Fortinet Inc (FTNT) 2016 Q1 法說會逐字稿

Good day, ladies and gentlemen, and welcome to the Fortinet first-quarter 2016 earnings financial analyst Q&A call.

(Operator Instructions)

As a reminder, today's program is being recorded. I would now like to introduce your host for today's program, Michelle Spolver. Please go ahead.

Thanks, Jonathan. Hi everyone again, thanks for calling in. I have Ken Xie and Drew Del Matto with me here.

It's a full Q&A session, but before we launch into the first question, just a reminder that the safe harbor and cautionary statements that we made in our prior call affect any content we deliver during this call as well. With that, Jonathan, you can take the first question.

Jeremy Benatar, Raymond James.

A quick housekeeping question. Last quarter, you guided to cash taxes between $25 million and $29 million for the year. Any change there?

No change there. $25 million to $29 million is right.

Got it. That's it for me.

Thank you. Gregg Moskowitz, Cowen and Company.

With the [cue-in] out of the way, are there any potential transition issues that you would say our associator could be associated with the unified global sales model that we should be aware of, or is this pretty much in the rear-view mirror as far as you are concerned?

We call it out. The changes have been made. Somebody asked if we had some new people still to come in. We're always going to have some new people come in, and there may be a few spaces to fill. But the structures are in place, the alignment is there.

Gregg, the big caution is that there is some remapping that occurred in the rear-view mirror. But when you transition, there's always six to nine months of a sales cycle, maybe longer in some cases, that we need to be cautious about, and we reflected that in our guidance.

Exactly, perfect. Ken, last quarter, you said there were some larger enterprises that were taking a little longer to make decisions. Has that dynamic continued, or has that changed one way or another?

Remember our commentary last time. I think what we were talking about is that, because security was becoming more strategic and there were more levels involved, I think because people were talking about emergency spending and that it's shortening the deal cycle, the deal time.

We were talking about that in some cases, it might elongate it because it's gone up to a Board level for approval and its customers are taking more of a strategic look to security. If I remember correctly, I think that's the context of the statement last quarter. I don't think that we've seen a change.

Okay, great. Just a couple of other quick ones. We've heard that you may be looking to make some changes to your website potentially. Can you speak to that and what you're aiming for there?

Wow.

Very good question, absolutely the case. We did change the website, and we launched -- we are going to, forthcoming. We've also updated our Partner Portal, which I believe has launched.

It's part of the awareness-building and the plan in terms of increasing awareness and building awareness. In addition to fresh content and a fresh look and feel for the website, there's a lot that's enabled in the website to help drive demand through better search engine optimization and things like that. I think it's all designed to help drive better demand for Fortinet.

It's more helping the lead gen and also more interactive, and also target -- customized for some of the regional customer which may have a different need. So that's where we feel we will be helping the sales a lot.

Okay, that's really helpful. And just lastly, getting back to the CMO search, I know you've classified your marketing team as being strong as it stands today. But what type of person are you looking for in terms of bringing in, and what would be a reasonable timeline in your view?

First of all, we are saying global marketing lead. We haven't given it a title, Gregg, but global marketing lead. And the time frame, it's hard to predict. It usually takes anywhere between four to seven months to recruit someone. We will be very careful and cautious about who we recruit.

Ideally it's someone with certainly knowledge of the security space. We want someone who can help enable our partners and drive better market awareness and build that go to market machine that would really help us drive the growth. And somebody with experience doing that and a proven track record of doing that is the type of person we are looking for.

We also need to line up what sells better, because that's where sales and marketing need to be closely working together to help grow the Company.

Okay, terrific. Thanks very much, guys.

Thank you. Saket Kalia, Barclays Capital.

Thank you for taking my follow-ups here. First, maybe a dumb question, but can you dig one level deeper on how a customer would use security fabric? Is this a separately billable product?

And realizing it may be early, but any idea how it might be priced whether it's a perpetual license or subscription? Any color on just how it's used and pricing.

It's all of the above. There is not a single product. The fabric ties together everything from the expanding tax surface basically, where you now have IOT, cloud, more apps, more use of data.

And so you could sell -- it's just the breadth of the portfolio you're looking to sell at the end of the day, from a Wi-Fi product into the FortiGate class into controllers. And it could be virtual or physical at the end of the day, and we fill the subscriptions as well.

