Fortinet Inc (FTNT) 2014 Q4 法說會逐字稿

Good day, ladies and gentlemen, and welcome to your Fortinet Q4 2014 earnings financial analyst Q&A.

(Operator Instructions)

As a reminder, this Q&A session is being recorded. And now I'd like to turn it over to your host, Michelle Spolver.

Hello, everyone, and thank you for dialing in to the second call. I have with me Ken Xie, our Founder, Chairman, and CEO, and also Drew Del Matto, our CFO. Before we get in -- it's obviously just a straight Q&A session, so before we get into the Q&A, I just want to remind everyone the disclaimer I mentioned during the earlier call related to forward-looking statements applies to all forward-looking statements we may make during this call.

With that, let me see if we have questions in the queue. We do. At this point, John, and you can go ahead and start Q&A.

(Operator Instructions)

Erik Suppiger, JMP.

Thanks for taking the follow up. First off, Drew, you had said the CapEx would be $40 million to $45 million, I think.

Yes.

How can we think of that over the course of the year? That's a good step up from where we are. What are the factors that go into that and how will that roll out over the year?

Well, so I'd stick with the $445 million. I could help you with Q1, if that's okay. So probably $10 million to $12 million in Q1. I don't see any -- not to give guidance for the rest of the year, but I don't see any particular anomalies to that trend. It's just more timing related.

If you don't spend it in one quarter, it pushes to the next quarter. Sometimes things get pulled in. So hopefully that will give you some thoughts on that. I'm sorry. What was the other question?

A follow-up on that is, what is the incremental increase? Are you still spending on the buildings or what -- ?

Oh, I see what -- fair enough. Look, our probably CapEx run rate just on equipment is probably 40% of that, something like that. And then the rest of it is, we are looking -- we're building out some offices. We are looking at a building somewhere, potentially just a little CapEx there. And then also, we still have some money in there for our ERP.

Okay, would there be reason to think that that would be flat in 2016? Just qualitatively, can you tell us what the longer-term prospects are for CapEx?

I don't really -- it would be hard to do that, Erik. Again, I don't have any anomalies for that, let's just say yet. You are somewhat opportunistic on a variety of real estate related matters. We aren't really getting margin guidance. It would be hard to really get into that.

Okay. All right. And then just coming back to the OpEx, the margin guidance that you gave. When I tried factoring in my model from an operating margin perspective, it would seem as though you have a step up in the March quarter, then the OpEx then kind of levels off at that range, which implies that maybe the hiring is really -- it took place in December and some in March and then you don't do aggressive hiring in the following quarters. Am I interpreting that correctly? Or kind of what is the qualitative perspective on the hiring front?

I think that what I've been trying to signal or illustrate is that it's hard to predict hiring. We have been -- to be quite frankly, we have been really aggressively pursuing hiring throughout the year. You did have a couple of lumps, which illustrates what I was trying to articulate earlier, which is that they do tend to come basically in groups. Some simply because it could be some seasonality relative to calendar year end, for instance. There may be that.

Certainly people wanting to exit after their calendar year ends for maybe they have commissions tied to a year end or a quarter end or something like that. That's definitely possible, or bonuses. But what happens is we do get these experienced people and people tend to follow them. Good people follow the good people. We did have literally a couple of bumps like that that hit.

One of them, clearly, I think actually a couple of those, if not several of those, actually hit in Q4 and that's where we ended up within the quarter adding on literally four vertical leaders in the US. Again, these are more expensive heads and then literally following those people were those leaders, were other people who wanted to be part of their team.

Okay. Okay. Then I'll pass it on. Thank you very much.

Sterling Auty, JPMorgan.

Thanks, hi guys. I apologize. I've been bouncing between calls tonight and not sure how much detail and I heard some of the call. I'm curious, within the bookings/billings that you had in the quarter, can you give us a sense, especially with the increased investment that you're putting to work here, how much of that is coming from existing customers in terms of up sell, further penetration, et cetera, versus reaching out and capturing new customers?

Fair enough, Sterling. I believe one of the metrics we did give on the call was that we added 8,000 new customers. And so our base is roughly -- got a base of around 200,000 and we added 8,000 new, which we think is a healthy number, quite frankly. We do get a lot of that. We talked -- you could probably do a little bit of math implying something from the renewal rates that we share and retention rates.

We are expanding -- we clearly are expanding the customer database. I think it would be also fair to say that whether it's new customers or even expanding an existing customers, if you look at the deal sizes, there is kind of a change. There's a transformation going in how people think about buying from us and how they are engaging us. They are either engaging us in a new fashion, which is a new customer let's say, or in a broader fashion for existing customers.

That's why I think the big bill statistics are very useful to help you kind of get some sense that whether it's a new customer, existing customer, something is profoundly changing in the model. We are certainly further enterprising -- we're still very good about our enterprise penetration, let's say. That's where investments really are aimed. That's where we -- I certainly would share that that's where the bulk of that growth is coming from.

