Fortinet Inc (FTNT) 2016 Q4 法說會逐字稿

Good day, ladies and gentlemen, and welcome to the Fortinet Q4 2016 earnings financial analysts' Q&A conference call.

(Operator Instructions)

As a reminder this conference is being recorded. I would now like to introduce your host for today's conference Miss Kelly Blough, Vice President of Investor Relations.

Hello everyone and thanks for calling back in again. Ken, Drew and I are here for this Q&A session. Before we take questions I just want to remind you all that the disclaimers and cautionary language I went through in the earlier call today still apply. With that we can begin the Q&A.

(Operator Instructions)

Our first question comes from the line of Walter Pritchard, Citigroup.

Hello, can you hear me okay?

Yes.

Drew I'm wondering if you could walk through on [KF-606] just help us understand if those companies are in the process of evaluating this impact. Especially from the perspective of functionality to date that is subscription that could change in terms of the way it is recognized. Is there anything to talk about their that is meaningful?

Well, no Walter. I don't have any -- we haven't realized any alarms on that front yet. Its still early. We're still in the evaluation phase to be honest. And it just -- its really to early to talk about it but hopefully the subscriptions flow holds with what it is right now.

Okay. And then just as you look at 2017 and one thing you called out is more stability in the sales force and the ramp of reps and so forth. Is there any way to think about just as those reps -- as you get more tenure and stable reps, how that potentially contributes to growth in 2017 independent of other factors?

Well, I can only analogize and there's a correlation between tenure and productivity. And I think qualitatively that's just the fact they get to know the customer base better. They get more familiar with the products. They grow the install base, they get better upsell and cross-sell.

We believe that we're -- the -- augmenting that with a stronger go-to-market engine. Where we continue to invest in marketing as we talked about [Legion] -- things like marketing Legion. And training ultimately not only help build the funnel but help them just drive a cleaner funnel if you will overtime. And those really are the dimensions to the go-to-market engine.

Great. And the last question.

As it relates to looking at duration of subscription, and how you expect that to impact your guidance and 2017. Is there any way for you to help us understand that versus maybe where it was in 2016? And then as well as factors that will be impacting that?

I think, we'll double check but I believe we are at about somewhere between 22 and 23 months and we're probably what 21-ish a year ago.

So you clear seeing longer duration. I think it is really driven by the price. Customers are skipping the value of just a lower price because you get a better price if you buy a longer term deal.

They know that they are going to renew. So it just makes sense to go ahead and pay it upfront.

I think as we get more in the enterprise, it feels as we get more in the enterprise, that we see more of that type of buying behavior. Where they're figuring, well look if I'm going to buy this architecture and I'm buying it because I get security orchestration end to end, and getting more value out of subscriptions. I'm certainly going to renewal. So why not pay for upfront.

And so you do have to assume that trend, there is some risk that trend will continue. I can't guide on it one way or another because there's no way to really see that until it happens. It leveled off a little bit of think Q4 to Q3, but I think it's really hard to predict.

Okay. Great. Thanks. I'll cede the floor. The next question comes from the line of Keith Bachman, Bank of Montreal.

Hello, thanks. I had two questions.

Number one, your competitor talked a little bit about pricing in certain areas including maintenance. Was there anything that you would like to call out in terms of pricing issues at all?

No. I think discounting is -- we haven't seen any issues there. We have actually --

Did it abate or was it a better environment? It seems like demand is pretty strong this quarter across the board.

So I mean obviously if you look at our product gross margins they're very strong. I think that's an indicator to some extent about pricing leverage. And that's it. Other than that, obviously with the enterprise bundle you're driving a higher deal set.

Also what's [in the] strong technology engineer, refreshed (inaudible) quarterly like the ( inaudible ) we just launched our high-end the [70/60] and also the (inaudible) [39/80] the most high-end quarter in the whole industry.

That's also increased a lot of like capacity to add additional service, additional performance. That's also making the system (inaudible) higher price combined with the more mega service subscription attached to the box.