We also think it's something that ought to be interesting to our partners because it's a broader concept, thought leadership to sell to customers. And we are calling it -- it's an open architecture, so the more people and technology partners we bring to the table also helps the partners sell more product as well.

That's a good way to think about it. So it's really more products, and that's what we're trying to do. In some cases, that could be upsell, cross sell. In some cases, it could be a new customer who's hopefully a larger PO coming in the door, a larger initial PO with more products on it.

It's more trying to help customer -- most enterprise customer to ease and also help them manage this better, because otherwise there are many different solutions from different vendors. And also, you talk about pricing as an area we keep improving, like have some enterprise-level pricing model compared to an individual product pricing model, which we are keeping improving on that. That is basically helping the enterprise customers do easy management, easy pricing and most secure solution working together.

Okay, got it. For my follow-up, Drew, you talked about in the Q&A session from last call the different subscription bundles, whether it's UTM or enterprise. And again, realizing it's early on the enterprise bundle, but are you seeing any change in attach rates just in early days for high-end appliances, maybe opting for the enterprise bundle versus UTM?

The enterprise bundle just started early this year. It's still relatively early. Also, the enterprise depends on a new FortiOS 5.4, which also just released this year. And so it's still early to say what's the attach rate yet.

I would characterize it as both an initiative. As Ken said, the 5.4 release is helpful, and then obviously just an opportunity for us going forward.

Got it, very helpful. Thanks guys.

Walter Pritchard, Citi.

Thanks. I'm wondering, Drew, if you could just help us through the specifics on -- you said that the way you're accounting for the price increase on renewals impacted the split between long-term and short-term deferred. I didn't totally understand that.

It would seem like what would be deferred, whether it's recognizing one year beyond the year, would both be impacted by the price increasing, and you wouldn't necessarily see some allocation shift. And it does look like whether you're looking quarter to quarter or year over year, your long-term did grow quite significantly more than your short-term deferred. That's the context of the question.

It's been growing about a point a quarter is the concept. My issue was, if you amortize something at a lower cost and it comes off at lower cost and goes back and renews at a higher cost, you would have more long-term deferred.

Okay, maybe we will send you some numbers and try to follow up on that one. Just on the segments, I know you're not breaking out Meru, but can you tell us, are you putting Meru into the low end, the mid range or the high end in terms of product segmentation? Or is that excluded from the product segmentation?

It's excluded from the product segmentation because the product breakouts we give now are FortiGate billing.

Okay, got it. And then lastly, just on margins, you talked about 20% margin I think a number of calls ago as a medium-term goal. Is there any -- I assume there is no change in that, but we didn't hear you talk much about that on the call. Where are you still on that medium-term goal, and any finer time frame around it?

That's the one we talk about by 2020, we try to reach 20% operating margin. We plan on improving 1% per year going forward. We still have that target.

Great, thank you very much.

Gabriela Borges, Goldman Sachs.

Just a quick one on FX at some of your international customers. I know you're priced in US dollars, but have you seen any impact from purchasing power at some of your customers overseas, whether that's last year into this year maybe stabilizing a little bit?

We haven't seen anything. We haven't seen an impact. Again, we had I think a very nice margin on the product revenue side this quarter, which we didn't feel like we saw a lot of pricing pressure, quite frankly.

Also, we tend to be competitive, both on the function pricing and the performance. So we don't see any pressure.

That's helpful. And then just on the marketing investments, I'm curious as we think about the run rate of investments through the year. When you get a CMO, I imagine it will be a little bit of a step up on the margin, but aside from that, is there anything big in terms of step ups in the marketing that we should expect through the year?

There's nothing we should call out specifically. We've guided for the year. It's all in that envelope.

Okay, thank you.

Brent Thill, UBS.

This is Fatima on for Brent. I just want to dig into the traction you are seeing in APAC. This business according to my numbers has now accelerated its growth from a revenue perspective for four consecutive quarters. So any help you can give me around what the market dynamics are.

Are they just behind from a refresh perspective? Are you pricing in JPY there? Was there some FX tailwind given what the yen has done. Any color there to highlight.

We bill in dollars. We believe -- it's just that we've not invested as dramatically there as we had elsewhere. And we were pushing the teams to do that, and the investments paid off. That scenario, we continue to look to invest more, quite frankly.

We tend to be very strong with healthy and broad partnership over there. And we invest in the sales capacity and the marketing there. So once we have some investment there, we can see the growth accelerate.