Would there be a way to -- I know these are going to be guesstimates -- I can't imagine you'd manage at this rate. Is there a way to think about the customer acquisition costs and how much of the sales and marketing line is actually going towards bringing in new customers? Because I think you talked about the lifetime customer value by going up [bocket it] and getting these larger enterprises, they've got very attractive, lifetime customer value so the investment is worth it.

Is there a way to quantify that by looking at the customer acquisition costs and then to your point, look at the renewal or return and try to get a sense of what that multiple might look like?

Sure, Sterling. I could do that qualitatively versus kind of sharing new numbers at this point. If you think about when you acquire something, you are really doing that on some kind of cost envelope. If you think of gross margin all in, gross margin, R&D, sales and marketing, and G&A.

And on a normalized basis, kind of taking out the ramp of the self productivity, you get back to something that's better than what profitability we're actually showing because effectively, today you are investing for the future. You're ramping up sales and marketing. In near term, don't really provide a lot of return. It takes a little while to get that investment back. We've seen some of the benefit of what we did early in the year in Q4, for instance.

And so if you think about it that way and then say look, you can expand that multiple by five overtime, expand the productivity of those sales people through tenured relationships, deeper relationships, broader relationships, and add to that the fact that with security, and this was part of the bet, the governance and the regulatory requirements make it very hard to kind of rip and replace. But if you back up to today, it's one of those rare instances where people are really looking up and saying, look, this is not working.

Adding some security onto my network gear is the [bide] really is not the right way to do it. It has to be security first. So that's a unique opportunity. And that's really one of the huge elements of our investment thesis. That's the opportunity today, the unit economics, take a bit of investment today.

There is a bit of obviously some kind of latent productivity in there that we are inferring. Just a bit to predict -- we go a little bit more longer term out when that is because we know the economics expand overtime. It's just a little hard to predict near term.

What's the average duration on the renewals, or I should say on the subscription, both renewals and obviously what people are buying as part of the [fund bill]? Let's just leave it at that.

We do anywhere between one and three year deals. Every once in a while, you'll see a five-year deal. Those do happen. I don't think it's material. The best way, I don't have the number for you, Sterling, but the best way to look at it and see if it has changed at all, which would probably be the most important thing, you can kind of look at the change in deferred revenue quarter on quarter as a percent of billings and that will give you a sense.

Okay. Great. Thank you, guys. Appreciate it.

(multiple speakers) do it for everybody. It will normalize over time.

Got it. Thanks.

Walter Pritchard, Citi.

Hi, thanks. I'm just wondering on -- you have sort of an R&D cycle or I guess there could be a consolidation cycle coming here in the securities base. It feels like everybody looking at buying endpoints. There's lots of small companies. It feels like we're entering a point where there will be a lot of acquisitions. You historically kind of stayed out of that as a company, had a very strong organic R&D capability.

How do you all think about a coming cycle here and is that part of your thinking with -- I know you answered a question previously saying that most of the incremental spend was sales and marketing. But do you feel need to uptick R&D spend given what may happen in the space and how is that impacting your spending for the next couple years?

I think the R&D spending will be keeping [pertius] mostly because a lot of R&D [prod] is really the long-term benefit compared to sometime the sales marketing you can see that's a short-term return. But on the other side, you can see we have the PNA to keeping you know where it keeping your [vas] in the new technology which can really change the landscape in a secure space.

We view really protection -- to do the protection of malware [satus too] the key. Took out a bite of security and point today, it is probably even more difficult to secure the computer to like it 10 years ago because you can see the mobile device, a lot of IOT, and even on the desktop side, it's the OS starting all mixed up together and the web applications that are even more powerful than some of the traditional OS. So that's making the endpoint market is the [word]. I think it's kind of a not growing, putting this way and also more difficult to effectively secure a lot of enterprise.

But on the other side, the traditional network securities, you at least secure the borders. We tried to move inside enterprise and secure the server, the data center, as security. That's we see the new trend. That's is really leveraged our investment in the last 15 years, in the ASIC chip and a huge performance and advantage we have.

I think this will benefit for us a lot in long term and the competitor has a more difficult time to catch up because the investment in this ASIC how we're take a long, long time to see the return. That's why so far I have not seen much other competitor even try to start in here. But generally in the network side, as the [others in ready] so far a lot of network vendor, a lot of network device, few can only handle the layer to layer three in the traffic.

For security relating to go up to the layer 7 to see the [u cert and comp] application. That's only the security ones can handle and especially trying to deploy inside enterprise. So if you -- like I said, the [umercy] R&D will be worth mostly keeping going forward and we'll see the lump firm benefit especially with the [foot in] we have huge resource in the R&D and will continue to lead innovation.

And then just a couple numbers questions. On the sales and marketing side, Drew, do you expect that channel incentives and -- it's not necessarily part of your spending. It might be a discount off the top or a greater margin. I'm wondering if that's part of more incentives, more margin for the channel is any part of your strategy here is you look to gain some share and invest in sales and marketing generally.

No it's not, Walter. No, we are not anticipating that. We are not offering that, certainly. You know, I think the feedback from the partners, they like the fact that we're actually doing marketing, as much marketing as we are versus kind of where we were a year ago, let's say, or before.

