Fortinet Inc (FTNT) 2014 Q1 法說會逐字稿

Good day, ladies and gentlemen, and welcome to your Fortinet Q1 2014 earnings financial analyst Q&A.

(Operator Instructions)

As a reminder, today's conference is being recorded. And now I'd like to turn the conference over to your host, Michelle Spolver.

Hi everybody, thank you for joining the second call, a half-hour after the other one ended. Before we start I have Ken Xie and Drew Del Matto here with me in the room, and again, the purpose of the call is to answer any remaining questions that you have.

Before we start, a reminder that all forward-looking statements made during this call are subject to the disclaimer I read during our earlier call. So with that, I think we can open up the line and take some questions.

(Operator Instructions)

Aaron Schwartz, Jefferies.

You obviously, and I know you covered this on the last call, but some very good traction with the large deal flow. And Drew, you alluded to this, but in the fact that you are seeing, I guess, a mix shift in your business, towards large deals, and I guess that's related to the dedicated investments in the enterprise team here. Can you just walk through how you manage the pipeline differently, how you manage the sales funnel, et cetera? Because those are always lumpier, and the timing's a little bit more difficult, but how you manage that process given the shift in the business?

Yes. I'll do that, Aaron. I'll also probably take you to what's changed, because I'm saying to people that's a bit -- I don't know the dynamics word, but we're building process, I think is the way to say it, and adding headcount, so it creates a bit of complexity until things normalize, is one way to think about it.

But yes, quite frankly we're using salesforce.com more religiously, probably, than it's ever been used before here. And so that happens globally, and people domestically may do that differently than people in the southern hemisphere, or the eastern part of the world.

It takes a little bit of time to get everybody online and using the same process, and doing it in the same way, and us analyzing it and being able to make valid conclusions on it, in the same way. It's a bit of getting the pulse to go and getting the beat at the same cadence, and then making sure that we're watching that, if you will.

The other thing that's changing is, I think we're doing a lot more -- we're spending more on marketing, let's say. And as we drive campaigns, we're trying to monitor those campaigns and the turn of those into leads, as we invest in generating ads from campaigns, let's say online ads, whether it's on LinkedIn. Or as you read some article, we know you had some interest in our products or network security, and we serve you an ad, and then we're monitoring those types of situations, and you do that in a variety of ways.

But we're linking tools to do that so that we can monitor those investments through to SMB -- salesforce.com, excuse me, SMB, salesforce.com, and then ultimately through to the billing. And again, new people, new process just takes a while to get it, I would say, normalized and working in a consistent manner globally.

And I know you're newer to Fortinet, so maybe this is a question you can't answer, but as you go through that process, and I think this question you hit on, with the guidance question on the first call, but is there a different coverage ratio that you think of until you get comfortable with the process? Or different conservatism that you take, just given if there is a mix shift. Just timing on large deal flows is obviously tricky for anyone.

Is that in terms of, are you asking that with respect to guidance?

Yes.

Is that basically how conservative is the guidance?

Or is it a different sort of coverage ratio you think about, with the sales coverage versus the plan or --

I think there's two questions in there. I'll try it this way.

Look, I think we're very comfortable with guidance. I think what we can see, the forecast certainly supports our guidance. That would be how I would answer that question.

In terms of coverage model, that's a variety of questions, because that could go to how many accounts per rep, if you will, and that is something we're looking at. We feel like right now we have a lot of accounts to cover, and probably more not enough people to cover them efficiently.

And then, the second, there's ultimately a piece of that in terms of what's your SC ratio, and so forth and so on. I think we're starting to figure most of that out. I think there's still opportunity for improvement, and I think as we get -- Ken and I are trying to be very metric focused as we go through this, and we've done some.

It feels like we made some progress last quarter, and to be true, even some of that was in place before I came here. But there is a focus of improving it, and I'd say Ken, just working with Ken to share a little bit he's very focused on measuring ROI, as much as anyone.

And people, as a leader, I think that translates throughout the organization, so he's a great leader on that front. It makes it very convenient for us, and makes probably my job a little easier, the marketing job a little easier, as we try to implement these things and tracking them, because he expects that we do that, and that's from the top of the Company.