Fair enough. Just shifting gears from a market perspective, you've seen a tremendous amount of success in the enterprise and the Fortune 1000 -- that's certainly a billing business. But you also have a stranglehold on the low end of the market.

I'm wondering if there's more room for share gains at the lower end where you had played more meaningfully historically. And to what extent is your managed security services business with service providers cannibalistic to the very low end of the market?

The low end we still keep growing nicely. Right now we are number one in the number of units shipped that deployed globally more than any other vendor. On the low end, we've done some research.

We have probably [20]%-some of market share, and with the technology we have since now shipped, we have a huge advantage compared to other competitors who have to load into a small PCL server to compete. We can put everything in a single chip. That's where we feel we keep gaining share

We will be about the same rate with growth overall. We also have the high-end care service provider with the Fortune 500 company. The area we under invest is the mid enterprise market side. That's where we study the sales capacity and marketing study and invest in that area.

It's very US, and that's also the global sales structure -- the additional marketing investment will also help drive the mid enterprise growth. That's where we have a pretty low percentage compared to the low and high end. We feel the investment will help grow in the mid market.

That's very helpful. And last one for me, if I could sneak one in, you talked about the momentum you're seeing with the FortiSandbox. It tripled two quarters ago, then it quadrupled and grew almost 300% year on year from a billings perspective this quarter.

Can you help us frame what type of penetration you are seeing with respect to your 270,000 customers? And what's the predominant method of adoption? Is it more the bundled service, or is it the appliance. Any color on those would be helpful. Thank you.

It's more like fabric, security fabric selling together, and that's where we're working with the FortiGate, FortiMail or FortiWeb together with FortiSandbox is the solution. That's where there's a few relatively small customer base because most Sandbox buyers lean toward the big enterprise, high-end customer.

But we also started offer the ATP into the FortiGate, the 5.4 FortiOS and also the enterprise bundle. That eventually will help drive the ATP solution to the more broad market. At the same time, the new FortiGate we announced that six others were selling will also help.

Sandbar is one example of security fabric working together, and that's driven the upsell/cross sell. That's where we see not only the Sandbox, but some other solutions like an email, FortiMail, FortiWeb, some other also starting to ramp up quicker than the overall company growth. So that will help drive the upsell/cross sell.

Melissa Gorham, Morgan Stanley.

I wanted to ask you about use of cash, and apologies if this has been asked because I was hopping back and forth. You bought back some shares this quarter, and you've been somewhat acquisitive but not unusually acquisitive in the past. I'm wondering how you think about the opportunity of continued share buybacks versus consolidating further security functionality through acquisitions.

Melissa, we haven't been asked that question today, which is unusual. The strategy remains the same. We continue to invest for growth, primarily organic. We're very good at driving organic growth as reflected in our security fabric we announced.

You're right, we've done really what have been I would say tuck-in acquisitions, and I would characterize Meru in some ways more as an acquire in terms of getting a pretty big R&D team with Wi-Fi experience, with wireless experience, and the idea to inject security. We're looking at opportunities that make sense. If you think about the fabric, consolidation makes a lot of sense. The customers are looking to consolidate.

We've heard customers basically complain about there are too many security vendors in their environment. They are looking to consolidate. We hope to do that with the security fabric, make that an open architecture as well and provide our ability to have a common operating system all the way from Wi-Fi into the data center and doing it at a high-level of performance and security.

So clearly we're looking at opportunities there. And then from a buyback perspective, I think you can look at our history of where we're buying back. We talk about there's $150 million remaining of the $200 million, but we will stick to the guidance on that. We tend to be very opportunistic when we buy back and to try to buy back more in the price ranges we bought back historically.

Okay, that's helpful. That's it for me.

Sterling Auty, JPMorgan.

I've been hopping around the slide. Two technology questions that hopefully haven't been beat on. When you look at both the ninth generation chip that you've introduced, especially in the Fortinet 6000 series, one of the speed differentials you've had has been the 48 gigabit backplane, the ability to do it without core aggregation, et cetera.

What have you done in the new generation of appliances, especially in the 6000, to take that maybe to a new level? Or are you still at the same performance level on the back plane as the previous generation?

The backplane definitely is a few times faster, but most importantly, we want to drive the computing power for the next-gen firewall function, which will involve prevention, malware, anti-virus and the other SSL level encryption. So that's the big need and change, because the traditional firewall function we take care on the network process level. That's where we have the firewall VPN.

