阿卡邁科技 (AKAM) 2017 Q3 法說會逐字稿

Good day, ladies and gentlemen, and thank you for standing by.

Welcome to the Q3 2017 Akamai Technologies Earnings Conference Call.

(Operator Instructions)

As a reminder, this conference is being recorded for replay purposes.

It is now my pleasure to hand the conference over to Mr. Tom Barth, Head of Investor Relations.

Sir, you may begin.

Great.

Good afternoon, and thank you, everyone, for joining Akamai's Third Quarter 2017 Earnings Conference Call.

Speaking today will be Tom Leighton, Akamai's Chief Executive Officer; and Jim Benson, Akamai's Chief Financial Officer.

Before we get started, please note that today's comments include forward-looking statements, including statements regarding revenue and earnings guidance.

These forward-looking statements are subject to risks and uncertainties and involve a number of factors that could cause actual results to differ materially from those expressed or implied by such statements.

Additional information concerning these factors is contained in Akamai's filings with the SEC, including our annual report on Form 10-K and quarterly reports on Form 10-Q.

The forward-looking statements included in this call represent the company's view on October 24, 2017.

Akamai disclaims any obligation to update these statements to reflect future events or circumstances.

As a reminder, we will be referring to some non-GAAP financial metrics during today's call.

A detailed reconciliation of GAAP and non-GAAP metrics can be found under the financial portion of the Investor Relations section of our website.

Also, please note that all growth rates referenced today will be in constant currency unless otherwise noted.

With that, let me turn the call over to Tom.

Thanks, Tom, and thank you all for joining us today.

Akamai delivered strong results in the third quarter, driven by the continued rapid growth of our security offerings and strong traffic acceleration in our media business.

Revenue in Q3 was $621 million, above the high end of our guidance range and up 6% over Q3 of last year.

Non-GAAP EPS in Q3 was $0.62 per diluted share, also coming in above the high end of our guidance range.

The fastest-growing part of our business last quarter was our cloud security portfolio, which grew 27% year-over-year to $121 million in revenue or about 20% of Akamai's total revenue.

Our leading Kona Site Defender solution continued to be a primary driver of this growth, and I think recent events can help you understand why.

By now, most everyone has seen the headlines about the breach at a major credit reporting agency that exposed the personal financial data of more than 140 million people.

According to published reports, the attackers exploited an Apache Struts vulnerability in a public-facing website to gain access to the confidential data.

The company's leaders believe they patched all their sites and apps after the vulnerability was identified, but when there are a lot of sites and apps to update, it's really hard to be sure they're all covered.

As it turns out, they missed one, and the gap was exploited, with bad consequences for their customers, their brand and their market cap.

This is one of many scenarios where Kona offers significant value.

Kona examines all the traffic coming to a website to determine whether it's safe.

Attack traffic trying to exploit the Apache Struts vulnerability or other web app vulnerabilities can be blocked by Kona so adversaries won't cause harm.

In other words, Kona can be used as a virtual patching layer in front of websites and applications, allowing our customers to be protected even when they haven't patched everything.

As this example illustrates, using Kona to provide defense in-depth is more important than ever when it comes to cybersecurity in today's world.

Kona also leverages the enormous capacity in the Akamai platform to protect our customers from large DDoS attacks.

By placing our servers in thousands of locations near end-users, we can absorb enormous volumes of attack traffic that could easily overwhelm traditional data center defenses.

Kona was a major topic this month at Edge, our 10th Annual Customer Conference, where we announced and demonstrated the latest addition to our cloud security portfolio, Bot Manager Premier.

Bot Manager Premier is designed to use the advanced machine learning technology that we acquired from Cyberfend to stop account takeover attacks.

Account hacking is becoming a huge problem.

In a recent study, we found that about 2/3 of the login attempts that we monitored across our customer base came from bots that were trying to break into accounts.

And these bots are becoming more sophisticated at evading traditional defenses based on things like the number of login attempts that come from the same IP address.

To counter this threat, Bot Manager Premier uses artificial intelligence to distinguish machines from human users.

It can detect the way humans touch a screen or press a key or move a cursor with a mouse or how people hold and move mobile devices in ways that machines do not.

Although Bot Manager Premier has only been in the market for a short while, we already have over 70 customers who are using it to help reduce the billions of dollars that we believe are lost to credential abuse annually.

Bot Manager Premier and Kona Site Defender are designed to protect external-facing apps.

At our customer conference, we also demonstrated a new solution that will enable enterprises to use Kona to protect their internal applications.

The traditional way to protect internal apps has been to use firewalls and architectures based on the notion of a perimeter defense.

We believe that such methods aren't good enough anymore.

That's because the majority of data breaches are now triggered from inside the enterprise perimeter.

Sometimes, it might be a malicious employee doing something bad, but more likely, it's as simple as an unwitting insider getting infected with malware by clicking on the wrong link or visiting a compromised external website and then bringing the infection inside the perimeter, thereby enabling the attacker to corrupt, destroy or even exfiltrate sensitive data.

Last month, a major airline deployed our new solution to protect their internal applications, and within the first week of use, Kona had discovered and blocked 85 attacks that came through trusted insiders.

Another new solution that we released in June, Enterprise Threat Protector, helps enterprises prevent employees from getting infected in the first place and also helps stop data from being exfiltrated once the enterprise has been breached.

Enterprise Threat Protector provides a recursive DNS service designed to block employee access to infected sites and to enforce compliance policies quickly and uniformly.

Customers love that it's very simple to implement, with most getting it up and running in a matter of minutes.

As we announced on October 11, we expect Enterprise Threat Protector to benefit from our planned acquisition of Nominum later this year.

Nominum provides recursive DNS and enterprise security products for over 100 of the world's leading carriers across more than 40 countries and serving 500 million subscribers.

Nominum's ThreatAvert service detects suspicious DNS activity and identifies malware in subscriber and enterprise devices, thereby protecting users, businesses and carriers from malicious noncompliant or otherwise undesirable content.

In addition to bolstering our enterprise security offerings through enhanced data and functionality, we expect that the Nominum acquisition will strengthen our access to the carrier channel for our enterprise security products.