Also, Drew, before joining Fortinet he worked with Symantec, a much, much bigger company for nine years. So that's where he knows the space better, knows the model better and also a lot of like the service-based model, so that's also helped Fortinet grow going forward. So that we see much better shape compared to like a couple quarters ago.

Great. Thank you very much.

Nandan Amladi, Deutsche Bank.

Gregg Moskowitz, Cowen and Company.

Congratulations everyone, on a very nice Q1. I guess the first question is, you alluded to some vertical areas where your current coverage is a little light. And I'm wondering if you could elaborate on where you see a particular opportunity there?

I don't know that we alluded to areas that were light. I think we talked about, there was certain areas that we saw better or bigger opportunities, and where we are putting our investment focus. One of those areas is the enterprise, I think is what we called out.

Yes. I think we called it out, but Gregg, I think there are opportunities in a variety of the verticals, the ones you would go to. And not to try to bind ourselves too much, but clearly, areas like healthcare and financial services are opportunities that probably any Company would be looking at, as they create their coverage models to be focused. If you look at the general IT spend in those areas, those are obviously areas that would be of particular interest.

Okay. That's helpful, Drew. Within the high-end, obviously very good performance there.

And you talked to some of the components. I guess I'm wondering, it sounds like service providers continue to strengthen enterprise, and obviously you're seeing more traction, but wondering how you might parse the strength there? Whether it be across enterprises versus MSSPs versus telcos, just if you have a little more color around that?

I think service provider, I wouldn't say it continues to increase. We continue to see a stabilization or normalization in spending with service providers, that took effect really in Q4 of last year. So if you look at the mix of our vertical mix, service providers gained a percentage point, it was 25% in Q4, 26% in Q1. So we saw a bit of growth there, but not a lot, and the large deals indicate that we saw a lot of growth, so therefore we saw growth and traction within the enterprise, large enterprise.

Yes. I think the deals, we picked the deals in the script that we shared to be a good reflection of what we thought was a good reflection of how the business felt to us, or appeared for the last quarter. I think we mentioned a variety of enterprise, some service provider, and a little bit of MSSP in there, as well.

I don't think there was a big mix shift. I think definitely there's a tailwind, as we invest in the enterprise, I don't know that tailwind is the right word, but we're actually obviously getting some traction there. As you can see in the deal stats.

Okay. Perfect. And my apologies if I was a bit off on the service provider commentary, just jumping across back-and-forth between multiple calls tonight. I guess a just of couple other questions, Drew, any thoughts I guess on this point as to how 2014 CapEx may shape up?

Yes. Actually, again, we're not giving guidance, but we're guessing that for the next three quarters, I think we're going to spend about -- we figured about $15 million for the rest of the year in CapEx? On top of our normal run rate, about a $15 million -- and we still have some building costs left here, we just moved into a new building. We have a solar project going on here, I think it's like $4 million in there.

And then we're also going to start an ERP project at some point. We haven't really started that yet, but that will start, and so it's about $15 million more than probably the prior year, with the run rate.

Okay. Fantastic. That's great.

And lastly, for Drew or for Ken, just wondering if you had any update on the COO search? Thank you.

We still open -- still going on and also, we feel that's the area we can also improve in the Company long-term. But like I said, we don't want to rush into someone in the position but we still open searching right now.

Perfect. Thank you very much.

Jayson Noland, Baird.

Just a follow-up to the previous question. Would this role, the COO role be focused on internal ops or external sales and marketing activities, or a bit of both?

Probably more towards the field outside. But also, depending on background, also probably help internal -- also would be helpful, but so far we see how to help in -- if we look at company spending, probably somehow in this network secure space. The sales and marketing really is the biggest spending, probably even like double, triple the spending on R&D and some other outside.

So that's where we feel if there is a way we can be more efficient and also better return on investment, and now even a more aggressive invest, that's where the sales and marketing and also build a bigger capability, that's probably the area we want to have the position to be more focused. At the same time, internally, we also see the team starting to do much better improving this area.