And the same with the sales. We're getting broader distribution, a broader presence, more support. And I think that's really what they appreciate.

Okay, and then just -- ( multiple speakers )

For the most part, we have a good channel program. For the most part I think the channel is relatively happy with margins that they have. So that's not an area that we -- that is broken per se in our channel program.

Okay. And then I guess last question I had is on your margins, your four- to five-year target here. When you were a company of less than half your size, you were pushing 20% margins. And I guess I'm wondering, obviously you're spending more aggressively now, but it feels like you get the leverage there. Why you wouldn't have margins go -- I mean, in 2011 you were at nearly 25% and the same with 2012. Why couldn't we get back to those levels given your anticipated revenue scale?

Again, Walter, we feel like now is just a unique opportunity to capture share. We feel like we've proven that. We more than doubled the rate of growth last year. It is a very competitive -- you know it's a competitive market at some level. You really have to invest. And think about it this way, we are transformational in a way.

The sense -- the idea of an enterprise sales force here is something that we're really building in a much more different way than what we have before and the lead generation engine, which is really a company for tomorrow, right. We had what we had which was working fine several years ago. I think if you look at 2013, it stalled a bit. And now we're really trying to build out for the future.

So we're built different than some others simply because we are transforming into a new go to market model. It takes a little bit of time and investment and that's why we're doing it. But you have to combine that with the fact that if you're going to do it, this is the time to do it. If you look at the security tailwinds out there, and from what I just said, the fact is when you get in these places, especially now, security is first.

It's not would you like some security with that network gear, right. You really have to have that security architecture defined. We have a great platform, an innovative platform to provide not only third-party certified security effectiveness, but also the fact that we can converge quite a bit of security technology on our network devices, which is what Ken's been doing for years. In many ways, we really are the leaders in that thesis.

And so I think that's resonating very well, and it is unique opportunity. Really, we feel like now is the time to go and really push ourselves into the enterprise which is a very high return over time, as we've talked about.

Okay. Thank you.

Jayson Noland, Robert Baird.

Great, thank you. Just a follow-up on Walter's question there. Is there anything more expensive about this approach with account teams that would warrant a lower operating margin target? It does seem like with more revenue at a much higher level than what you saw back in FY11 that something in the mid- to high-20s would be an appropriate longer-range operating margin target.

Jayson, I'll stick with the guidance we've given. We've given the guidance for 2015, which I believe was 14% in Q1 which was 7, and so you could probably infer from that that some of the quarters are going to be better than others. And then we also have gone out to 2017 and said we're committed to margin improvement by that point in time. Then we talked about the 20% growth, 20% margins, four to five years out.

There's nothing more expensive about your approach today than what you had a few years ago?

I'm not sure I follow.

I think we have -- we definitely have more new people that are coming on board. A few years ago, we were hiring less. We had probably a mature -- more of a mature team across the board because there were fewer newer people on the sales side and it takes some time for those people to ramp up and become protective. I would think that's probably the only big change.

Okay.

Also we winning much more larger deal that's taking technically more time and also we see the long-term value of this large customer is better than some [Rs and Bs].

Again, and I mean just to lay into that the cycle times. Again, a lot of these things have to be tested. There is -- you have to basically be able to test security products now before you put them into play, for the most part. People are very conscious about what they're putting into their environment and so they are testing things more and more. That kind of combined new relationships, new people, we're just giving ourselves a little bit of room here. Okay.

The bigger customer take a longer time to test compared to the medium size enterprise, they tend to lessen some of the marketing message. We try to hedge both. One is the [emergency] marketing and try to grow in the mid-enterprise. On the other side, the bigger deal tend towards the bigger enterprise which has more long-term value.

Okay. Thanks for that color. And then the second question on service provider. Have you seen much revenue traction yet with the NT 6-based product in the 5000 series family, the higher end chassis. Is that still mostly in test or are you starting to put some of that equipment into production?

I think internationally there probably has a little more progress compared to the US side. US, you do see a lot of testing going on.

When we talked about that I think two quarters ago, I think we had gotten a question about when do we expect to see an impact. I think we said in 2015 just because there's a long testing process that goes for with service providers for that high-end of a product.

For the new customer in [care red] like I said, tend to take two to four years. For the existing customer, for the new party, also tend to take one to two years.

Okay, that makes sense. Thanks a lot, guys.

Keith Weiss, Morgan Stanley.

Excellent. Thank you guys for taking the question. Let me shift gears a little bit. Can you give us some color on sort of your [trash] on when some of the products outside of FortiGate? We talked a little bit about that last quarter. I didn't hear about it as much this quarter. Is that part of the sort of the larger enterprise deals or is it just sort of larger FortiGate deals in terms of selling bigger boxes to bigger customers?

We have a more broad part of this than some of our close competitors, like I mentioned from the Sandbox into the FortiMail some other ADC or DDoS. We see more cross-selling, up selling opportunity. In general, probably this like a new part of the new technology. They work well with FortiGate and also with see grow probably faster than the FortiGate opportunity right now. Because all this up sell cross out in opportunity.