Right. Right. Understood. The second question was on the first call. There was a question asked in the cloud what's a mix of businesses if you will. And your answer was selling more appliances than the virtual firewalls, but how do you see that unfolding over the next three to four years? Do you see that same mix or do you think virtual will take a bigger footprint of your cloud business.

Definitely the cloud; starting to grow faster than real appliance, but also a lot of the enterprise. Depends on the application. They need achieve some imbalance.

Some application can be deployed on the cloud, but some other still need to be deployed in the [promise] they have. The other issue they face is really when you put all the data -- when you put all the, whatever they call it, (inaudible) into the cloud, they also need to have a secure way to access that.

Whether (inaudible) or mobile device. That also increase the sales of our price. I'm trying to make the connection between their company to the cloud.

Okay. The only other thing, I just quickly would say, we constantly hear that it creates more complexity. And the examples we gave were hybrid models. And if you look at the large deal growth within that is some reflection of the problems of hybrid models.

Okay. Makes sense. Many thanks. I'll cede the floor. Congratulations.

Thank you.

The next question comes from the line of Fatima Boolani, UBS.

Hi team and Kelly. Nice to meet you telephonically. Thanks for taking the questions.

Ken a question for you on the IoT trends that you referenced in your prepared remarks, and just the opportunity that you're seeing on wireless security side. More tangibly, how has the assets from Meru that you've integrated into your portfolio, how is that business trending vis-a-vis the broader multi-billion-dollar wireless LAN opportunity?

I think we're in the middle probably half way there to call it unified -- a unified (inaudible)Wi-Fi and security together. So some of the Meru employees that are in transition into that part; some are still not there yet.

And I think maybe give us a few more quarters we can have a more productive transition into what we call a unified access. That's what would drive additional sales, additional value. We are still in the phase to reengineering some of the part up and also making some other additional service of the equipment what works mostly better. But I think its the market space and also where you position ourselves to give the customer the whole fabric approach.

And that's where you will find the endpoint to the access to the network. And then there is additional applications like the web email and in the cloud.

That's helpful and a question for you Drew.

On some of your competitors and peers have evaluated alternative commissioned recognitions (inaudible) something that we've mentioned in passing. So I'm wondering how you think about that. Because clearly increasingly larger portion of your sales are moving real estate to the balance sheet.

So I'm wondering how you are thinking about that into 2017. If that is something that's a focal point and if that is reflected in your 16% margin [guidance]?

Well our 16% margin guidance is on consistent accounting. And we don't -- certainly there's no revenue. We don't anticipate this point any revenue or commission accounting changes in 2016. I'm sorry, 2017.

In 2018 we have to go to the new accounting standard. And that was what Walter was referencing earlier. And in that it also forces you to recognize the commissions in principle match the commissions I believe it's over the life of the customer.

There's some debate I think of this right now whether it's over the life of the customer or the subscription. And that's what we're working through right now. I would just go back to what I said to Walter, we are still evaluating.

Okay. And last one for me if I can sneak one in.

International demand and just demand patterns in buying behaviors, is there a disparity or disconnect versus what you're seeing in North America?

I know North America typically tends to be strong and long. But you've historically had a very strong presence internationally. So I'm wondering if you can contrast or juxtapose the demand complexion domestically and internationally?

That's it for me.

I think the demand is good in both places. I think for us it's just the newness of the go-to-market engine, and sales model in North America that has made it more challenging for us historically.

The next question comes from the line of [Patrick Folzo], [Aready]

Can you give any color on the revenue or the growth mate or some metrics around the cloud and first editions?

On the revenue side? Well it is growing very nicely. I think cloud is the fastest growing part of our business. We don't share the numbers.

You can't get any more specificity?

There small numbers. There very small numbers.

And then on cloud pricing, can you help me understand broad pricing dynamics on the cloud versus on prem. I know they are different. Just help me frame it to help understand it. Do customers think of it in the same way? Pretty much talk me through.