I'm happy to report that many of our carrier partners who attended Edge were enthusiastic about the planned acquisition and the opportunity for us to work with them on technology that is so important to the operation of their networks.

Most people don't pay a lot of attention to DNS, but DNS is the on-ramp to the Internet.

It's key to managing traffic flows, and it's a great place to insert security.

Some of you may remember a year ago, when a small DNS provider got hit with a DDoS attack that resulted in a large number of important websites becoming unreachable.

By integrating Nominum and working more closely with our carrier partners, we believe that Akamai will be able to help make our customer sites and the Internet as a whole much more resilient in the future.

While many of the conversations at our customer conference were dominated by concerns over cyber threats and interest in our new security products, there was also a lot of interest among our media customers in the rapidly evolving OTT landscape and especially direct-to-consumer offers.

Based on our meetings with key executives from several of the world's largest broadcasters, Disney's direct-to-consumer announcement in August is an example of what we think may become a trend over the next several years.

We believe this would benefit Akamai because of our close relationships with many of the world's leading broadcasters.

During our last earnings call, I outlined several steps that we're taking to accelerate revenue and traffic growth in our media business.

I'm happy to report to you today that these efforts have started producing results.

Year-over-year traffic growth on the Akamai platform accelerated in Q3 to the highest levels that we've seen in 2 years.

In fact, Akamai's traffic grew faster than the 24% year-over-year growth rate that Cisco estimates for the Internet overall.

And our video traffic grew at nearly double that rate.

Just last month, we supported a single media event that drove more than 17 terabits a second in traffic.

And on that day, we delivered 60 terabits per second overall on the Akamai platform, establishing a new record for our business.

To put this number in perspective, 60 terabits per second is the equivalent of 12 million people watching a video streamed at 5 megabit per second at the same time.

And 5 megabits per second is a typical bit rate for the over-the-top streams provided by major broadcasters today.

It's not hard to imagine audiences that are 10 or even 100x as large in the future and viewing media at even higher bit rates.

This is a key reason why we're so optimistic about the future of our media business and why we continue to focus on improving the scale, quality and affordability of our over-the-top media delivery services.

At our customer conference earlier this month, we demonstrated our new low-latency streaming technology, which provides an experience that is several seconds ahead of satellite.

This is a big improvement over other Internet streaming solutions that are anywhere from 10 to 90 seconds behind satellite, a delay which creates a much less desirable experience for live events and linear broadcasts.

At the conference, we also demonstrated our new media client software, which draws content from multiple sources to produce an optimal viewing experience for end users while also providing cost advantages for our customers.

We still have plenty of work to do to return to historical levels of revenue growth in our media business, but it's encouraging to see the hard work starting to pay off as we gain share in the market.

There's one last item that I'd like to mention about our customer conference and that's how favorably our customers reacted to the progress we've made in dev ops.

Many customers told me how pleased they are to see our advancements in tooling for developers and the ecosystem of partnerships we built to make developers' lives simpler.

We're making great strides in dev ops so that developers and system administrators can work with Akamai their way, with the agility they need to deliver their software innovations faster.

In summary, I remain very confident about Akamai's prospects for future growth.

I continue to believe that our solutions deliver excellent value for our customers and that our innovations will help our customers thrive as the Internet landscape continues to rapidly evolve.

I'll now turn the call over to Jim for an update on our financials and the guidance for Q4.

Jim?

Thanks, Tom, and good afternoon, everyone.

Before I dive into the details of the strong third quarter results, please note that all revenue growth rate references will be in constant currency.

As Tom just highlighted, Akamai delivered a strong Q3, with both revenue and earnings coming in above the high end of our guidance range.

Revenue in the third quarter was $621 million, up 6% year-over-year and up 8% if you exclude our 6 large Internet platform customers.

Revenue growth was solid across most of the business, with the overachievement versus our guidance driven by a higher-than-expected uptick in Media Division traffic and continued strong growth in our Cloud Security Solutions.

Looking more closely at the Q3 revenue results by our customer division lens.

Revenue from our Web Division customers was $328 million, up 14% over a particularly strong Q3 last year.

We continue to see solid growth and diversification into this customer base, particularly with our cloud security offerings, with 53% of Akamai's third quarter revenue now coming from our Web Division customers.

Revenue from our Media Division customers was $273 million in the quarter, down 1% year-over-year and up 3%, excluding the impact of the large Internet platform customers.

As we outlined in our last call, we have been working diligently to increase traffic growth and share in our media customer base, and we began to see early traction in the quarter.

Our year-over-year traffic growth rate in Q3 was the strongest we've seen since 2015 and exceeded our expectations entering the quarter.

Traffic growth was strong across both our Internet platform customers and our core installed base, with particularly strong growth coming from our video delivery customers.

Finally, revenue from our Enterprise and Carrier Division customers was $20 million in the quarter, up 1% over a particularly strong Q3 last year for our carrier business.

Turning now to Q3 revenue results for our solution categories.

Revenue from our Performance and Security Solutions was solid, coming in at $381 million, growing 10% year-over-year.

This growth rate is a deceleration from historical levels, primarily driven by a moderation in the use of performance solutions by our Media Division customer base.

Third quarter revenue for our Cloud Security Solutions was $121 million, up 27% year-over-year, led by our flagship Kona Site Defender and Prolexic offerings and our continued portfolio expansion into new areas like bot management.

Our cloud security business now has an annualized revenue run rate of nearly $0.5 billion and continues to grow at a very rapid pace.

Our Media Delivery Solutions revenue was $183 million in the third quarter, down 3% year-over-year and up 2%, excluding our large Internet platform customers.

This growth rate is an improvement over Q2 levels, driven by traffic acceleration throughout the quarter.

Finally, revenue from our Services and Support Solutions was $57 million in the quarter, up 12% year-over-year.

Moving on to our geographies.

Sales in our international markets represented 34% of total revenue in Q3, consistent with Q2 levels and up 3 points from Q3 of last year.

International market revenue was $213 million in the third quarter, up 18% in constant currency, driven by continued strong growth in our Asia Pacific theater.