Okay. Thanks again. A question on NFV. We hear a lot about the potential for architectural change in the networking market with SDN and NFV and carrier.

In the carrier market, do your customers engage with you in regards to network function virtualization? And do any of them tell you they're uncomfortable with a custom basic, in preference to off-the-shelf hardware? Just wondering how much this comes up with your carrier customers?

I think first, we use an older CPU, older network processor chip, whether like a Cavium or Worldcon or Golcon, whatever, the same as our competitor. We never lack of any usage of the commercial chip on the market.

And then, in addition to that, it's what makes unique for Fortinet also gives us a lot of advantage is, we also build our own chip because the network security market is relatively small, compared to what the network in other server market, PC market. And that's where almost no chip companies try to build a chip dedicated for the network security, which also needs more heavy processing power to process all these security functions. So that gives us a much better performance advantage, and also we can offload the CPU or other network processor from the market, so we can offload that to the own basic chip.

And that gives us like a huge advantage compared to some other companies just using software to do the job. Which whether it would be slow on the CPU or the CPU cannot handle market function, wireless security function or the virtualized function, network function of SDN comes in. So that's where -- in the platform side we have a better platform, more performance, better function, more function compared to the software only, using the standard PC server to do the job.

On the other side, I still feel the virtual network function, or NFV, or some other SDNs, still relatively early stage and the market -- whole market is still kind of not quite defined yet. That's if you look in the report a couple weeks ago from Gartner, they say in like 2017, that's probably three years from now, the whole network security is still less than 10%. Doing the virtualized deployment, that's where like in the data center and enterprise, are in the tiered service provider.

So still relatively small. So that's where I believe the platform we have gives us more advantage, compared to the software standard server, because we do use all the latest CPU, network processor chip but additional ASIC boosts the performance beyond the CPU and network processor can handle, give us much better advantage, and also free up the CPU to handle some other functions.

Okay.

The only thing I would add, Jayson, is that I think the first part of the question is, do we engage directly with carrier customers? We do. That type of deal, and that in terms of the size of those deals, we would engage directly with.

And then the other thing is, that from a stats perspective, we're in pretty much in all of the top carriers in the world. And the reason, the primary reason that they buy Fortinet is for the performance, which is enabled by our ASICs. So I understand the question, but I think the answer is more than comfortable dealing with us and our technology approach.

Yes. That's where you see like some competitor introduce some chassis-based solution earlier this year. I have to say, they are way behind, and also they are still using the standard server, which is difficult to survive in the high-speed high-redundant like carrier environment.

So we are doing this, we have the dedicated chassis, dedicated hardware for 10 years now, and dominant in this carrier space. So far we're not seeing a software-based competitor that can able to really compete or survive in this very high-speed and high redundant environment.

Makes sense. I appreciate all the color.

Michael Turits, Raymond James.

It's James sitting in for Michael. I had a couple questions on the model.

First, Drew, I wanted to make sure I heard you correctly on CapEx. Did you say that you'd spend $15 million on top of the normal run rate for Q2 through Q4? Or is that for all of 2014?

Q2 through Q4.

Q2 through Q4? Okay. Thanks.

You had it right other than that.

Got it. Okay. I had a comment, sorry I missed it, I have been going back-and-forth on a couple calls, you had said you were comfortable with 14% growth consensus for the year, but was that figure for revenue or for billings, for both?

What we said was, one, we're not giving annual guidance. We said that we're just directionally telling people that we believe we can grow 2X or better of the indicative benchmark, how Gartner and IDC are calling the market, call those the benchmarks of 6% to 7%, so those come up to the 14% range.

As you've seen we've been -- obviously Q1 is doing better than that. And I think Q2 is north of that as well.

And the 14% number was the high end of that, of the market growth range.

Yes.

Okay. So basically no change? You're keeping the 2X guidance?

Yes.

Taking a look at cash flow, for cash flow from ops, is there any reason that shouldn't grow in line with net income for this year?

Yes. So we gave you -- from operations -- oh, I see. Yes. Of the -- let's see, the tax rate -- I'll give you, does it help to give you where we see the anomalies?