Cloud is such a broad term. I think we need to be a little careful. What I'm talking about is more the metered model with my last response. When you're thinking like what you would get from AWS or Microsoft Azure.

If that's what you were referring to, that's the metered model. Again is very small. It's a utility-based model and so basically unit based on time. And that's really how it works.

And also, for the price compared to the cloud virtual deployment, what's unique -- what's different for that really we do have our own secure process (inaudible). We just (inaudible) security function like (inaudible) faster then the normal CPU. In which they were using the cloud or more other competitor platform.

So that's where -- so they see -- I think for us we have a nice balance with the appliance with the virtual cloud deployment. But even the cloud provider, even the service provider carrier, they do see the benefit of some of our appliance because the secure process higher demand that we have.

So that's not our competitor, or some even cloud provider they don't have. So that's where the secure process can also be open up some new market for the cloud. Secured results is also one huge function value added for a lot of service provider, cloud provider.

And I think -- Patrick, ultimately just a broader term here, the fabric concept is (inaudible)is to the cloud or for the cloud. We partner with all major cloud providers and service providers. You can buy from them. You can buy from us. It is the fastest-growing segment of the business.

And it is just one of the things that we believe is a factor of growth for us. Has been somewhat, it certainly has driven some of the hybrid deals that I mentioned some of the larger deals. Which I referenced earlier and we're hopeful that trend continues.

And can I ask just one more? Over time in the medium to long term. Is the expectation that you on (inaudible) flat lined or maybe even declines? And if not why would that be wrong?

I don't there will be decline and depends on how the cloud application. When you move the data to the cloud you do need to increase the access or secure access for the cloud data there. So that's where you do need the promise to do that.

And at the same time, even for the cloud itself, security function also need to have whatever you need to run secure function. Just like now you see the GPU for the (inaudible) you see the TPU for the AI (inaudible) quickly add a huge value. I do believe secure processor eventually can also huge value for whether the cloud provider or for any eco-structure.

Okay. And we need to move on to other questioners please.

Thank you guys. Good quarter.

Thank you Patrick.

Our next call comes from the line of [Mike Passado], Pacific Crest Securities.

Hi guys. Sorry to disappoint; it is Rob Owens for Mike. Hope you're well. A couple of things.

Number one on the non-FortiGate side. I know there wasn't specific quantification. But any more color you can give us it becomes significant from an accounting treatment standpoint will you break it out for us? Or will it just remain in the numbers, from a general perspective?

Hi Rob. And not a disappointment by the way. (Laughter) The -- we don't have any plans to break it out right now. We don't really characterize it as another segment of the business in accounting terms.

So I wouldn't call it that yet. But I would say it is growing nicely. And the story right now is more around fabric. The fact is that just buying more products per customer I think is really the driver there. That's how we think about it.

And then secondarily, any color you can give from a high level regarding unit volumes? And does the product revenue line, I know that there's been more that is being deferred especially as you have different bundles and so it's offsetting that a little bit.

What headwind is that creating overall? I know it is a difficult metric because you sell to multiple bases at different price points. But is there any general color you can give us just in terms of what unit volumes look like to be?

I could try a little bit generally. They continue to be fine. If you look at the mix, we were more weighted probably toward the mid and highs this quarter.

And I wouldn't call that it trended all. We said this many times. We always qualify the mix from one quarter to the next; is generally not a trend and its hard to predict. But, fair enough question on the units but I think fine.

When you get more low end to this you obviously drive more units. What was the other part of your question? The --

How to think of that as a headwind; the puts and takes with regard to balance sheet revenue versus a year ago.

Honestly, the way I think about it is its ultimately more predictability, and it is a higher margin recurring revenue stream. It's hard for me to get away from that.

A lot of this, we believe, a lot of this is intentional. Clearly we sell the enterprise bundle with the price increases a couple years ago, and we do have breath-of-product portfolio across virtual and into the cloud. And so, all of those things tend to be way towards more the balance sheet side than the deferred revenue side.