And it's coming from a smaller base. Yes, I think we talked about it. We have a lot of Advanced Technology products that work hand-in-hand with FortiGate or separately. Some of the products I think that have been doing well and we've gotten traction with our Forti AP product and then also the FortiMail and Ken talked about in his commentary about the advantage that we have.

A lot of the FortiMail product sales were along side of the FortiSandbox or APT subscription sales where you can do, those parts can work together to protect against advanced threats. Clearly, it's still a immaterial part of our overall revenue. It's an opportunity for expansion.

Got it. And should we think about this as add on sales? So that these are all customers that are buying a core FortiGate appliance and additionally buying these products, or are these products sold more on a stand-alone basis?

It doesn't usually happen. I think what you're -- it doesn't usually happen at the exact same time of the sale. A lot of times it's being able to go into our installed base which are usually FortiGate customers and sell them a non-FortiGate product. It occasionally happens in the reverse but not really, with the exception of maybe like I just said doing the FortiSandbox and FortiMail together.

Right. Right. Considering you guys do have a very broad breadth of product portfolio outside of FortiGate. Is that one of the sort of the [lay in] opportunities you guys are looking to invest more in or is it still sort of just coming on that core initiation firewall opportunity?

I think we probably imagine up front and once we see the opportunity some time we [like sor din ear] we investigate more. We see some good growth in the web security and some other like wireless WiFi combined with the traditional FortiGate. And now we also see even FortiGate has internal deployment that we announced the new FortiGate 3200D which has a 48 [tanga] 10 port single appliance.

That's even can replace some other internal switch but also can secure in all the seven layer. So that's really for the internal data center deployment. It really depends and but we do see some new tech grow faster than the FortiGate.

Got it. Good segue. Ken, you have been talking a lot about the growing opportunity within the data center. I think a lot of us -- when we hear about what is going on within the data center, our thoughts immediately turn to software defined data centers and more virtualized type solutions, where the ASIC advantage that Fortinet brings to the marketplace doesn't seem to be as -- doesn't seem to come to bear in the same way.

How should we think about what you bring to market with your appliances versus what a lot of other competitors are talking about in terms of integrating with a software-defined network?

First, I think we have the best software-defined network or some other virtualized solution and also we're working with some leading partners in this space like Amazon, VMware, (inaudible), so we are the same or even better than some of competitors in this SDM virtualized solution data center. But on our side, we do uniquely have the ASIC solutions which gave much better performance and also much lower cost of the protect the traffic.

So that's also needed for data the center. I have to say today there is a lot of talk about SDM resolution but it's on the real deployment few relatively small. I think you can see the Gartner called it the virtualized security in the next three to four years, still relatively small, not quite mature there. But on other side, a lot of data center they need security, they see the huge cost savings, leverage our high performance solution compared to traditional software only security function.

So that's where we see a lot of new opportunities inside a data center secure in the service provider. We are the only ones who can provide like a terabyte solution for the data center that also has 100GB interface which a lot of data centers that deploy the 100GB interface right now have not seen the competitor come close to what we have today. There's a lot of potential opportunity, but we also launched both SDM and also the virtual solution in the last few months, which is really leading in the industry.

Got it. That's very helpful. And then one last one from me. You mentioned on the main call building up some inventory and some potential of supply constraints with some of your chip providers. I just want to be clear. Do you guys feel comfortable that that's not going to get in the way of you guys meeting demand over the next couple of months? It shouldn't be any sort of constraint on you being able to fulfill everything coming in?

Ken, I am going to let you take that, too. We spent a lot of time on this recently, to be fair. We don't think so. That's really why we're trying to probably erring on the side of building inventory, quite frankly, to avoid that situation. We don't want to lose customers because we can't ship, for instance. We are more worried -- that's really the business for us. But Ken?

We have some good calculation about what was the opportunity costs, what's the inventory costs, what's the shipping costs. Once you have to say in Q4, we do have some rush shipping costs, have to fly in some product from the manufacturer which making the shipping costs relatively high because we ran pretty quickly.

Long-term wise, we also see sometime sudden high-speed network component can be take a long lead time to get and also even to manufacture. So that's where we tried to build certain inventory to really protect some of the off side if we have some, especially some bigger cost from bigger carrier, they tend to order in bunch and that's where we try to have a little bit more actual inventory even to help that.

Excellent. Very impressive quarter, guys. Thank you for the time.

Gregg Moskowitz, Cowen and Company.

Great, thank you very much. And I'll add my congratulations. And apologize by the way if this is already asked on the main call. I was just jumping around a little bit. But clearly you showed strong a billings number in each region. What I wanted to ask there is some other software and tech vendors that have recently reported Q4 earnings have cited issues in China, in Japan as well. Are you seeing any signs of sluggishness in certain countries within Asia-Pac or was demand healthy across the board?

We did get asked that question earlier. I'll repeat the answer, Gregg. Really the answer is first of all, I would start with we really kind of grown a 2X market rate there in Q4. That's the glass half full side of it, if you will. That being said, we had a couple of leadership transition issues throughout the year. I think there were really two. One was in the earlier part of the year.