Sure.

I think so. I think the one thing that's always the wild card, by the way, is collections. We obviously had a good collections quarter in Q1.

Somebody asked me the question about linearity, you can see that the benefit of the linearity shows up in Q1. So hard to predict how that would happen to Q4, but I think that's probably reasonable to assume that we'll have a similar pattern to the prior Q4, in terms of how that works.

In terms of cash tax, I think for the year, we will end up spending, I believe, about $36 million, $37 million -- call it $37 million. And we guided a one-time payment last quarter of $18 million in Q1 that we actually did make. It was slightly -- it was roughly in that range. I think we paid about $23 million in taxes overall last quarter, with the anomaly -- there was a one-time true up a piece in there, that gets you up to that amount.

And so, I can't think of any other anomalies on cash from operations. You have the inventory turns number too, so that can vary quarter-to-quarter, a little bit, but that's how I would think about it.

Okay. Great. Thank you.

Keith Weiss, Morgan Stanley.

This is Melissa Gorham again, for Keith Weiss. A follow-up question for Ken, on competition, specifically related to Cisco.

Just wondering if you are seeing increased competition with Cisco, given that they have Sourcefire now and that's integrated into the business? Does that does that change how you feel about Cisco from a competitive perspective?

We have not seen much in the last couple quarters, after Cisco acquired Sourcefire. I think first, we're not quite directly competing with Sourcefire before the acquisition. And after that, we also don't quite see much change so far.

On the other side, like Drew mentioned, 70% to 80% of enterprise customer really looking for the network security, they want to buy from a non-networking company. They try to be different, deployment compared with the traditional network deployment they have there. So that's also making a lot of enterprise, they need to have a different vendor, different layer protection.

Melissa, this is Drew. Just a little bit of color on it, I think the point of that is that best-of-breed matters in this space, clearly in security, and we obviously have a network performance advantage vis-a-vis our competitors, and certainly Cisco.

Longer-term, when best-of-breed matters, their ability to focus on security when they are doing all these other things, we believe that's potentially an advantage for us, generally becomes diluted focus over time, when a big company acquires a small company. So I'm sure they're doing fine so far, but I guess we'll see how it goes longer-term.

Also from the function part, if you look like more than 20 years ago when we started what are called are UTM, or multi-function firewall, which already have the intrusion integrated long time ago, more than 10 years ago. So far after Cisco acquired Sourcefire, probably they put a lot of effort trying to integrate that into their existing firewall, but we have not seen much in the field yet.

Okay. That's helpful. That's it for me. Thank you.

Jonathan Ho, William Blair.

I know you went through this on the call, but I wanted to make sure I understood the accounting around the agreement not to sue from Palo Alto. Can you maybe walk through that transaction, just help us understand exactly where the different pieces fit? And then we'll go from there.

Sure, Jonathan. You're asking -- I think you're asking about the accounting question, how we're treating it from an accounting perspective?

That's right. The $20 million shows up in cash flow, but then where does it show up on the balance sheet? What amount this quarter, is it amortized?

Yes. I'm not staring at the balance sheet, but there's going to be basically a years' worth -- think of it this way. We're amortizing it over the period, the six-year period, evenly.

So much what is it, 72 months in six years, if I get my math right here? Whatever $20 million is divided by 72. That would be how much per month that will go as will be contra expense, contra OpEx.

And then we basically we pay them, so we obviously credit -- they pay us, excuse me, so we get cash. If I get that right. So the cash appears on our balance sheet, and then the other side of the ledger is appearing on what you would call accrued liabilities, I'm sure it shows up in accrued liabilities.

And there's some current portion of that, and then some long-term portion of it. The current portion would be one year's worth, at any given time.

So I think that was roughly -- I think I saw the number when I was reviewing it, roughly $3 million or slightly north. Something like that.

20 divided by six years would be about $3 million? Yes. About $3.5 million.