But then it becomes more predictable and it is recurring and it is higher margin generally. I think of it as a tailwind.

As your moving up market with more bundles be going more strategic with the fabric play. Are you seeing any change to renewal rates at this point or is that still coming down the road?

No. No change there. It's interesting our retention rates and I think right now -- I think Walter asked a question earlier about duration. And if they're doing three-year deals or five-year deals, they generally switch to another appliance probably at the end of that.

Maybe not three years. Maybe they renew for a year or two or whatever. But some of that plays into it. Plays into just waiting for more tenure with the customer if you will is a good thing. Your baking in a renewal if you will.

But we don't count that into renewal rates. We look at the renewals coming up. I just wanted point that out. Your clearly seeing that shift, but renewal rates remain similar to what the been in the past.

Great. Thank you for the color.

Yes.

Our next question comes from Sterling Auty, JPMorgan.

Thanks sorry to disappoint, Sterling Auty for Rob owns; just getting. (Laughter) If Rob and I don't pick on each other nobody will.

I apologize if you answered these on the main call. There was a bunch of us jumping back and forth between you guys and FireEye.

On the CapEx can you break out how much of the CapEx in total is dedicated to the real estate initiative versus CapEx for normal course of business?

Sure. I can Sterling.

The CapEx, again this is high. I would think our run rate CapEx -- let me make sure I get the specific numbers. I'm going to go back to my script because of want to make sure I've given you exactly what I said. Give me just a second.

And then I'll try to reconcile it for you if that's okay. So I said we anticipate CapEx of approximately $140 million to $150 million in 2017. Roughly $15 million to $20 million of this number was pushed forward from last year. We estimated, when we estimated $80 million to $90 million in CapEx but only used $67 million.

So $140 million to $150 million. And the run rate CapEx embedded in that is about $25 million; $25 million to $30 million.

The hard part of this and as you can see from Q4, it's hard to predict how these things close. Because we negotiate, and we try to get the best deal. And we also try to -- the other side is doing the same thing. And then there's complexity in buying real estate.

So is hard to predict but again that's the answer on just more directly your question.

Got it. Got it.

And then, the midrange of the product line as a percent was the strength in the quarter. Any sense as to -- was there any change in deployment?

And what I mean by that is if I go back a couple of years ago in the last refresh that Checkpoint did. One of the things that they did is the performance got so good that people traded down. Meaning they were using cheaper solutions in areas that historically were using the more expensive.

In the midrange in this, are you seeing that same creep? Or is there just a broader deployment because of micro segmentation other things for that midrange line?

We don't quiet see the trade down part. Because we're keeping even the model number the same. Like just sometimes from the D version to the E version its already in the fifth or sixth generation now.

So customer pretty consistent keeping buying the same model after few years. And at the same time because we increased additional service or function into the box. So they do need to have like a more powerful model to adopt (inaudible) enterprise service or other ADP mobile security service.

So that's where -- and then also we do have some new high-end coming out. Like I say the two most high-end in the industry. The 7000 Series also the new 3000 Series. That's also helping drive some high-end big enterprise sales. And also for the service provider the same thing.

So that's where we see -- we don't quite see the same trim down. Because of additional function needed additional service that they needed.

I would like to say that I'm not being super strict with follow-ups, but we do have a bunch more people on the line for questions. So maybe we can take further questions at another time Sterling.

That's fine. Thanks Kelly; thanks guys.

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

Hello again Ken and Drew.

I would like to ask your assumption on product revenue growth in your Q1 and full-year 2017 guidance. If I assume sequential growth in services revenue in Q1 or (inaudible) analysis that says service revenue grew by 25% in 2017. The guidance may correspond to a low single-digit growth rate in product revenue. That steadily presents some conservative assumptions in your guidance.

We don't really guide on product revenue. Again you have -- I would only just caution that again we've seen the mixed shift which a lot of the conversation has been focused on.

I'm wondering whether I missing something if I assume sequential growth in services quarter after quarter.