Foreign exchange fluctuations had a positive impact on revenue of $1 million on a year-over-year basis and $5 million on a sequential basis.

Revenue from our U.S. market was $409 million, up 1% year-over-year and up 3%, excluding our large Internet platform customers.

Moving on to costs.

Cash gross margin was 76%, consistent with Q2 levels and in line with our guidance.

GAAP gross margin, which includes both depreciation and stock-based compensation, was 64%, down 1 point from Q2 levels as we introduced more capitalized software projects onto the platform in late Q2 and early Q3.

GAAP operating expenses were $310 million in the third quarter.

These GAAP results include depreciation, amortization of intangible assets, stock-based compensation, acquisition-related charges and other nonrecurring items.

Excluding these charges, non-GAAP cash operating expenses were $245 million, up $6 million from Q2 levels and in line with our guidance.

Adjusted EBITDA for the third quarter was $226 million, up $2 million from Q2 levels.

Our adjusted EBITDA margin was 36%, down 1 point from Q2 levels and in line with our guidance.

Approximately 1 point of the Q3 EBITDA margin compression is due to the impact of the SOASTA acquisition.

GAAP depreciation and amortization expenses were $97 million in the third quarter.

These GAAP results include depreciation associated with stock-based compensation, amortization of intangible assets and amortization of capitalized interest expense.

Excluding these charges, non-GAAP depreciation was $84 million, up $7 million from Q2 levels and slightly above the high end of our guidance as we introduced more capitalized software projects onto the Akamai platform.

Non-GAAP operating income for the third quarter was $142 million, down $5 million from Q2 levels.

Non-GAAP operating margin came in at 23%, down 1 point from Q2 levels and in line with our guidance.

Moving on to the other income and expense items.

Interest income for the third quarter was $4 million, consistent with Q2 levels.

Noncash interest expense related to our convertible debt was roughly $5 million.

As a reminder, this noncash expense is excluded from our non-GAAP results.

Moving on to earnings.

GAAP net income for the third quarter was $61 million or $0.35 of earnings per diluted share.

Non-GAAP net income was $107 million or $0.62 of earnings per diluted share and $0.02 above the high end of our guidance, driven by the strong revenue achievement and a slightly lower tax rate.

For the quarter, total taxes included in our GAAP earnings were $26 million based on an effective tax rate of 30%.

Taxes included in our non-GAAP earnings were $40 million based on an effective tax rate of just under 28%.

This tax rate is about 1 point lower than our guidance due to a higher mix of foreign earnings.

Finally, our weighted average diluted share count for the third quarter was 172 million shares, in line with our guidance and down 2 million shares from Q2 levels, driven by increased share buyback activity.

Now I'll review some balance sheet items.

Days sales outstanding for the third quarter was 58 days, down 2 days from Q2 levels.

Capital expenditures in Q3, excluding equity compensation and capitalized interest expense, were $108 million and coming in slightly below the low end of our guidance for the quarter, benefiting from our focus on network CapEx efficiency.

Cash flow generation continued to be solid, with cash from operations of $236 million in the third quarter.

Our balance sheet also remains very strong, with roughly $1.4 billion in cash, cash equivalents and marketable securities on hand at the end of the quarter.

If you factor in our convertible debt, our net cash is approximately $725 million.

During the quarter, we spent $129 million on share repurchases, buying back roughly 3 million shares.

As we have discussed in the past, we believe the strength of our balance sheet and cash position is an important competitive differentiator that provides us with the financial flexibility to make key investments at opportune times, including the repurchase program and M&A.

We believe our SOASTA acquisition in Q2 and our expected acquisition of Nominum in Q4 are great examples of how we have used the strength of our balance sheet to invest for future growth.

In summary, we are pleased with how the business performed in the third quarter.

We exceeded the high end of our guidance range on both the top and bottom line.

We have continued to see a robust pipeline of innovation across the business, and we have already seen good traction in our traffic share initiatives, which we believe will improve the performance of our Media Division.

Looking ahead to the fourth quarter.

Holiday seasonality plays a large role in determining our financial performance, driven by online retail activity for our e-commerce customers and traffic for our large media customers.

As a result, the fourth quarter remains the hardest to predict.

Also impacting our guidance is the close of our Nominum acquisition, which is expected in mid-November.

As we shared in our press announcing the acquisition, we expect Nominum to be dilutive to earnings over the next 4 to 5 quarters due to the impact purchase accounting has on revenue recognition and costs associated with integrating and scaling the business.

We expect the acquisition will be accretive to earnings beginning in 2019.

Operationally, we plan to integrate Nominum into our core carrier business, and going forward, we will not be reporting on the Nominum business separately.

Lastly, we expect modest foreign exchange headwinds from the strengthening of the U.S. dollar over the past several weeks.

At current spot rates, foreign exchange fluctuations are expected to have a negative impact on Q4 revenue of $1 million compared to Q3 levels.

Factoring in all these variables, we are expecting Q4 revenue in the range of $638 million to $656 million.

This guidance assumes roughly 1.5 months of Nominum revenue.

And due to the impact purchase accounting has on revenue recognition, Nominum's revenue in the fourth quarter is expected to be a modest $1 million.

At these revenue ranges, we expect GAAP gross margins of 65% and cash gross margins of 77%, both up 1 point from Q3 levels.

Q4 non-GAAP operating expenses are projected to be $265 million to $270 million, up $20 million to $25 million sequentially as the business absorbs 1.5 months of Nominum spend and as we continue to make targeted investments in our web and enterprise divisions.

Factoring in the cash gross margin and operating expense expectations I just provided, we anticipate Q4 EBITDA margins of 35% to 36%.

As a point of note, Nominum is expected to compress EBITDA margins by roughly 2 points in the fourth quarter and into 2018 while we integrate and scale the business.

As we have said in the past, factors that could impact our EBITDA and operating margin profile would be M&A activities and revenue volumes quarter-to-quarter.

As we absorb the SOASTA and Nominum acquisitions, they will weigh on EBITDA margins in the near term.