This is really the continued after -- since the market like three years ago, there's I think it's a $6 million payment. And then also looking in the last few years, and I will not say every quarter, but every year, we have a multiple like IP kind of income. Because we have very strong, and we're innovating in the space, and that's got a lot of technology and the pioneer relatively early in this area, which some other companies tried to follow. So that gives us very strong position on our APM technology.

Got it. Got it. That's it for me in terms of questions. Thank you.

Erik Suppiger, JMP.

Two quick questions. One, just on the LTE opportunity. You said that you're not seeing much of that at this point. Do you think that you're going to start seeing some of those deployments this year, or is that beyond 2014?

I think first what I see is really I see the first tier, also probably more in US, they're not quite as deployment yet, but if they do accelerate a lot of testing. And we do see the second tier, third tier and also some other international region started doing early-stage. Because if you try to do the cloud, the mobile solution especially securities mobile device, it probably goes to the carrier, the LTE, that's pretty much only the variable solution today.

So that's where -- also we see some other vendor, they initiated called a clean pipe, and that's where we try to help the enterprise to add additional layer of protection from the infrastructure side. That's also starting, I think probably some European country a little bit ahead, compared to some US first tier carriers right now.

Is there testing at this point, first tier carriers? Does that mean that you would expect deployments within the year, within 2014?

There's testing going on but I cannot give the timing of when we would be -- land a deal.

It's really up to them, Eric. It's really not up to us. The testing, it's thorough testing, it will take some time, and there's other things that they are considering, that you have to consider too, so it's really up to them.

But all I have to say we are way ahead of any other player because the speed, the technology, the broader function we offer, and also I think pretty much all of already existing customers which we have experienced overall solution in the last 10 years. So we are in a much better position compared to any other one try to get in there.

Okay. Second question. Just to be clear, your free cash flow was $50 million.

If we adjust for the Palo Alto payment, it would be around $30 million. Do you expect that your adjusted for that Palo Alto payment, do you think that your free cash flow going forward would still be in the second quarter, would still be below what you were in, in the year ago quarter, for the June quarter? Or is there anything going on in June that would cause cash flow to pick up from those levels?

I think CapEx is going to be slightly higher. Let me -- give me just a second here. Yes.

So we're going to have about $5 million of cash CapEx over the prior year. I think we obviously had a strong collections quarter in Q1. And so I believe we have -- I believe ARR is lower, $20 million going into the quarter versus the end of December. And I would -- you would probably want to take that into account, or certainly this change year-on-year.

We had very good DSO. I think it was 59 days versus 66 last year. And so that's going to be $10 million, $12 million of collections benefit in Q1 versus Q2.

Okay. So sounds like --

You definitely have some downdraft -- you've got a couple of downdrafts there.

So it's going to be considered down significantly from the year-ago quarter, the June quarter of 2014? Is it a presumption that you're going to be, for the year, probably below last year's free cash flow levels?

Again, we're not giving annual guidance, but I think I said earlier I gave somebody, that we'd spend -- I think the DSO thing hopefully you can figure out how that -- in your model, how you believe that normalizes year-on-year. Clearly there's a collections delta in Q1 versus Q2.

Potentially, that normalizes by the end of Q4, right? It's just hard to call Q4 linearity at this point. But over and above the typical rate, I think over and above, we had said about $15 million of CapEx this year versus last year.

Additional CapEx versus last year?

Versus last year, that would be the big thing, and then obviously you had Pan, and you had the tax payment. Those roughly offset.

Okay. Very good. Thank you.

Nandan Amladi, Deutsche Bank.

It's Kazim here for Nandan. I have a couple of questions.

One is on the NP6 processor. You mentioned that it's included in the new high-end appliances you released last year. When was this chip released, and how many appliances today field the NP6 chip?

I think we announced the 3700D in October, end of October last year, and then the 1500D probably end of December. And then shipments started around Q1. And then there's a few others, go to the five cell, the three cell, and in the middle range or in the process to be upgraded later this year.

I think I mentioned NP6 compared to NP4, on chip level performance we see 5X to 10X, depending on function improvement. And we have a three chip family, the network processor is more towards high-end and middle range and then the system chip is more towards the low end, which we refreshed the product last year.