We just don't guide on mix, we just have to stand by the revenue guidance we give.

Sure. That's fair. Thank you Drew.

You're welcome.

The next question comes from the line of [Saket Kalia], Barclays Capital.

Hey guys. Thanks for taking my questions. One question and if I have time for a follow-up, and if not let me know Kelly.

Okay.

The first one, a little bit of a housekeeping question for you Drew. I noticed there was an other expense item in the fourth quarter as you often do. Are we baking in any non-operating expenses in the 2017 guide? So basically expenses both below the operating line?

Am I baking in anything unusual below the operating line for 2017? Yes, no.

And for my follow-up a little more strategic. But now that we've got the third quarter miss step well behind us in the rear view mirror. Is there anything that you've learned -- anything incremental about, beyond on what you said, about what you think is true of that miss step?

Is the business maybe becoming a little bit more backend loaded through the year? Or did maybe the higher security fabric content in deals maybe contribute even more to some of that young sales force in the longer sales cycle? Any color for maybe a two quarter postmortem would be helpful.

Sure. The environment as a whole, I think it's consistent across the environment that people are seeing -- everyone seeing elongated sales cycles.

They're some of the digestion. There's clearly the consolidation that we talked about and a lot of it -- you can think about as digestion but I do think the flip side of that coin is evaluation. And so as they evaluate the next-gen architectures it takes time.

I think perhaps the changes at the beginning of the year sometimes with all of that perhaps they take a little bit longer to take hold, and I don't have a lot more to add with that. Other than the lessons learned is keep it flat and focused.

Keep people focused on customers and make sure you don't build to many layers in the Company.

Got it. Very helpful. Thanks guys.

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

Great. Thank you for taking the question.

Maybe just gross margin. Sounds like there was some really nice tailwinds in the mix shift part of the business. And if I look at the results pretty consistent gross margin expansion across the last couple of years.

Just curious Drew is there some other factors that we should be thinking about that's going into the guidance for gross margin to be about flat this year?

Well, we definitely left the year with a product mix towards the higher end. Which has a higher margin. So that's it.

I think we're just waiting in -- its hard to predict how much is going to mix shift into next year. I think we're just -- we're building in based on what we've seen in the trends, and just been cautious to be frank.

Being prudent I would say is the word.

That makes a lot of sense, and as a follow-up just from the announcement from the other day. The defense approvals for the classified networks. Maybe help us understand how material that could be, and then broadly speaking how the demand environment is tracking in federal? Thank you.

You're talking about US side?

That's right.

We're keeping the (inaudible) in the US side definitely. Also we saw new [see saw] on board come from the government side. Also keeping [evas] in team there also.

The approved by the Defense Department for the classified network deployment really help a lot. The same time also helping to drive additional sales. The government side is come about 18%, that's really the global government. By the DOD these approvals also can help drive additional government sales. Even for some overseas government there.

So that's where we feel this is pretty helpful for the overall business for us

That's helpful. Thank you.

Takes Gabriela.

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

Okay. Thanks very much.

Ken wondering if you could elaborate a bit on what you saw on Q4 with respect to internal segmentation, firewall activity, and is there any variance in average sales cycles for those type of deals?

Yes, I think the big, it's a large enterprise do see the internal segmentations are very important. Also we started promoting internal segmentation two years ago. We expand this into the whole fabric because a lot of data not only just moving internally between department or some other networking gear.

But also go to the cloud, go to the mobile device, and also even some to the IoT. So that's the whole fabric to the extent beyond internal segmentation. But for the big enterprise the first step usually changing from on the defense on boarder to the internal segmentation.

We see still in the early stage still ramp up pretty quickly, and it's become -- in the next few years will become a standard for all the big enterprise. On how they want to secure the internal network. It has to be to a lot of segmentation, and how to secure different department, different server.

So that will become a new standard for the whole Compass network, and the fabric will be expanding beyond that to the cloud, to the mobile device, and to the IoT.

Perfect. And then just a quick follow-up for Drew.

