to 14.5%.

Got it. That's helpful. That's it for me.

Okay.

Sorry, go ahead, though. No, sorry if you had something to add, please do.

I was just going to also -- going back to EPS for a moment, I gave the annual number a moment ago. For the quarter, you're probably looking at $0.02 or $0.03 a share.

Q1.

Q1.

Our next question comes from the line of Brad Zelnick from Crédit Suisse.

Just following on Saket's question, you've used the words to say that the benefit to margins will be at least 150 basis points. So related to that, can it be 250? Can it be 200? What are going to be the drivers that could take it well beyond? And are you going to call out what the impact is over the course of the year?

Well, there's not much history here, obviously, for other companies. Nobody's gone through this yet. It's a very complex part of the standard. It can be whipsawed significantly by, one, your product mix, between products and services; your geography mix; your benefit payments; the term of your contracts. There's so many variables going into it. I think right now, at this point in the call or the quarter, we're most comfortable saying it's at least 150 basis points.

And he will call -- I do believe, Brad, they're going to call it out going forward in due time.

That would be helpful. And my other question, if we look at subscription and support, can you just remind us -- I mean it's just an increasingly important part of the mix given the margin profile and the momentum there. But can you give us a sense, even if you don't call out exact numbers, the contributions to growth, growth rates from better attach of enterprise bundles, 24/7 support, virtual and subscription mix? Any guidance there just on the relative mix of it all and performance would be helpful.

Yes, Brad. Just -- there remains a big opportunity there not only with new customers but in our installed base, and so it's really hard to call that out. It's hard to kind of forecast on a deal-by-deal basis. But the trend, if you follow the trend, the trend has been more fabric and higher-priced bundles and higher-priced services. And I think someone asked just a few minutes ago, do we have more opportunity to do that? And I should let Ken probably respond to this, but what I believe I heard was there continues to be an opportunity to find ways to add value in higher levels of service and more content through -- more security content in the subscriptions and charge more because that -- we've been successful doing that, and we'll continue to do that in the future.

Our next question comes from the line of John Weidemoyer from William Blair.

I'm sorry, I thought I took myself out of the queue. I had a question that Saket asked on the 606 and guidance. That's been answered.

Our next question comes from the line of Michael Turits from Raymond James.

Since there's just not enough 606, I figured I'd add some more. Just what are you guys thinking from a protocol perspective? I assume because you're modified, you're going to be reporting both 606 and 605. Will you be guiding both in the future? And where would you steer -- where are you trying to steer consensus to be, 606 or 605, for the rest of the year? I mean, obviously, it's going to be 605 right now but as we go forward.

Right. Going forward, it'd be 606, right, for guidance. We will obviously -- we will certainly be disclosing both 606 and 605 numbers each quarter as we go forward, so The Street and others can make the bridge, if you will, from what revenue would have been under 605 versus what it is on 606. That should be fairly easy to accomplish. And hence, we'll also be providing what the commission impact is.

Okay. So basically, let's expect the consensus will be 605 for now and then, after next quarter, it should flip over, right?

606.

Do you expect consensus to be under 605 currently, like tomorrow morning? And then -- or do you expect it to be 606 tomorrow morning?

606, tomorrow morning.

Our next question comes from the line of Walter Pritchard from Citi.

I look at your deferred revenue as a source of cash flow over the last 2 years. I mean long term, deferred I think has driven almost 85% of the volume, and short-term deferred has been a pretty significant driver. And I'm just trying to get a sense as to -- you did mention that you expect duration to be up in '18. Do you expect it to be up, sort of in line with what the contribution to cash flow has been the last couple of years? Or any sense you can give us there because it's a pretty big swing factor in our model?

Good question. And I think, as I've said before, it's difficult to predict the -- because it's really based on the mix of the business. To the extent we're successful with the enterprise and continue to grow that, I would expect that the extension of that would be we'd have longer-term contracts which, as you're pointing out, that builds free cash flow.

So up but indeterminate amount?

Again, depend upon mix.

And our final question is a follow-up from the line of Fatima Boolani from UBS.

Keith, could you remind us what proportion of the base has either a UTM bundle or an enterprise bundle? I'm trying to get a sense of what the opportunity within the base looks like as you push through more higher-value bundles.

Yes. I don't think that's a number we've disclosed previously.

It's still substantially UTM, Fatima. There's a longer -- there is a broad opportunity for sure of selling the enterprise bundle. We're just -- I would consider early, an early opportunity there. And it's been growing. We've been doing a nice job. Sometimes we pick it up in the renewal process, sometimes we pick it up with new sales. And we continue to look at ways to communicate that opportunity to the customer and also incentivize the sales force to drive that, so it remains a significant opportunity.

And this does conclude the question-and-answer session of today's program. I'd like to hand the program back to Peter Salkowski for any further remarks.

Again, thanks, everyone, for being on the call today, both calls today actually. And if you have any follow-up questions, please feel free to reach out to me. Have a good day. Thank you. Bye-bye.

Thank you, ladies and gentlemen, for your participation in today's conference. This does conclude the program. You may now disconnect. Good day.