And that region has picked up actually a little bit in Q4, but still did better in Q4 versus the early part of the year versus Q4 a year ago. So you're still kind of ramping up. He had kind of a tough compare for Q4. But doing a nice job there, and then the other one is just went into transition into Q4 but that one was forthcoming all along. I think that created a little bit of -- probably a little bit of -- explains a little bit of that.

That being said, across the board the numbers were okay. China wasn't the hottest spot for us, to be fair. But it was okay, and Japan was fine. I think we -- Japan was fine, definitely fine.

Okay, very helpful. Thanks, Drew. And then last quarter, I think you saw a bit of a pause, nothing significant but a bit of a pause with service provider activity for internal use purposes. Just wondering if that's changed at all from your vantage point.

It's Michelle. I don't think we saw that last quarter. It was several quarters ago. I think it was like last year. But no, our service provider business performed well and has been performing sort of at normal characteristics, I guess, for the last few quarters.

Okay, thanks for the clarification. Roughly how much of the way would you say, Ken, that you are kind of through the ASIC product cycle at this point? Is it kind of a little bit more than half still ahead at this stage?

I'd say MP6 based probably about half, or a little bit over half. But we also have the new chip come out like the other CP-69 will come out which also has pretty much all FortiGate like content level performance. CP is the content processor. NP in the network processor. NP tend to go to middle high end speed up network hard side performance and the CP tend to [ascend] out of FortiGate function there.

And then the other one, we have the [search] we call the system-on-a-chip, which we launched, I think, almost 2 years ago. That's where the 60 [thees] are using that chip. We also in the process of doing the [associate] three. That's probably maybe toward the end of the year sometime. Based on the term every three or four years, we have a new chip for each category.

If I could actually just clarify about the NP6 cycle. So as Ken said, about -- when we say 50% of the way through, what we're saying is 50% of the models that will have the NP6 processor in it are currently shipping. They are at various stages of shipping. For example, we announced the 3200 -- FortiGate 3200 last month. When I say various stages, a lot of them are in the very early stages so there's plenty of ramp-up opportunity for those models.

Okay. Got it. And then just one last one. It sounds like FortiSandbox did quite well, and I realize in a lot of cases it can be a facilitator for some deals and some larger deals. But just wondering if you've been able to sign any sizable orders for FortiSandbox, specifically or if it's a bit too early for that.

We did talk about it, Gregg, in terms of the deals that we won. Ken talked about it. Let me actually see if I can find it. So basically it was a -- these are an example of some of the deals that we won. One of them would be a large government service provider, a global plastics manufacture, and a US-based law firm. So obviously we won more than three but those are the three that we called out in the script.

Great. Thank you.

Michael Turits, Raymond James.

Hello. Good evening. Again on the margin guidance, you've got your guide for 2015. Can you just clarify, you may have done it before. I apologize. The comment I think that you said that you get operating leverage in 2017. Was that the implication that margins are flat 2015 to 2016? How should we understand it?

Sure, Mike. So, again, we've given guidance for 2015, 14% for the year, 14%. We talked about certainly margin implement in 2017. We really didn't talk about 2016, to be fair. Just to give us a little bit of time in our new model, we're still trying -- look, try to go back to the strategic part of this. It's a unique opportunity we believe to take share not only at market growth but from incumbents for all the reasons we mentioned, security tail winds, ability to converge on our platforms, certified (inaudible), all of that.

And you know, look, it certainly worked last year for 2014. So we'll continue that. We're giving ourselves a little room. It's a little bit hard to predict exactly how people will come on board. And then how that will turn into productivity on some of the larger deals over time. And so we're giving ourselves a little bit of time to get a sense for the pulse of that before we really guide more specifically on 2016, if you will.

But look, as we get better information, we will absolutely share it. But again, we just have a lot new here and we're trying really to find a reliable pulse before we really try to signal what that is. I think we just wouldn't want to miscommunicate it.

And then, that's understandable. I guess next question was just on the five-year plan, I guess. Is that what you said? There's a five-year, 20%? (multiple speakers)

Four to five.

Someone asked it before and I just wanted to ask it again. Which is that we understand the unique opportunity you talked about to take share now but I guess the implication -- I would infer from you saying that it's four to five years out, that that's your kind of long-term steady-state margins are 20%. And I just want to re-ask the question. If you had done 25% before, are you saying that you don't think you should be able to get back to that level? Has the business model changed in some way?

Michael, you certainly get into the law of big numbers. I think 25% growth on $400 million or $500 million.

No, no, no, no, no. I'm sorry. We're talking margins, margins. You had done 25% peak margins. So like why wouldn't you be -- I understand near term there's a share opportunity. But long term, why wouldn't I think of the kind of steady-state operating model for Fortinet as achieving the same 25% margins that you had reached at your peak?

Is there some difference in the environment? Is there some difference in the model? Why couldn't you get back to that?