Just getting back to CapEx. I understand that the building expansion in Sunnyvale and Vancouver is a multi-year process. But from a high level modeling standpoint I'm trying to get a sense of whether or not you think that CapEx should be directionally lower than the $140 million to $150 million level in 2018?

Anything that you might be able to provide there would certainly be appreciated.

Yes, we're not guiding out to 2018 and beyond 2017. But Greg I would certainly include some provision at that. Its hard to predict the levels will do each year. But you have to go through -- again there's just a variety of people to coordinate, and zoning approvals, and variety so its hard to do.

But I would certainly -- you'll want to bake in some provision for that.

Okay. Will do. Thank you.

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

Hey Drew and Ken and Kelly.

Hey Michael, we can barely hear you.

It would help if I move my microphone down. There we go. Hey Ken, hey Drew, and Kelly. Good to see our paths crossing once again.

So first of all you talked a lot about multi product this quarter. Can you rank in order of -- what's with the additional products that most contributed to the acceleration growth this quarter? So multi product, which are the additional products non-core.

Sandbox and Mail -- FortiSandbox and FortiMail.

Okay. Anything else after that?

I think one side is really the access. Whether the Wi-Fi, some network gear, and the other side is application. That's ties most closely to the network side.

Both we see some nice growth there. Then the more further side -- the one side is really the end point, and also the cloud, is still relatively small number. But also see some growth. But is safe to say from a contribution from the numbers side probably the application; which is like a mail Sandbox and a web. And then the access part is really Wi-Fi. They do help.

Yes and Micheal part of my answer reflects the trend over the last several quarters.

Got it. And then for my follow-up.

When you say the cloud -- I think you mentioned, you said that cloud had doubled. Did you quantify it is some way?

Well it depends on how you define it. The cloud --

I wonder if you could define it for me. When you say cloud -- when you guys say cloud exactly which do you mean? Do you mean the virtual version, do you mean the revenues that are going through the Amazon and Azure store; what exactly do you mean when you say cloud?

Some one asked me earlier about the growth. What I was talking about was AWS and Microsoft Azure; which is the metered model.

But it is a small number Micheal. More than -- I am pretty sure more than doubled from last year. But is still a small number. The broader definition of the cloud would include virtual machines, software, and then could even be the hybrid model. Where we are doing -- and this is where is gets harder to define exactly how much it is. Because a lot of the appliances could be going to a hybrid model somebody could be managing them for that. Could be a -- I'm trying to think of the words again, the a -- [Portacloud].

There is a two part of deployment (inaudible) to the cloud. Where is really the service provider model. The other one is for the cloud really is the cloud provider himself. Also deploys both on the software version and also the hollow version of the [six] security function there.

Not to confuse you more, but the word I was trying to think of was BYOL which is bring your own licence where they could by the box through any channel, and then just deploy it somewhere else and then be using a metered model with it. So it could be a com point. Its hard to define.

Okay guys. Thanks very much

Our next question comes from the line of [John Lucia], JMP Securities.

Hey guys. Thanks for taking my question.

You are the second firewall vendor to report to pretty good results here. While there are some other security vendors that haven't faired as well. I know you've been asked about security spending, but I just wanted to ask if there is anything specific to the firewall market that drove the strength this quarter?

Have you seen any reallocation to the security budgets to the firewall? Anything like that? Or is the improvement here really specific to Fortinet in terms of sales execution?

Well I think it is the fabric. Its the difference between the fabrics and the competitors. What they might often refer to as a platform, and its -- we're unlike isolated competitor platform solutions that provide visibility at a certain point in the network.

But security fabric has a common operating system, has a single pane of glass, offers customers a unified security posture, and complete visibility across the entire network infrastructure. We do segmentation, we can integrate end to end, security orchestration.

Those things differentiate us. And that's important because the enterprises are evaluating the next-gen architectures. That, along with the better performance from the self ports, is really what drove our quarter.