And while we are not providing specific guidance beyond Q4 on this call, we are striving to operate the company in the mid-30s EBITDA margins and low 20s operating margins in 2018.

If we are able to continue to improve our media growth rates and deliver continued strong growth in our web business, we would expect to see EBITDA margins return to the high 30s.

And we remain committed to balancing both top line and bottom line growth over the longer term.

Moving now to depreciation.

We expect non-GAAP depreciation expense to be $84 million to $86 million.

Factoring in this depreciation guidance, we expect non-GAAP operating margin of 22% to 23% for Q4.

And with the overall revenue and spend configuration I just outlined, we expect Q4 non-GAAP EPS in the range of $0.60 to $0.65, inclusive of an estimated $0.04 dilution impact of the Nominum acquisition.

This EPS guidance assumes taxes of $41 million to $44 million based on an estimated quarterly non-GAAP tax rate of just over 28%.

This guidance also reflects a fully diluted share count of 171 million to 172 million shares.

On CapEx, we expect to spend $90 million to $100 million, excluding equity compensation, in the quarter.

For the full year, this level of CapEx would equate to roughly 16% of revenue, at the low end of our long-term model of 16% to 18% of revenue.

In closing, we remain confident in our ability to accelerate revenue growth rates for the business, execute on our continued product and customer diversification strategy and deliver a compelling financial model for our shareholders.

Thank you, and Tom and I would like to take your questions.

Operator?

(Operator Instructions) And our first question will come from the line of Sterling Auty with JPMorgan.

I wanted to dive into the media strength but the performance weakness in the quarter.

It sounded like it might both be coming from the same set of customers.

So can you give us specifics as to what, within those changes, resonated to gain share and drive media strength?

And what drove the weakness in the performance piece?

Yes.

So as I talked about in my remarks, we have some really strong new capabilities for media, particularly around video.

The idea that now we can do live and linear broadcasts ahead of satellite, that's a pretty big change over what is otherwise available with OTT streaming, typically, being 10 to 90 seconds behind the satellite broadcast.

The new capability is, from the client-side software, to be able to take sources of the content, multiple sources, integrate that and provide a seamless viewing experience for the user and also at a lower cost point for our customers is important.

Broadcast operations support is really important for high-quality OTT, the accelerated ingest technology, origin-connect capabilities, so we can connect directly with the large broadcaster origins.

The focus on the big 250 global media companies that we talked about in the last call, giving them the kinds of packages and capabilities that they want for things like subscriber base pricing, background downloads for that to make sense, peer-assessed delivery, regional pricing based on our deep carrier relationships, things like managed CDN services.

So a lot of stuff that we do that's very unique in the media business and focused on the big media broadcasters and customers, and we're seeing really strong response there in terms of gaining market share, traffic growth, particularly in video.

On the Performance and Security side, obviously security doing very well, with several new capabilities coming in to market over the last several months that we think will drive continued strong growth for our security products.

And in the Web Division, which features the Performance and Security products, 14% growth there, a little bit less than it was last quarter, and part of that driven by some of the things going on in public sector, which tend to be lumpy, and there's some uncertainty in Washington right now.

And so that had a little bit of an impact on the Web Division customer base growing at 14% this quarter.

Great.

And then just one follow-up for Jim.

So as March should have a full quarter of Nominum headwind, shouldn't that be the trough in margins?

And would you expect then a steady increase from there?

Or is it one of the things where you take a hit, maybe it comes up a little and just kind of flatlines until you see some of those top line drivers that you spoke about kicking in?

Yes, I think just thinking about it the right way, whether it's a trough or not, Q2 is also a quarter for the company where we do tend to seasonally have an uptick in expenses just in the core business.

And so it will probably be a trough in Q1 and Q2, and then it should scale from there.

And our next question will come from the line of Siti Panigrahi with Wells Fargo.

Just a follow-up to Sterling's question on the media traffic side.

This quarter, you talked about 2 record peaks in terms of media traffic, September 12 at 60 terabytes -- terabits, and that beat your August 29 peak.

So is that any kind of one-off event that you experienced this quarter?

I just wanted to understand how sustainable this traffic growth, or is it something to do with your focus now back on the top 25 media customers?

Yes.

So there's always events that take place, and the big events drive the new traffic records.

And as we talked about before, any given event itself, because it's a short-duration event, usually doesn't have a lot of revenue associated with it.

But when we're talking about the overall traffic growth, we're talking total bytes delivered.

And that's not driven by events.

That's driven by the top 250.

And we were very pleased to see that our total traffic growth in the quarter, the growth rates there being the highest we've seen in 2 years, higher than the Internet as a whole, and really pleased with the video, the aggregate video traffic growth being nearly twice what you see the established reports saying for the Internet as a whole.

So really gaining strong share there, and that's not measured really by these one-off events that sort of get the headlines because that's where you hit the peak of traffic.

We're talking about the core business in media, total bytes delivered, and that's because of the focus on the top 250, the focus on scale, quality at an affordable price and all the capabilities that I talked about around media.

And our next question will come from the line of Keith Weiss with Morgan Stanley.

When you think about the media business, particularly the platform side of the equation, you guys have seen revenues around $51 million for, I think this is like 3 quarters in a row.

Is it safe to say kind of like the worst is behind us in terms of the declines that you guys expect from those platform customers and then that should grow on a going-forward basis?

Or is this more of a pause in what could be a longer-term declining trend line?

Yes.

I think with the big platform companies, it appears as they've sort of stabilized at about combined 8% of our revenue.

Seeing the same quarter-over-quarter, now we've got a little bit of time still for the year-over-year wraparound effect.

So there's a little bit left to see there.

But basically, it looks like things are -- have stabilized on that front, with the 6 giant platform companies accounting for about 8% of our revenue.

Got it.

And then just one question on pricing, particularly so it's the pricing that you guys get on sort of the video traffic versus just the rest of the traffic.

How should we think about that, sort of the relative pricing on one versus sort of the rest of the traffic?

Yes.

I think the pricing tends to be higher where it's more difficult to do and there's a perceived higher value.

So in terms of bulk bit delivery, if it's something like a background software download, you'd see the lowest probably pricing there.