Fair question, Michael. First of all, we said at a minimum. Secondly, it's a little bit difficult. You got to figure out how big are you going to be and how much can you grow, and I think the balance is between growth and margin at that point.

We're trying to get some indication, hopefully, for some people to think about us longer term just to get some direction. Look, I think what I would take away from that is we think we can continue to grow. We've shared some of the reasons why. And then also that we're committed to profitability and returning that obviously over time at a minimum at that level.

Okay. And then to the CapEx question again, also you had been doing -- if you do your $40 million to $45 million, that's about 5% of revenues. I guess it would have been my thinking that you had really peaked with some of your ERP builds last year which was closer to 4% of revenue.

They said that 40% of next year's $40 million to $45 million is equipment replacement. So I just -- we're trying to understand where or that goes back down to the 4%. We were only modeling like $25 million in CapEx for you and now it's $20 million more than that. Is that really one-time stuff and it heads back down? I could really use some guidance on that.

There is some one-time stuff in there to be fair. Again, we still -- I think I pointed out that we're looking -- right now, we're trying to determine the balance between rent versus own on some buildings, quite frankly. That's a bit of a thought in there for this year.

So if I think it'd go out to 2016, obviously given the guidance but -- .

Yes.

It doesn't make sense to think that it comes back down again in 2016. (multiple speakers) It's real important to me. Fee cash flow is the basis for the way a lot of us value you. We're trying to understand what's a one-timer versus what's a long-term trend.

I think -- maybe here's a way to think about it. Kind of buildings aside, I think we were growing -- we were spending run rate -- the run rate CapEx was probably closer to 16-ish, something like that. I think I said 40% of it. It's about that or less. And then the additional items in there are building out for an ERP system. And then we parked some money, quite frankly, for some real estate that we may or may not do.

It does make sense. At least that's an issue to think about it declining again. Obviously not giving guidance but it sounds like those one-timers should be over in 2016?

The real estate, we don't know. ERP, you hope so.

Right.

Think of it this way, Michael. It kind of occurred then or now. We're kind of building them into now to be.

Right.

Again, without giving 2016 guidance hopefully qualitatively, that will give you some sense.

Okay. And then lastly again -- (multiple speakers) something. And then lastly, as I think about modeling cap cash flow from 2014 into 2015, 2014 you had a benefit from the $20 million payment from Palo Alto. So if I'm thinking of the base for 2014, I would take it down by that $20 million.

And then it seems to me that you've said that your cash taxes which were pretty high this year. Your cash taxes were -- what were they?

42% and I think all in about north of 40%, about 42.5%.

Right. And your 20 to 30 going down to 25 for next year so that gets me a little bit of a benefit. Outside of those two, if I add back, take out the $20 million from last year and then I normalize for the cash taxes going down, should I think about cash flow from ops as growing roughly in line with net income 2014 to 2015 outside of those two adjustments?

Yes, that and the items that we discussed on the non-run rate CapEx items that we talked about.

No, no. Cash flow from ops. Just cash flow from ops.

Fair enough. I think that's right.

Okay. What about inventory? It sounds like you're building inventory.

The only thing to do, Michael, the only other thing I would add is take a look at DSO because you have to kind of -- I think DSO was relatively high at the end of December. I'm not sure how you model that going forward. But that seemed to be a little higher than run rate. (multiple speakers)

December's fine but in other words if that reverses, that actually would then help us. Right? Or do you feel like that's the new normal and it stays that way going into next year?

What, the DSO at 74, the new normal?

Yes. Is DSO permanently higher?

No, I think it contracts.

It contracts. (multiple speakers) That could be helpful in other words going into next year.

I think the -- that's my point. You have a high AR balance which you didn't collect in December this year, 2014. That's what I'm trying to point out.

Right, and then -- but wouldn't that have been -- you mentioned that you might build some inventory. Does that offset it? Is the inventory use of cash?

Not by that extent.

Okay. Okay. I think that's it. Thanks very much, guys.

Brent Thill, UBS.

Hi, everyone. Thanks for taking the question. This is Fatima Boolani on behalf of Brent. I just wanted to dig into the reinvestments on the sales and marketing side. Last year there was certainly a push to become or have a better partnership with some of your enterprise resellers. So I wanted to get a sense of the dynamic that you have in place currently as you build out what is effectively a field -- vertically focused field salesforce. I was just wondering if you could just walk us through that.

I think that's a lot of a like a top 10 or 20 national resellers in the US. We started engaging last year but [sumo] in the ramp up stage. We definitely will benefit from all of this new partnership especially for the some verticals and large deals. That's where we keeping [you mast in] this year and every few -- that's also the potential growth area.

It's two different areas. One of them is the channel investing in the enterprise reseller channel. As Ken said, a lot of that happened in 2014 but some of it will continue in 2015. And then what we talked about in 2015 in regards to enterprise and verticals is building out teams of which we've already hired leaders for the teams. The ones that Drew talked about specifically were financial services, healthcare, government, and cloud.