We don't have a huge performance advantage, but go to the content side, that's where the intrusion, the malware virus and the web content, that's where the CP9 processor and also the new 6000 architecture may double, triple than the previous 3000/5000 platform on the single blade side. And that eventually will also have -- the backplane speed will also help a lot.

So that's where the advantage we will have both leverage the system-level architecture, which has much better improvement then the past in the underside, and also the computing power and also the CP9. This eventually will help with security from the current firewall VPN traditional connection level security to the content application level, which the common processor that the 6000 we will lead in the space.

Separately, when you look at the security fabric, is that something that when you're trying all the things together as you showed on your slide, is this an indication of other areas that you plan on expanding into with your own FortiNet solution, whether I think about end point or other? Or how much are you going to rely on the partnerships and the echo system to provide that tie-in functionality in some of those spaces?

We keep developing both internally. We have a pretty strong engineering team here, but we also keep open for the third-party partner, including both from the endpoint.

Even some solution we have, we also have one open for the partner because the management cost is huge than the product cost in most enterprise environments. That's where we're making all of the different solutions working together.

I think we try to leverage -- we are the number one most deployed network security solution there. And also adding the Wi-Fi and some endpoint and some application like email web and the DDoS and the ADC will also help to get all of this together. And that's where the fabric is really open for both internal and also for the partner to help manage together to reduce the management costs.

Got it, thank you.

Catharine Trebnick, Dougherty.

On your geographic reach, is there any way -- you just talked in the last call that in the United States, roughly 65% of the service MSSP is related to the service provider business. I'm trying to get a handle geographically, like in your EMEA or Asia-Pac, how much of that revenue is service provider related and how much of that might be related to the MSSP business?

I don't have a number, but my gut feeling is really in the US, we are leaning more towards the MSSP compared to some international because the US -- like we said in earnings, we have pretty much all of the major carrier service provider customers. But the US is very -- the MSSP also helping drive a lot of enterprise and other service for us. International, some of the service provider is not as large or global as some of the bigger US one, and that's maybe towards -- not quite --

Catherine, let me make sure what I said was gotten correctly. What I said was, within our global service provider business, roughly 65% to 70% of that is for MSSP deployment. I wasn't saying that that was 65% of US business or for American business. It was of our global service provider business.

Okay, thank you for the clarification. Another question is, if I take EMEA, how much of the European business would be related to the service provider versus enterprise?

Yes and see, we don't break up the verticals by region.

I know, I was just trying --

We don't know the answer, we don't break it out anyway. But service provider is 22% of our global billing, and we don't break it out as to how much it is per region.

All right. I thought I would try. Thank you.

Ken Talanian, Evercore ISI.

First one, do you have the deferred breakdown between product and services for this quarter and last?

There's not much -- we don't break it out, Ken, but there's not much product.

Okay. Very small. Okay. And then my second question, just thinking broadly about the competitive environment, you're seeing increased spending from Check Point in sales and marketing. Palo Alto continues to be strong.

As I listened to some of your commentary around your own sales and marketing investments, it sounds like you're larger, there are a couple people you need to add in, you need a global marketing leader. And then at the same time as you look forward, you remain somewhat cautious on the environment.

Is part of that the competitive environment? Do you expect that you will either have to add more product to win a comparative deal, you expect to do more discounting? Is there any of that factored into your expectations?

I'm not sure I understand the question. Ken.

(inaudible) I can answer. You can look at the NSS report. We also assured that [ISHO]. We're much better both on the security function and also the price-performance compared to the other competitors. And also we're very unique with the basic chip and also heavily invested in the systems side.

That's the new product we're keeping driving leading the space. And like internal segmentation, we don't see any other competitor come close to what we have. We're keeping the pace and keeping leading, but we don't see price pressure from a competitor because we're leading that area with the performance the function we have.

We absolutely don't look forward to discounting prices. That's not something we would certainly plan for. It's not part of our strategy.

If you think about it simplistically, it's rate and volume. You want to require customers, which is the volume, and then do it at a higher rate, which is price. And the more we can add to the bundle and charge more, the more we can upsell and cross sell really helps with those economics.

Again, we've been investing and very successful at acquiring new customers. And so that's really how we think about the model.