If it's doing something for OTT, live linear video, where the -- where our customers' users are paying, that's where quality matters the most and where our capabilities matter the most, and so we would be, generally speaking, paid more there.

Now, of course, the largest volume customers will also pay the lowest rates.

And so when you have very large customers, that will drive a lower price point per byte delivered.

So those are sort of the 2 ways of looking at it.

And one is sort of the scale, and the other is in terms of the value that's being provided.

Got it.

So when we see the video traffic growing twice the rate of the overall traffic, is that a positive for pricing?

Yes -- well, I think it's a positive for Akamai.

I think when you look at the future, the media business is really being driven by video.

There will always be software and gaming delivery, but really, the future is the adoption of OTT at real scale.

And we're really still at the beginning of that.

And when we see the video traffic growing, that's something that we see as a very positive sign for us.

And our next question will come from the line of Michael Turits with Raymond James.

Michael Turits.

So one question, maybe one on performance.

So just it's great, the acceleration in volume traffic growth rates.

Is that a function of you guys regaining share, given the fact that you had a new strategy on pricing and how you look at customers?

Or is it a function of the customers accelerating their, let's call it, organic growth rates?

And then I had a question on DNS also.

I think it's -- well, probably there's some combination there, but we do believe we're gaining share, and that's specifically account-by-account based on delivering what a major broadcaster needs or a major media company needs.

Also, when you measure, based on the Internet as a whole, we're going faster than the Internet as a whole, which would suggest for growth taking share.

Right.

Yes, so we're comparing against the established reports out there, what the Internet's growing as a whole.

In some cases, we have competitors that publish reports on how fast they're growing.

And so when we see ourselves growing faster, a lot faster than that, we take that to believe as further proof that we're growing our market share.

Now that said, I think OTT will continue to grow as a market as a whole, and there's a long way to go there for future growth, as I talked about with some of the ways of thinking about 60 terabits a second today and what viewing audiences might look like in the future.

Right.

And then maybe I'll follow up on media instead of talking about performance.

But on media, it seems as if it's -- is there a change in your view of -- relative to the Internet platform guys?

I guess my thought process or previous as well, your thought process as well, is that there was probably still some significant share movement of the DIY guys from this point forward.

Has that changed, that you feel like you've actually bottomed at this percentage level?

I don't think there's a fundamental change.

The giant platform companies are going to do a lot of DIY.

They are doing a lot of DIY now.

We're seeing a stabilization in terms of percentage of our revenue, 8% being from those 6 large platform companies.

I think as we look at OTT in the future, those 6 companies are going to have a reasonable share of OTT.

And I think as we look forward, we see a world where there will be a lot of direct-to-consumer and other distribution mechanisms in addition to those 6 large platform companies, and that presents a very large opportunity for Akamai.

And our next question will come from the line of Brad Zelnick with Crédit Suisse.

Nice to see stabilization in the media delivery side of the business.

Most of my questions have been answered on that front.

I want to turn just to cloud security, which, in the quarter, decelerated but at a consistent rate with, I think, what we've seen in prior quarters.

And as we think about the trajectory going forward, I wanted to just drill into Nominum a bit.

So as you stated in your prepared remarks, it builds on Enterprise Threat Protector.

And if I'm not mistaken, it's a Carrier and Enterprise play.

Can you talk about the go-to-market and the types of investments you need to make in this channel and how you see them paying you back over time?

And perhaps, if you looked out into the future, how much of security over time comes from the carrier channel, looking out 3 to 5 years maybe?

Yes.

I think the carrier channel is very important for us.

Carriers are incredibly important partners.

They're our fastest-growing channel by far.

We're deepening the relationships with the world's major carriers.

Nominum helps us do this.

Provide -- they provide a very important technology for the carriers.

And then on -- with the recursive DNS, it's what controls the on-ramp to the Internet.

It's what controls traffic flows across the carrier backbones.

We're in a great position to combine that with our capabilities to really help the carriers manage large traffic flows.

Now on top of that, Nominum's built their enterprise services, the enterprise security services that are very synergistic with our Enterprise Threat Protector services.

And so this will help Enterprise Threat Protector, and it will help grow the carrier channel for us.

Already, our fastest-growing channel, but I think helps it further.

So it's important for us in both senses, first with enterprise, security products and capabilities, and secondly, deepening the relationship with the world's major carriers.

And our next question will come from the line of Mark Kelleher with D. A. Davidson.

Sorry to do this, but I want to go back to the media again.

You saw some great growth there, 24% year-over-year, but revenue was down year-over-year.

So can you talk about the pricing dynamic?

I know you talked about some steps you were going to take last quarter competitively to regain some of that.

Can you talk about -- I know one of your competitor's in the process of being acquired.

Is there -- yes, just talk about what you're seeing in pricing.

And was part of the dynamic where, in the value curve, the bits where that you were delivering, maybe not OTT, but maybe more on the commoditized side?

Yes.

First, our growth was in excess of 24%, which is the Cisco published stat for the Internet as a whole.

Pricing is competitive in the media segment.

It has been from the beginning, for the 19 years we've been in the business.

And I think it will always be competitive.

And we work really hard to provide the best possible pricing for our customers so they can grow their business on the Internet.

We work really hard to lower our costs so we can be competitive out there and still be profitable in our media business.

So I don't think there's been a fundamental change in the pricing dynamic.

Every year, there's more traffic and at a lower price point per gigabyte delivered.

We are looking forward.

As we gain share, we definitely want to do that.

And we do believe it will help us grow revenue going forward and accelerate revenue growth going into 2018.

Jim, do you want to add anything into that?

No, I think you covered it pretty well that, obviously, the revenue is a function of traffic and price point.

And I think, as Tom outlined, that what we saw was an acceleration in traffic growth.

Our expectation is if we can continue to execute well, you'll start to see that acceleration in traffic growth manifest itself into an acceleration in revenue growth in the media business.

And our next question will come from the line of James Breen with William Blair.

I think it might have been last quarter, you talked about, on the video game side, seeing a little weakness in some of the updates.

Are there any particular verticals this quarter that you saw strength in than you had in the past?