I believe it's all complementary in terms of channel. That's actually something that's favorable channel because everything we sell it goes and normally gets credited through a partner. So everything -- we're 100% indirect in that sense. However, it gives them more coverage and broader distribution, if you will. It's a good thing for them, right.

But also you [made] a lot of in the lead gen and some other marketing activity.

Yes.

And just to clarify, that 43% growth number that you cited for US enterprise, was that the US enterprise business as part of the -- your salesforce that you're continuing to build out or was that via your enterprise reseller partners? I just wanted to make sure I get the distinction correctly.

Sure, and thank you. That is enterprise reseller channel. Those are what we call national resellers, larger ones, like you've probably heard of Vacuvont, Fishnet, CompuCom, those type of resellers. It's not all of our enterprises business in the US. It's the business driven by those partners.

Understood. As you build more strategic partnerships with these larger resellers, are you seeing any revenue concentration with some of the larger ones?

Revenue concentration how? More revenue?

In terms of accounts receivable or even revenue concentration with larger distributors or resellers?

It's a relatively smaller percentage and both compare before and also compared to the competitors.

Fair enough. I was just trying to juxtapose the 28,000 in channel partners you have. I imagine some of the larger ones would be pushing more of the volume for you. That's helpful. And then maybe to touch on an earlier comment that Drew made around pricing leverage. Just curious if you're seeing sort of an easier promotional environment.

Are you seeing less pressure to discount, just given the transformational nature of the purchases as you highlighted earlier? I just wanted to get a sense of whether unit economics are improving alongside with ASPs not seeing a downdraft. Just would love to get your views on that.

I think, Fatima, the best way to think about that is you bring in new people, what we're very focused on, and people who know how to sell value. The better they are at selling value, the less you have to use price as a lever on a deal in a negative way. And so we're very conscientious. We do more training. We're very thoughtful about how we price. We spend a lot of time on that in the training.

We talk a lot about that, and really trying to focus on ensuring that we are getting the best value for our deals. And we see a lot of that in our bigger deals. That's what we're really trying to instill is that nature. This works hand-in-hand. If you get good people because smart salespeople obviously have a shared interest in that, not only do they get higher performance against near-term quota, but they don't undermine themselves are undercut themselves on the longer-term relationships.

You know how it is. You always kind of start your next negotiation on where you landed on your last negotiation in terms of price or discount. And so that's really where the focus is. It's hard to correct anything you've ever done in the price -- passed on pricing. Not that I'm saying that that's bad. But your opportunity really is on the new where you bring those people in and arm them with the best information, the best negotiation skills and training and support so that they could drive the best value out of every deal and customer.

That's helpful. And a very last one for me, if I could. I was wondering if you could reconcile the increasing proportion of the entry-level appliances in the billings mix relative to the high-end just reconciling that with the tremendous large deal momentum that you're seeing.

I think it's really -- the high-end grew more, the 3700D and the 1500Ds have just resonated extremely well in the enterprise space. That's very much the story.

Yes, it was a 10-point increase in the high-end products year over year so that's really, really where we saw the mix shift. We've talked in the past about product billings or product mix for billings. It really just depends on the models that make up the type of deals that we do.

In some cases, in some past quarters, we've had a higher concentration of low-end products but it's because it could be a million -- multi-million-dollar deal with an enterprise that has a -- it's a very large distributed enterprise that in addition to buying the high-end units, they also need to buy 100s or more of the low-end unit so that changes the mix shift. In Q4 particular, really where we saw the biggest change is at the highest end portion of our products and that was a 10-point change.

I appreciate the color. Thank you very much.

Aaron Schwartz, Macquarie Research.

Just two very quick questions, if I could. Drew, I think you mentioned a couple of times that hiring was extremely strong at the end of the quarter. It does look like net headcount was not as great as the last couple of quarters. So I just want to make sure I understood that or was it were you alluding more towards some of the continued hiring in January?

You get a lot of -- it's a really good question, Aaron. Actually, a lot of them are Jan 1 starts, early Jan starts, let's say. They're not all on Jan 1. You do have a lot of people I think, especially in sales and marketing who basically finish the quarter and their whatever compensation is tied to that, then they join. A lot of people really joining. We have a nice ramp in January, for sure.

The other thing I would point out, though, the people we hired, we mentioned some of the vertical leaders and some of the marketing. They are more expensive. Those types of people are going to be a little more expensive as well. Just leadership side of it. There's a little bit of that in the play.

You're bringing that cost on really with -- usually there not productive, they're just building teams. You are incurring the related cost of ramping that piece of it up. So that's really the story.

Right, and then just secondly real quickly. You are obviously not the only security company that's hiring pretty aggressively right now. What are your views on wage inflation? I think a few people have asked this already. It just seems like your reinvestment to drive revenue, you're paying a higher cost per unit of individual at this point. If we tried to assume some productivity over time, it just seems like your cost structure is going to be higher per individual than it has at any time in the past. What you seeing on the wage inflation side?