And also, what we call the content processor, that's pretty much the coprocessor, pretty much next to the CPU. That's where pretty much apply to all the products, both the low end, middle, and the high-end that's the one like early question we mentioned about, we do use the latest CPU but it is kind of a coprocessor. And also network processor gave us much more additional performance boost, compared to what a CPU can do, because one function moved to this chip, they can be more than 10X performance again there.

So that's where the processor still going on, so every quarter we may have a couple products we announced using the new chip. But as a kind of strategy so far really, when we're announcing, we want to make sure we have all the inventory, we have a lot of planning, other things going on.

The timing we tried to make sure once we announce, when customer need it, we have the product available. So that's the strategy we have.

Right now you have two appliances the 3700 and 1500 with the new chip, as --?

Yes. There's more coming.

How long does it take for you to like, put that in all your products, when you have a new chip released? How long is the rollout of that new chip to the entire appliance range?

It's usually about three years, I think, to go through our whole product line, based on ones that use ASICs.

Each chip take about two to three years, and then doing that process, we gradually update. Sometimes because better CPUs, but really also the chip will help a lot.

It also depends on the timing. Really we try to manage the release, not everything come up in one or two quarter. Just kind of during the process, kind of gradually apply it to the product line instead of try to rush out in the first few quarters.

Do you typically see any kind of refresher product upgrade after you introduce a new chip into the appliances?

Put it this way -- we're starting to see more opportunity, which leveraged higher performance. That's where we mentioned the 3700D is the first one in the industry which has the 40-gig interface in appliance.

So we announced October last year, so we see a lot of opportunity, which because so far, none of the other competitors or software product can deliver this warp speed 40 gig performance in a single interface. And we also mentioned like, I think this year would be the opportunity to upgrade to the 40 gig to the 100 gig interface, so a lot of us are starting testing right now. But it's not announced yet.

Got it. And just one follow-up on the accounting change that you had this quarter. So just to be clear, there's no change in the revenue recognitions, your revenues will remain the same as you had previously. You're seeing product income go up slightly because now you're including the gain on sale also, when you calculate the operating income?

Yes. So again the only accounting change really was non-GAAP. And yes, the way the prior -- it was $6 million over three years that was contra, they removed that from contra expense, so in other words, they took the favorable impact out of OpEx.

And so going forward, again it's now recurring, it's a recurring number. And it's insignificant per quarter so that will be a contra expense item in OpEx, contra OpEx going forward. Again as we discussed, divide the $20 million divided by 72, times however many months in the quarter, that being -- the reason I say that is because we didn't have any benefit in January. It just was two months this quarter, because we started this in February.

And then the other change was we collapsed the ratable line, I think -- I should just point out why we're doing this. We collapsed the ratable line into the services and other line. But that has no impact on margins, revenue, anything, there is no revenue, to be clear, there is no revenue recognition change.

But so the operating income is going up, what I want clarify is your OpEx changing or your operating income is going up, because now you have higher revenues for your non-GAAP operating income? So when I look at the sheet that Michelle just sent around, the operating income for the previous quarters is slightly higher in the new presentation?

On constant revenue, you would have a slightly -- I think it's about 30 BIPs due to this change. It was about $600,000 a quarter, I believe, that was contra expensed. That's it. That's the only change on a retrospective basis.

So the OpEx goes down, right? The non-GAAP OpEx goes down?

That's correct. No impact to revenue whatsoever.

Got it. Okay.

Just one final one, I think Ken mentioned on the last call that your exposure to certain verticals is lower in terms of sales capacity? What would be -- what verticals would you classify as the ones that you think you need to increase your sales capacity?

Can you repeat the question please?

In the last call I think Ken mentioned that some of verticals you're not as staffed up in terms of sales capacity. What verticals would those be?

I think Drew mentioned about like finance service, healthcare, that's the two big industries, especially in US.

I think the way we answered the question was, just logically, we are not really trying to define our verticals on the call yet. But logically, if you look at where the big buckets of IT spend are, you're going to pretty quickly go to healthcare and financial services. And that would be -- assuming we take a logical approach, that's where we'd be looking.