And then just with the big 6, are you seeing -- the revenue's been stable now for 3 quarters.

Any different types of traffic that you can tell that you're getting or not getting from them?

And has that shifted over time?

Yes.

I think we were very pleased to see, within the media area, very pleased to see the growth in video traffic.

That's, of course, where we're focusing.

And that's been, I think, the major change from prior quarters.

And I think, with the big 6, again, we're going to get used more for video.

We're going to get used more for live and linear events within video.

So not a fundamental change there.

Of course, we did do some very large events that we talked about, and those were for some of the giant platform companies.

So not a fundamental change.

Again, it's -- if quality matters, if -- which it does more for video, there will be more business turning to Akamai.

And then just quickly on housekeeping, how much did you spend in buybacks during the quarter?

And then how much free cash flow did you generate?

So we spent about $129 million in the quarter in the buyback.

And our cash from operations was about 200 -- mid-230s.

And our next question will come from the line of Vijay Bhagavath with Deutsche Bank.

It would be very helpful for us to know in terms of what are the sustainable drivers for traffic volumes.

And the reason I ask is traffic volumes have been lower than expected recently, and then now we saw this acceleration.

So what's driving the acceleration?

And is part of the acceleration due to this move to your offpeak pricing strategies?

Is that starting to resonate with some of your top customers?

This way, perhaps, we could see like sustainable -- sustainably high traffic volumes moving forward.

I think the sustainable drivers for traffic are focused around OTT and the growth there.

And for us, to gain share there is the focus on quality, on scale, on affordability, on the functionality that we talked about; low-latency streaming being as current or better than satellite instead of 10 to 90 seconds behind satellite; for the client-side software that provides even better quality and potentially lower price for capabilities like accelerated ingest or connecting directly to our large customer's origin; broadcast operations support.

All these things that Akamai uniquely does I think help us and are sustainable drivers for traffic growth for Akamai.

Perfect.

A quick follow-on would be on the enterprise security business.

I mean, arguably, you've done many interesting acquisitions recently.

So when could we start seeing enterprise security inflect as part of security revenues?

Would it be sometime like first half or second half next year?

So it helps us model the security business as enterprise starts waterfalling in.

Right.

Now a good question.

Very early days.

We've just released our first product, as we've talked about, and we'd like to see it start to make a difference at the end of 2018.

So it's big enough that you'll start to notice, and we'll report it out separately.

Our goal is to have the same kind of growth trajectory we had for web security, which was nothing about 5 years ago and now is on $0.5 billion a year run rate.

So that's the goal, which would mean we'd start to see something meaningful by the end of next year.

And our next question will come from the line of Rob Sanderson with MKM Partners.

I want to dig into the web performance business a little bit.

Growth of 4%, by my calculation, that's pretty soft, pretty significant deceleration from the 8, 9, 10s that we've been seeing recently.

So I got a few questions.

I think the large majority of this business is recurring revenue.

Should we assume that bookings and renewals are both fairly weak here?

Or is there more lumpiness or what we used to call bursting revenues years ago in this business?

Second question is I think this is an area that management has identified as where you could be growing faster.

I know there's been some reorg and some leadership changes over the past couple of years.

Can you kind of comment on your satisfaction or dissatisfaction in those efforts?

And then finally, how would you characterize the competitive dynamic as it relates to the slowdown in growth from web performance products?

Yes.

We principally don't look at it from the product perspective there.

There's such a blurry line between the media products and the low-end performance products.

And so the web performance product number gets caught up in media.

So I don't think it makes sense to look at it that way.

I think the right to look at it is what the Web Division number is, and that grew at 14%.

Of course, security is an important part of that, but so is performance.

And we sell more, have more revenue from our performance products there than we do security still.

So I would say that business is strong and growing at a good clip.

Like all of our businesses, it's very competitive.

And we have seen some burstiness in public sector, which is part of our Web Division customer base.

And as we talked about, there's a little bit of softness there right now.

And that helped contribute to go from a 16% growth rate in the Web Division, to 14%.

So I think that's the right way to look at it.

You don't get caught up in anything that's going on around media and whether they're buying media products or performance products.

Think of it as the Web Division customer base.

That's how we think about it, and it's at a very healthy 14% growth rate.

And our next question will come from the line of Heather Bellini with Goldman Sachs.

I just wanted to follow up on the performance business as well.

I was wondering if you could talk about year-over-year pricing trends in the quarter for performance in particular and also kind of how this compares to what you've seen for the year.

And also, I mean, maybe I missed it, but can you give us an idea of the amount of revenue there that's driven by the media customer overlap and how we should think about that going forward?

Yes.

I think the pricing trends are pretty typical in the performance business.

Every year, the pricing based on traffic comes down per byte delivered.

We do have several new products there that are very successful.

There's the entire security product line that we talked about.

We also have Image Manager, which is new in the last year, getting a lot of success in the web in terms of the web performance products.

We have now the mPulse and the SOASTA products now in the marketplace, getting traction.

We have the mobile SDK for the mobile apps.

An increasing amount of the traffic is going to mobile apps, and so the API traffic there is very important.

So those are new products that are coming, that we're bringing to market.

And that helps to offset the normal decrease in pricing, and so that we actually grow the revenue we're receiving for our performance products in the existing customer base.

And I wouldn't really think too much about the overlap for the low-end performance products that our media customers can buy.

Think about it in terms of the Media Division and what they're doing.

And a lot of that, what we're focused on there is on video.

And then you think about the Web Division customer base, which is pretty much everybody but media and carriers.

That's the vast majority of the Fortune 1000.

And they focus on the web performance products and the web security products.

And we're seeing very strong growth there.

No unusual behavior, I would say, in pricing.

And the good news is there's new products we're bringing to market, not only in security, but also in web performance that are gaining traction.

Great.

I guess what I was trying to get at is just the performance business decelerated year-over-year.

And it used to be a line that I think people thought of as kind of a high single-digit performer.

Are these new products that you're talking about just is that stuff that will help the growth next year?

Or is that something that they were in the market for all of Q3?