Honestly, we haven't. With some of the leaders you negotiate your -- you're obviously negotiating packages. But by and large, we're holding -- we hold pretty strongly to our benchmarks which are surveys of the industry and we pay at market. Really, I think the opportunity they see is really on the opportunity of performance and we pay for performance. It's not -- the OTEs -- I would expect similar. I have not, essentially, it's a fair question you're asking because there obviously is demand.

To be fair, I'm not sure if there's that many experienced security people out there. But we seem to be getting a lot of interest to join our company. We generally have the OTE at market I haven't seen a big shift there. But where I think people believe -- I believe -- people believe they can over perform. They believe they can make money here and I've heard that time and time again. That's really why they join and that -- we've seen some of that.

Great, thanks so much.

Erik Suppiger, JMP Securities.

Ken, quick question. Yesterday, Juniper discussed some change in their product roadmap and they talked much more about developing high-end firewalls that can leverage their high-speed routing ASICs. I am just curious, that's been where your advantage has been with the terabit performance. What is your take in terms of them trying to leverage those -- their silicon technology to presumably match you or maybe leapfrog you?

I think there's a few networking companies talk about like doing whether the high-speed and that was security all integrated the security into a networking gear, whether routing a switch in there. But the difficult part really is all on security because security there is two issues. One they need a more high layer compared to the routing switching, and it's really layer to layer on the devices that follow the standard routing particle. But security go out to the layer 7, the content application user.

The other part really is security tend to need a like a 10 to 100 time computing power to process the same traffic in routing switching. So that's where if they want to integrate security with routing switch and security is a heavy part and they need much more effort. So far I've not seen networking company successfully doing that because they are pretty much all behind on the security, on the function there. Just a part of the regular routing switching chip into the security not working. But really (inaudible) building some security chip (inaudible) layer 7 function is the key.

All the security gateway, they have all routing switching function. All the routing particles are all built in there. But on the other side is a networking switching gear is all lacking a lot of security function. (inaudible) the chip is really eventually the enterprise, they want to look beyond the connection. They want to look at all the application and content of the user. They really need to process this layer 7 process.

That's the reason I talk about internal security because internal all connect by the switch. It 10 to 100 times faster than the way they connect with the internet. Last they need to have high-speed processing off the security traffic. Which the chip, the (inaudible) is more critical. It's very important. Because using a sulfur load on a standard PC server will not working because all the internal traffic, all the other servers all have the same computing power so you really need to have most dedicated power, much huge computing power advantage to process the security traffic in high speed environment.

I see some networking companies still behind because they need first deliver the secure function in sulfur first and then keeping invest in making the sulfur while function the security device to the power and improving the performance and also free up a CPO for some new function. That's the process we have been doing for the last 15 years. That make us have the best platform, both on the multifunction and also on the performance. We feel that secure internally inside enterprise will be huge opportunity and we have a huge advantage here.

Okay. And Drew, do you -- can you give us a sense, you said one of your components has a five-month lead time. What kind of component is that?

(laughter) I am not looking to upset a vendor. Ken, what could you?

I think sometimes some high-speed network can take a long time but sometimes it can be different. It's really about a forecasting about some other things we try to do better. But also from time to time the larger deal whether the huge enterprise or some [carrot] can be lumpy.

Very good. Thank you.

(Operator Instructions)

Michael Turits, Raymond James.

Hey Drew, I was told that you answered my cash taxes question while I wasn't on the call. So I missed out.

We were waiting for you to ask it. (laughter)

I am always ready now.

Thanks. I didn't want to deprive you of my question here. So I was pleased obviously the cash taxes will be lower this year, 25%, $25 million, which is close to like 25% rate out of cap free tax. So I guess that's my question. I'm always on the sustainability of cash flow items question. That's my question. We had a big cash taxes in 2014, kind of lower than I would have expected in 2015? Do I split the difference when I go into 2016 and go back to a 35% rate?

It would be hard for me to guide that. Quite honestly, I've hired recently a new VP of tax who is wonderful. And she has been very helpful on really getting a lot of comfort in the tax. But it would be premature for me to jump into 2016. Very, very comfortable about 2015.

Asked another way, is there any reason at this point not to think that that rate, that tax rate that you looking at for 2015 that that's a sustainable rate as a percentage?

I think 2014 is high. 2015, I think there is probably some benefit in there in terms of credits and other things. The other thing you have to realize is you are always dependent on the government for reinstatement of the R&D credit. There's a variety of other kind of international tax things going over. You have jurisdictional mix and just a variety of other factors. I will certainly do my best to follow up on kind of the long-term strategy. More tax but I think that's the best I can do for now. I would say 2014 is hopefully an anomaly here.

2014 high and maybe 2015 just a little bit lower than trend.

Yes, I would think so, Michael. I would share that much. I could say that we definitely have some benefit we're counting on we believe we're going to take to the bank in 2015.

Okay. Thanks very much.

Ladies and gentlemen, that does conclude our Q&A session for today. I'd like to turn the call back to Drew Del Matto to close the call.

Thank you for everyone for your support and your insightful questions. We really appreciate your time and support and look forward to talking to you soon.

Okay, ladies and gentlemen. That does conclude your conference. You may now disconnect and have a great day.