Okay. Got it. Perfect.

That's all for me. Thank you.

Robert Breza, Sterne.

Just a real quick simple question. As you think about the linearity of the year and you talk about growing at 2X times the growth rate, anything you see from just a seasonal perspective? How we should think about our models for either -- obviously second-half, given that you gave us the June quarter? Thanks.

Sure, Richard. Again, we're going to have to stick with what we just said, which is we're looking to grow at 2X or north of the IDC Gartner benchmark of 6% to 7%. Really can't respond any more than that.

Thanks.

Sterling Auty, JPMorgan.

I have been bouncing around between calls as well, so sorry if this is a repeat. But the billings growth in Asia, I think you mentioned a tough compare, but you mentioned Japan was good, China was good. Were there any areas in Asia that lagged behind, and maybe what's happening there?

Sterling, repeat the question one more time?

He was asking the question, Sterling was basically saying he heard that China was good, --

We don't break out anything more in terms of percentage of contributing to business other than we're trying to give color of what regions did well. And in Asia, clearly Japan, what we have said is Japan is the biggest, most significant contributor to that region. China's very small in the overall part of APAC for us.

I would just share, Sterling, we had a bit of organizational change over there. Not going to mention countries and so forth, but there probably was a little bit of impact of that from that as well.

Okay. And then Ken, technology wise, Heartbleed has been obviously in front of everybody's minds.

Did you have any exposure in terms of the exploit in the products, number one? And number two, did perhaps that open up the door for you, especially seeing as how Cisco and a couple others did have some exposure? Did you that give you some opportunity to go talk to customers about your solutions?

I think our solutions can protect that one long time ago, and also we have like a special protect all the data center, all of the e-commerce all this area. So I think combined, the FortiGate and some other product we have, that we call complementary to FortiGate like a FortiWeb DDOS, is really a very nice deployment solution. And that's actually starting to raise more interest in how we can protect their data center or e-commerce center there.

So that's where, since awareness of sometimes quite dangerous to do a lot of -- whether e-commerce whatever, or the Internet so you need to have better protection in our business. I think we see some more interest in this area. And also the products up in there, even before the news come out, has been protecting all the customer already, so that's where we see, so far the customer base is happy with our solution.

Okay. Thank you.

(Operator Instructions)

Michael Turits, Raymond James.

Just a follow-up on the cash taxes issue. I think you said $37 million in cash taxes for 2014, after you paid just $25 million last year.

Can you just remind why that's accelerated payment this year? And what it might be in 2015? I'd hate to run that out.

Well, it was a true-up, Michael. Obviously, if we had a better number going forward, we would be accruing it now. We would be -- you would have to build it into your GAAP or non-GAAP tax rate, your non-GAAP, actually, over time.

So this was just a true-up from things from the past. And we assume the best that is we do our taxes correctly, and we don't have true-ups going forward.

So I wouldn't predict that -- you wouldn't obviously be wise to predict that there's some tax true-up going forward. Although look, quite frankly, there's always something. I'm sure if we were looking in our past, but I would expect that in the normalized rate and that we would get back to that going forward.

So should I think after this kind of like 35% GAAP effective tax rate and the same thing on cash taxes, 35% GAAP?

I think -- again, we're not giving long-term guidance, but I think 33% is the right rate to use for non-GAAP. And then the cash rate for the year, I don't really -- it would be hard to give you the cash -- the actual cash tax rate, but I can give you the number, which is the 37%.

I'm trying to figure out going forward after that, what's a safe way to model 2015 and afterwards? I know you can't give guidance but -- ?

I haven't thought it out that far, Michael in terms of how that would look. You'd have to look at it pretax income, and I haven't modeled that out. I would use -- just to play it safe, I would use the 33%.

That's what I want to do. Okay. Thank you very much.

Thank you. I'm showing no further questions at this time.

Thank you, everybody, for joining both of our calls. Appreciate it. If you have any further questions, feel free to reach out, and happy to answer anything else we can. Thank you very much.

Ladies and gentlemen, this does conclude your conference. You may now disconnect, and have a great day.