We go the other way. Several years ago, we reached the bottom price, and now we offer the enterprise bundle, which has two additional new service, the ATP mobile and cloud security. So that's also helping make the per deal per box total solution be higher price.

And I appreciate that. I guess part of the question is also, it's obviously to your advantage to add more richly configured subscriptions to deals, but do you feel that that's a necessity at this point?

A necessity -- I think it's a strategy.

Okay.

Also, we have the technology the product can do that because that's where the new common processor, the new architecture had a lot of computing power and a lot of function which were helping drive the multi-function -- integrated consulted function there, which the other competitor -- they all had the CPU, which will be more and more difficult to consolidate more function into the CPU, which we have to [ASSIC] offload both on the common processor side and on the network processing side.

And then just to close off on the strategy part about it, customers want to consolidate vendors. They're looking to make it simpler to run the business in security. The more that we can add, we're providing more value to the customer. It also makes for ease of deployment, ease of management.

When we look at the overall bundle and the opportunity to enrich the platform, whether we added to the bundle or elsewhere, you may add it on top of it and price it separately. That's really the strategy, but again, go back to rate and volume, we absolutely look for ways to increase the ASP on the bundles if that makes sense. If not, we may just look at doing that separately.

But again, those vectors are there. Consolidation clearly that's out there. People are looking for a platform, security fabric does that. The more we can do to make it easier for them along deployment or management, then the more value we're providing and the more we're entitled to charge.

You can see that's been in the SMB space. We became number one three years ago, but with leading technology, we are probably more than double the number two now.

That's where the advantage we have on the system platform and the system chip solution will helping in driving a lot of additional value compared to competitor and also gaining a share quickly. And that's where we're hoping to expand into the mid enterprise and also the service provider space.

Understood. Thank you very much.

Imtiaz Koujalgi, Deutsche Bank.

You mentioned on the last call that the contract duration had not changed much from previous quarters. What is the duration in this quarter, and what's the rate in the last four quarters?

It tends to range between 18 and 22 months. It's within that range.

It hasn't changed.

Got it. And then you mentioned that the price increase was the main driver for the long-term deferred mix going up. I think I understand -- you've had a few price changes in the last few quarters I think. I just want to walk through which of the price changes you're referring to when you talk about the increase in deferred mix, because you had a change in the plans early last year which I'm assuming should not have a big impact --

It would be on subscription, the deferred piece. It's not product. It would be the subscription.

Right, so I think Ken said the enterprise --

That's supposed to be long-term. There wouldn't be any long-term product. You ship the product.

Right. Then Ken mentioned that the enterprise bundle was very early, so I assume that didn't have much of an impact on the long-term deferred. Is it the UTM bundled price increase you had last year which impacted the long-term deferred mix?

I think we will follow up on the math on this, but the theory is that if you're ruling something off at one price which is lower and renewing it at a higher price, you're building deferred over the last year at a higher long-term deferred at a higher price.

Okay, I will follow up, but I'm just confused why the mix would change. I thought this chart on deferred should impact it equally.

It went up about a point. It's been trending up.

Could you repeat that?

It has been trending up over the last years. And you're building and growing -- we will follow up.

Okay, no worries. On the product revenue side, you beat -- your seasonality on overall revenues was fairly in line with your typical Q4 to Q1, but on product revenues, it seemed a bit lower than what your typical Q4 to Q1 sequential growth is. I remember you had a price increase on the appliances early last year, so why the lower seasonality from Q4 to Q1 on the product line this quarter?

The product we're down -- it's a little more than last year. It always drops from Q4 a little bit. I don't have a better explanation than that.

Again, it could be a function of pricing or how they buy or what they're buying at that point in time. Again, we're more focused quite honestly.

To be fair, again, we're more focused on driving the recurring revenue stream at a higher rate of speed. And so as you tilt toward that, it will influence the balance of that mix more toward the services line.

The product price I don't think has been changed. We just had one product last year. We tried to line up the price with US price would be better.

I don't think we changed anything else. And also, that's not an increase. It's really just off the global price typically better.

There is no change in product mix last quarter or last year, which drove the seasonality lower than typical?

I think we're just more focused on driving recurring revenue streams. You've seen that manifest itself, and the services line slowed up.

That was a good uptake. If I were to follow up on that service line, is there a way to quantify the attach rate? Your service line did go up quite a bit seasonally versus last year. Is there a way to quantify what the attach rate was typically and what the attach rate of the services was this quarter, which drove that services line item?