Okay. And you have highlighted sales to hybrid cloud being strong in part do to the fabric. I just want to understand is that because your win rates are higher there in hybrid cloud because you are more differentiated?

Or is that because the move to hybrid cloud actually drives incremental firewall purchases? Or is it a combination of both?

Well look I think when -- the way we always talk about that is, generally when we feel that when we test we win. So if they're testing, if they're big enough to be testing we are going to have a pretty good at getting the deal. Because clearly the performance shows through, and the single pane of glass, and the security orchestrations I talked about. Those things are salient, they come across as salient I believe in the test phase.

So that's how we think about it. Other than that we don't really have -- its hard to predict win rates by a type of sale so to speak.

Okay. Thank you.

Our next question comes from the line of [Taz Hajoji], Deutsche Bank.

Hey guys. A couple questions from our side.

So you had issues with the reorg in Q2 and than you obviously struggled in Q3, and you back strong in Q4. Are you -- are all the issues that you had with reorg behind you now? Are you still the same (inaudible) perception because of the reorg you had in 2016?

Yes Taz I just think there is still some newness to the model. We of course like to see trend in consistent performance. We continue -- I think we have our engine generally in place. I think there is always some fine tuning to do, and performance improvement. Which we will seek to do.

We continue -- we are a learning organization. We will continue to learn and adjust. But we are focused on making sure we have good Legion, we continue to refine that model if you will. And make sure we have the right marketing to drive right ROI.

So we are very focused on those items.

Got it.

And then on the margin guidance for next year. You had a very hard hockey stick in margins going from first half of 2016 to second half of 2016. Your 2017 do we expect something similar? Or will the margin growth be more moderate throughout the year?

I have only guided for Q1. But they tend to ramp through out the year. That 's been our traditional -- if you go look at our past over years. That is generally consistent.

Then how can you say from more first half to second half? Or because you had -- you basically had flat margins.

I said ramp. Clearly Q4 had a -- I agree with the hockey stick comment. But it would be nice to see more of a ramp, but again it gets easier as you perform more earlier in the year of course.

Got it. And then one last one. If I could.

You guided to [buildings growth] of 16% for the year. And then margin sales are going up from 2016 to 2017. So would it be safe to assume that cash flow growth would be higher than your buildings growth, given that your margin sales are going up?

Cash flow from operations --

Are you talking operating cash flow?

Yes. Exactly.

Obviously we have talked about free cash flow. So -- I am just trying to think of the math. Sure if billings are going up, and margins are expanding then it should go up. We even said that we would even try to find operational benefits, and operating cash flow. Working capitol type items to the extent that we can.

Got it. Perfect. Thank you guys.

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

Hi guys. Thanks for taking the question. And congrats on the quarter again.

I was just wondering can you remind us again of your process for building out your billings guidance? Just directionally what proportion of 2017 guidance is coming from existing costumers?

Well I will start with the process. We have a bottoms-up process. Of course there is a bottoms up and a tops down.

The bottoms up is basically what's the deal-by-deal forecast from the sales team; how big is the funnel, what stages are they in, and what are they predicting to close. Clearly its could to good to get them to commit. That's what you ar always liking to do.

Then we look at other statistics consistently. How big is the funnel -- the more gross macro consistence -- how big is the funnel. Is there any concern on close rates.

Then obliviously from the tops down you want to look at the other factors. The environmental factors, or economic macro factors, and take those into consideration. Then you do a lot of questioning on the sales people. On confident are they quiet frankly. And you try to be as scientific as you can, and make sure there is dose of realism in there.

Okay, And then --

Then costumers -- yes renewals generally you are looking to -- we don't really break out what we are going to get from existing; although we are very focused on capturing new customers.

Okay great. Thanks very much.

There are no further questions at this time. I would like to turn the call back over to Kelly Blough for closing remarks.

Thank you everyone for calling back in. We look forward to seeing you at various conferences through out the quarter.

Ladies and gentlemen, thank you for participating in today's conference. This does conclude the program. You may all disconnect. Everyone have a great day.