I'm just trying to get a sense for the deceleration and how do we think about just that performance line when we think about a sustainable growth rate going forward because that also includes the SOASTA acquisition, if I'm not mistaken.

Right.

I think the new products will help the performance product revenue going into '18.

They are new this year, so they're very early stage in terms of their overall revenue contribution.

But yes, I think you will see improvement in '18 based on those new capabilities.

And our next question will come from the line of Michael Hart with Guggenheim Securities.

There's been a lot of questions that they had already asked, so I wanted to dig into something that hasn't come up as much yet.

In recent quarters, you consistently called out the Asia Pacific region as driving strong growth within your international segment.

I was wondering if you could provide any incremental color on maybe what kinds of customers or solutions or maybe any particular countries that are really driving the strong performance there.

And then also, one quick housekeeping question.

On the Nominum deal, have you -- or guys, can you tell us what the price you paid was?

And I assume it will be funded from cash on the balance sheet.

I just want to confirm that.

Yes.

APJ continues to be a strong growth region for us.

We also had good growth in EMEA, but APJ leading the way.

It's similar kinds of customers and solutions that we sell here, but there's just a lot of growth in that region as a whole and for us in particular.

When you think about it, a lot of the Internet users are there, a lot of the major device manufacturers are there, a lot of the world's major enterprises are there.

So it's not too surprising.

And it's the countries you'd expect that are leading the way, Japan being very strong, Australia being strong, Singapore and Southeast Asia having strength.

So I would say there's nothing really unusual there, sort of what you would expect knowing our business.

And Jim, you want to talk about Nominum?

Yes.

I mean, I don't think it would be appropriate to talk about the price for Nominum.

The deal hasn't closed yet.

The deal is supposed to close kind of in the next couple of weeks.

You'll certainly see it in our Q4 cash flow statement, but I'd rather not comment on it right now until the deal closes.

Our next question will come from the line of Mark Mahaney with RBC Capital Markets.

Yes, I'll just ask one question, please, on OTT.

Is there any way you could quantify what percentage of traffic, what percentage of revenue that currently constitutes and just an outlook, just on the OTT segment or that channel?

Yes, we don't split that out.

But obviously, that segment is growing more rapidly within the entire media portfolio.

But we don't give that split on a quarterly basis.

Our next question will come from the line of Barry Sine with Drexel Hamilton.

Hey, Tom, at your Edge Conference in Las Vegas, I thought the presentation with SOASTA was pretty compelling.

I know you've only had it for a quarter or so.

But could you talk about how that's driving -- I guess that really drives revenue from other products, but what's been the customer reaction from that presentation and then show it using that tool with customers?

Yes.

We got a great response at our customer conference.

For those of you who weren't there or didn't see the video, we offered a free 30-day trial, a freemium capability.

And we were really pleased to see that a lot of customers signed up on the spot to try it out.

And we derive revenue in 2 ways from SOASTA: one is when they buy the service on an ongoing basis because it really helps them manage their websites and apps more effectively and their online businesses more effectively; and then, of course, when they see the benefit they can get by adding new Akamai capabilities, there's a very strong tendency to buy those capabilities from Akamai.

And so we get revenue from our other performance services on top of just the plain SOASTA revenue.

So we were very pleased with the response and really excited that customers on the spot there could turn it on and start seeing what's going on with their websites.

And then a question for Jim on the buybacks in the quarter.

Where you are right now in the aftermarket, the stock was as much as 20% below that in August.

So you had a pretty good opportunity for buybacks.

What was the philosophy there?

What's the philosophy going forward?

Do you have the ability to drive buybacks higher?

The stock is still relatively undervalued where it's been historically.

Yes, I think we've talked about it before that the way our buyback program works is it's an algorithm, and it basically is an algorithm that buys back more shares and spends more money when the stock price is at lower valuation levels and it buys back less when the stock price is in higher valuation levels.

So you certainly saw that the uptick in Q3 was a function of our belief that, obviously, the company's valuation was low.

As you can imagine, we're fortunate to have the benefit of generating a lot of cash for the company and having a fair amount of cash available to be able to do both buybacks and M&A.

And so buyback has been an important component of our capital kind of strategy.

As you can imagine, one of the things we've talked about is that we've been acquisitive in areas, most recently with SOASTA and next with Nominum, and so we'll modulate the buyback program based on also firepower that we want to have available for M&A.

But we view Q3 as an opportunistic time to spend more on the buyback.

Our next question will come from the line of Jeff Van Rhee with Craig-Hallum.

So 3 brief ones.

One, if you could just touch on Nominum's pre-acquisition growth rates, to give us a little context of the trajectory of the business.

And then the second one, around network utilization overall, I know you don't quote a number, but as you stand now, how do you -- how are you positioned with respect to overall network utilization than maybe compared to the last 2, 3 years?

And how do you envision maybe the next year or 2 playing?

Maybe asked differently, do you intend to operate the network at a higher utilization rate?

Yes, the first one is that, as you can imagine, Nominum's growth impact on Akamai is going to be fairly minimal early on because of the impact of deferred revenue and purchase accounting because a very large percentage of their business is a license and perpetual revenue stream, where we're not going to be able to recognize their revenue until we resell their business once we bring them on.

But their business on an organic basis was growing kind of the high single digits.

But combined with Akamai, we think that we're going to be able to integrate that, as Tom mentioned, some of the capability into our security offerings and increase the revenue growth rate.

It's going to take a while to anniversary this deferred revenue impact.

So you should expect to see their business kind of begin to accelerate, from a growth rate perspective, throughout 2018 and more notably, into 2019.

Yes, and on the second question, I think we're in very good shape there.

We talked about reasonably doing 60 terabits per second of traffic, and I can tell you there was plenty of headroom on top of that available for us.

And as you may know, we put a ton of work into always improving the efficiency of our software and platform, and that helps us get more bits per second out of every dollar we spend on co-lo or CPU.

And so that -- we have plenty of room now, and as we go into next year, what we have, we'll be able to produce even more.

So that's an area where we're pleased with the progress we're making, and I think we're in pretty good shape.

Maybe if I can just try one other question here.