We haven't been sharing attach rates.

We shared -- right now, the renew we track based on the product, which is the mid-70%s. But if we consider the product units survived in the network environment for four or five years, that's where it translates to over 90% (inaudible) of the customer retention rate there. Because even same customer, when they buy the new product, we don't consider it renew. We consider it as new, because have the chain amount, and sometimes it's difficult to track a customer directly.

That's where we do keep the customer retention rate, the mid 70% in the product level. We believe the customer level probably over 90%, in the high 90%s.

I was referring to the subscription to the appliances.

I know we're spending a lot of time, but it's an important question. The attach rate, the reality is that from the beginning, we had roughly 80% of our customers that were buying subscriptions and buying bundles since inception. Whereas other companies have introduced new subscriptions services along the way, we haven't done that with the exception of the enterprise bundle that was just made available in Q4.

So outside of that, we've been enriching with features of the existing bundles, but it's hard to talk about attach rate trend when from the very beginning, the vast majority of our customers were buying the bundles. I wanted to explain that.

That's helpful. The reason I was trying to build on that is because you had a price increase in the UTM bundle. I was trying to get a sense of how much of the increased service revenue is coming from the increase in attach versus increase in the pricing you had.

No, I understand. It's almost all from the increase in pricing. We're starting off with a big base of customers, and the buying trend for Fortinet customers has always been to buy it with -- unless you're using it for a straight firewall, the trend has always been to buy the subscription.

You're not going to see a trend line go up like you have with other companies because they started with nothing and are introducing subscriptions along the way. The growth in services was almost all driven through the price increases.

Got it, that's very helpful. Thank you.

Erik Suppiger, JMP Securities.

I had a couple on the cash flows. Looks like you're going to be spending about $25 million more over the course of the year on CapEx. Can we assume that's a linear $8 million per quarter?

We guided $55 million. I'm not sure what the $25 million is, but we guided $55 million.

You did $30 million in the quarter.

For the rest of the year, yes, so $55 million. We don't really -- it's hard to guide on that. It depends on the timing of a few things. We're still buying some real estate deals. It's really hard to predict the timing of those.

Okay. Anything to think about from the seasonality in terms of your free cash flows for the June quarter or how we want to model that going forward for the next three quarters?

I don't have anything to add on that.

Okay, then lastly, deal size. You said the deal size for some of these internal segmenting opportunities, the firewall segmenting opportunities is in the six-figure type range.

Is that because this is still very early and maybe over the course of this year, we could start seeing some seven-figure size deals in the internal firewalling? Or is that not in the nature of that market opportunity?

It depends on how big the customer's network is. If you're talking the largest -- a huge, huge investment bank, the largest network in the world, they probably do more segmentation.

There's more areas to segment, they would use more firewalls. If you're talking to midsize enterprise, the network isn't so broad that they would put that many in.

So I don't know that it's that the deal size will grow. It's not really an issue of that unless -- as we start getting massive amounts of customers, which we are starting to do that, you could, but that's not really the way you would look at it. I think you would look at the growth coming from additional customers or expansion opportunities.

In a given account, can you give us a sense for the opportunity for traditional perimeter firewall buildout versus an internal segmenting, how you might look at that?

Are you asking -- I don't understand what you're asking. You're asking from a dollar perspective?

What would be the question, is the internal segmenting a quarter the size of the perimeter deployment, or how might we think about that?

Again, it depends on the customer. A large investment bank would probably be different than a midsize company. I mean if you want, I can do a follow-up call with you.

Maybe what I can do is dig out, and I don't have it in front of me, but give you a real-world example of a deal. Because we have talked about expansion deals, one that started as an internal firewall and expanded as added core to a traditional perimeter or core firewall. I could probably give that to you. I don't have that at my fingertips right now.

It all totally depends on the type of customer. If you have a very expansive huge network that you will do a lot of segmentation, you'd need a lot of internal firewall.

If it's a midsize company that you want to protect your finance server and your R&D server, you need two. Just so you understand, there's a large enterprise, traditional enterprise, midsize enterprise, and they're all going to be different.

We see some early research based on the switch market, which is about $40 billion, $50 billion. Some portion of that -- there were some studies with NPRs -- that has some segmentation, whether it will go to the internal segmentation firewall. But internal segmentation firewall offer much more than the switch level, whether the VPN or some other.

They also cover the inclusion, provision, the content and the application. So there's a much richer function, higher level spending compared to what the switch can do. And that's where -- there's some research. A few billion dollar would drive 6% to 8% additional growth, but it's still early stage.

It's sometimes dependent on vertical. Like Drew mentioned, sometimes the finance service industry, the healthcare industry and some other big enterprises may own an early adopter of the internal segmentation because the switch, obviously (inaudible) it's not quite a secure working for all this vertical space. But they see the huge value, but on the other side on the per-gig throughput, network security is still almost 100 times expensive than a switch.

That's where it's difficult to use in the switch market because network security, you don't have much higher huge computing power to process the same throughput of traffic compared to the switch. That's where we're still in the early stage, but we see as a huge opportunity.

Very good, thank you.

Rohit Chopra, Buckingham Research.

I want to come back to the last question. Are you saying that the internal segmentation firewall is or can replace the traditional switching? It's just expensive now, but as the price comes down, you could actually see yourself taking share in the traditional switching market?

I think if deployed and where the switch deployed and also together with the switch, because it's different than the traditional deployment, which I only deploy on the one interface because we're a company connect with internet. If it's deploying internally, which are all part of replacing some of the switch. But switch is still much lower cost, more popular compared to the internal segmentation.

You have the same deploy model, the function of the switch, but additional security function which the switch networking gear cannot cover. But on the pricing side, is per 10 gig still like a almost hundred times more expensive than the switch.

Correct me if I'm wrong. Segmentation isn't necessarily new. It's just that the functionality is different than older segmentation as we know it, like port-based or whatever people were doing before.

It wasn't very flexible, and now on an application basis, it's much easier. And what you're really trying to do with segmentation is trying to minimize any issues that could happen if there were some kind of threat. Is that a good way of thinking?

Yes, it's a little bit different than the traditional NPR segmentation, which they separate as different groups cannot connect together. They keep the connection but control the higher-layer intrusion malware virus content, which the switch cannot distinguish. That's where the internal security segmentation really keeping the switch level connection there and at the same time screen the content -- the malware intrusion.

Okay. My last question is on HP. I think there was an announcement with HP back in March. Any progress there with HP and any integration? I think there was some integration activity going on.

We do partner with quite a lot of companies including HP, both on the network side an on the access and even the endpoint side. That's where the security fabric -- I think in the security fabric will be helping accelerate partnership including the partnership with HP.

Thank you, okay. That's helpful, thank you. I appreciate it.

Hendi Susanto, Gabelli & Company.

Ken, you sound excited about security application for IOT. Some may argue that IOT is still in the early stage. In your view, what IOT application swill be the early adopters of your FortiNet security solution, and how should we expect the trajectory of the IOT security opportunity?

I think IOT would be the trend in the next 5 to 10 years. It's still relatively early, including the (inaudible) connect our appliance (inaudible). But we do see the security need. And also we developed the technology from the chip level to the cloud software and the service provider infrastructure level to address that.

There's also some research come from some other research firm, say the segmentation for the IOT would be the way for the solution, which we agree. But also, IOT has the -- basically the IOT security is a little different than the traditional user human level internet security. That's where we bundle like a invest early and also make sure we can have a best solution with the market rampup.

So is it fair to say it's not a 2016 or 2017 opportunity in terms of its size?

I think the point here is that vision matters to customers who are thinking about the future. There is some IOT out there.

I think that the large -- the ubiquity of it is probably a few years out, but customers addressing architectures today need to consider it. And so one, it helps to get something to talk about. You probably don't get near-term deals specifically from that, but you are selling an architecture longer-term that this a piece of, and customers know that's something you can help them be ready for over the longer term.

We already mentioned we had a competitor like (inaudible) another chip in that can be in the kindness [sense] there. And also recognize solution for the outdoor [for some other], and also the Wi-Fi and the access point.

So that's a lot relates to the IOT and the future of IOT. But we do want to keep it invested. We see that as a huge opportunity. We want to do some investment and capture the huge opportunity going forward.

That's helpful, thank you.

Thank you.

This does conclude the question-and-answer session of today's program. I would like to hand the program back to Michelle Spolver for any further remarks.

I want to say thank you again everybody for joining the call and asking some good questions. We owe you some follow-up information. I will get that to you, and feel free to reach out to me if you have further questions. Thanks a lot.

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