With respect to last quarter, you touched on your intention to drive some differentiated pricing.

I think that's the words you used.

As you look at the contracting structures and pricing methodologies, particularly on media this quarter, can you color that in a little bit?

What did differentiated pricing look like as it sort of played out in the field?

Yes, sure.

Things like per-subscriber pricing, per-geo or per-carrier pricing, pricing depending on a particular customer's needs, so just a lot of flexibility there so that we can align our pricing to our customer with their revenue models, so that it takes a little bit of the risk off of them and that they can run their businesses more effectively.

So that's the kinds of things that we are doing in the marketplace as customers come up for renewal.

Our next question will come from the line of Sameet Sinha with B. Riley.

A couple of questions.

And just looking at fourth quarter guidance, if I make my assumptions about revenues from SOASTA and Nominum, it seems like you're guiding toward the midpoint, a slight slowdown from the third quarter at about 4%.

I just wanted to understand the dynamics in the fourth quarter.

I can understand -- you spoke about e-commerce and how this is an unpredictable quarter.

Is there some impact potentially from the presidential elections on a year-over-year basis as well?

So if you can provide some clarity.

And secondly, last call, you had spoken about just managing costs and cost reductions.

Can you talk about your initiatives in that direction and how we should think about it as we head into next year?

Sure.

So on the Q4 revenue guidance, actually, we think this is a pretty robust guidance.

We had a very, very strong Q3 actually and as we mentioned, significantly better than what we expected going into the quarter.

And as we mentioned, that the quality seasonality plays is a large role in where Q4 revenue lands.

So we think it's a reasonably strong guidance.

I think it is fair to say that we had an exceptionally strong media quarter last Q4, so there's a pretty significant difficult compare in the media business year-on-year.

But I think, as I color the guidance, that I think that if the holiday season and the e-commerce season is strong, we would expect it to be kind of towards the higher end of our guidance, and if it's not as strong, maybe to the lower end.

And kind of to call it the midpoint is, call it, having a solid holiday season.

And again, last Q4 was a very strong holiday season and was a very strong e-commerce season.

So tough to predict what's going to happen.

Those things are somewhat out of our control, other than making sure that we're the provider of choice for that business.

And as far as cost efforts, that I think Tom commented a little bit on it, we are continually focusing on costs, in particular around driving more network costs out of the platform, driving bandwidth costs down, getting more efficiency out of co-location, software efficiencies to get more kind of throughput out of our servers.

So it is a lot of activity that's not something new.

It's something that we continue to do.

You kind of heard from my guidance for the fourth quarter that we are expecting an uptick in gross margins from Q3 to Q4.

And inevitably, some of that is seasonally even revenue grows like it does Q3 to Q4.

But you can also read into that, that we're making continued progress on our network cost efficiencies.

And we do the same thing on the OpEx side, that some of what you've seen here recently, with some compression for the company's EBITDA margins, has been heavily driven by acquisitions that we've done, and it takes a while to absorb and scale these acquisitions that we think are the right things for the business.

And then there are targeted areas of investment that we want to make sure that we're investing in even with what's been happening kind of more recently with our media business slowing.

So we're -- this is a pretty balanced approach we take to the business, focusing on managing costs in the network and focusing on managing and kind of prioritizing costs around OpEx as well.

Our next question will come from the line of Brandon Nispel with KeyBanc Capital Markets.

2, if I could.

I was wondering if you guys could give us a mid-year update in terms of adjusted EBITDA margins within the media and web performance and security categories.

And then just looking at the growth rates, coming back to a little bit higher this quarter in the media side, is your expectation still to achieve double-digit growth in 2018?

Yes.

The first one is we provide an annual update on EBITDA margins.

Again, we don't manage the company that way.

That is just a -- we do an annual exercise to go through allocating the company's costs into the various businesses.

And so a media update would kind of be not appropriate.

I'll certainly provide an update on the margin profile of our businesses and probably, more notably, even our divisions at the upcoming Investor Summit that we'll do in Q1.

But I think it's fair to say, call it, in aggregate, without even having done the model, that some of what you're seeing is probably a modest compression in EBITDA margins for media and probably stable margins within the performance and security business.

And with the growth rates, we are looking to get to double-digit growth rates for the company by the end of '18.

We are -- we believe that we can do that, and looking forward to that next year.

And just one quick one to follow up.

The compression in the media and some outperformance in the web performance, is that relative to the margins you laid out in your Investor Day presentation?

Yes.

And again, as I said, that we will update them again in Q1, that -- because we don't do it and don't manage the business that way.

I'm just giving you some general qualitative color.

But yes, it would be compared to what I shared in Q1.

And our last question will come from the line of Colby Synesael with Cowen and Company.

As it relates to video, which you've stated as obviously a driver of traffic growth, I'm wondering if we could just tease that out a little bit further, to try and get a sense of where that's coming from.

Maybe you thought it was just a bucket of either social media or broadcasters or those providing subscription OTT players.

Or maybe there's a different way that you would think about it.

I'm just trying to get a better sense of where that's actually happening or where that happened in the third quarter.

And then secondly, I think, last quarter, you mentioned that obviously in -- you're focusing on the top 250 media customers.

By definition, you're also deemphasizing those below that.

I was wondering if is that anywhere done with that already and if that's leading to any type of margin improvement that might be happening with the company as well.

Yes, on the last question, you want to think about it as reallocation of resources, which we've done a bunch of in the process of doing to get greater focus and more return for the investment.

As Jim talked about, we are very serious with how we invest every dollar.

In terms of the video traffic growth, we saw strong traffic growth really across the board, but I think you want to think of it as more broadcaster OTT high-quality video than social.

And I think, in the long run, OTT will be -- the big bits will be driven by the broadcaster, the high-quality long-form content.

All right.

Thank you.

In closing, we'll be presenting at a number of investor events throughout the rest of this quarter.

Details of these events can be found in the Investor Relations section at akamai.com.

And we thank you all for joining us, and have a wonderful evening.

Ladies and gentlemen, thank you for your participation on today's conference.

This does conclude the program, and we may all disconnect.

Everybody, have a wonderful day.