Edgio Inc (EGIO) 2009 Q3 法說會逐字稿

Good day, ladies and gentlemen. Welcome to the Third Quarter 2009 Limelight Networks Earnings Conference Call. At this time all participants are in listen-only mode. At the end of the prepared remarks, we will provide instructions for those interested in entering the queue for the question-and-answer session.

I would now like to turn the call over to Paul Alfieri, Senior Director of Corporate Communications. Go ahead, Paul.

Thank you and good afternoon. Thank you for joining the Limelight Networks Third Quarter 2009 Financial Results Conference Call. Speaking today will be Jeff Lunsford, Chairman and Chief Executive Officer, and Doug Lindroth, Chief Financial Officer. This conference call is being recorded on November 4, 2009, and will be archived on our website for approximately one week.

Some portions of this conference call may include forward-looking statements as defined in the Private Securities Litigation Reform Act of 1995. Forward-looking statements are all statements that are not strictly statements of historical fact such as statements regarding future events or future financial performance including but not limited to statements relating to Limelight Networks market opportunity and future business prospects; guidance on financial results; statements concerning anticipated future growth and profitability; as well as management's plans, goal, strategies expectations, hopes and beliefs.

These forward-looking statements are subject to risks, uncertainties and other factors that could cause actual results to differ materially from those contained, projected, or implied in the forward-looking statements including the inherent risks associated with litigation, particularly intellectual property-based litigation. Reported results should not be considered an indication of future performance. Factors that could cause actual results to differ are included in the Company's periodic filing with the Securities and Exchange Commission.

I would now like to introduce Jeff Lunsford.

Thank you all for joining us. Today Limelight Networks reported $32.5 million in revenue for the third quarter of 2009. During the quarter, gross margins expanded by 1 percentage point and adjusted EBITDA remained steady. Operating highlights in the quarter included the launch of Limelight next generation XD platform with adaptive intelligence, which provides advanced levels of performance and increased insight into real-time Internet and delivery conditions. And the launch of LimelightREACH and LimelightADS, two new services that provide turnkey capabilities for customizing and monetizing media delivery on mobile and other connected devices.

The Internet continues to transform and advance the way people live, work and play around the world. Tomorrow's Internet will involve more real-time data flow and content consumption across a growing mix of computers, laptops, netbooks, game consoles, television set-top boxes, and mobile devices; the combination of which will allow consumers to enjoy content, access information, and conduct transactions at anytime from anyplace.

With our XD platform and mobility monetization solutions, we are positioning Limelight Networks to be a core provider of cloud-based services within this transformed world, helping forward-thinking media, entertainment, enterprise, and government organizations leverage the next generation Internet's capabilities to differentiate from their competitors and better serve their constituents.

The $1.4 billion CDN market is healthy and continues to grow. This market operates in two distinct segments. First, the large object segment where media and technology companies use CDNs to deliver large video, music, and software files; second, the small object segment, where all types of companies use CDNs to deliver their entire websites, including small files like JIFs and JPEGs, and to accelerate and end transactions that involve dynamic data.

In addition to these two markets, a CDN is an attractive supporting platform for many value-added services that could be delivered over the Internet otherwise known as cloud-based services.

Limelight's strategy is to, number one, continue to build and distinguish ourselves as the most capable, high-performance CDN in the large-object space, which is our traditional area of strength. Number two, expand capabilities with LimelightSITE, XD and future services that will help us penetrate the small object and enterprise markets; and, number three, launch higher-margin value-added services such as LimelightREACH and LimelightADS that leverage our high-performance CDN platform. At present, our business is gaining traction along all three of these growth vectors.

At the end of Q3, our customer count was approximately 1,370, which includes mobility and monetization solutions customers. We signed contracts with over 100 new customers in the quarter but also continued to attrite less attractive ones. The value of our customer relationships increased slightly in the quarter as average annualized revenue per customer grew to $95,000 in the quarter.

We are pleased with our overall bookings performance, which is an increase over Q2. While the large object market is characterized by higher volumes and greater price pressure, this segment continues to be important, as it is where we get our scale. It is also the segment where we forge new ground with innovative customers utilizing new, advanced, consumer-oriented technologies.

During the quarter, we added or expanded relationships with customers in this segment such as Nintendo in the gaming sector, ABG and Checkpoint in the software sector, Guvera and Jango in the digital music sector, National Public Radio in the online radio sector, and Channel 4 in the UK in the online television sector.

Our small object capabilities are growing rapidly also, albeit from a small base, and we are signing up more customers for whole site, small object, and other enterprise-related services each quarter. During the quarter, we were pleased that companies such as Overstock.com, Blog Talk Radio, the Not Kelly Blue Book, Nikon, and Online Marketer Maximizer selected LimelightSITE and LimelightDELIVER to power their online businesses.

Since its launch, we have signed approximately 20 customers for LimelightSITE services, and we are optimistic about the growth prospects for this service, going forward, particularly with the anticipated launch of SITE 2.0 early in 2010.

Our value-added services are also growing from a small base. These include mobility and monetization solutions, highly specialized consulting services and other services we have in the works for launch in 2010.

In the third quarter, average traffic levels continue to rise, and we again experienced new records for peak traffic on the network. In October, Arbor Networks published their latest Atlas Observatory Study, which estimates total Internet traffic at approximately 14 terabits per second. The report names Limelight Networks as one of the Internet hyper giants along with companies such as Google and Facebook as one of the next-generation networks, which are gaining traffic share from traditional telecommunications providers.

Now I'd like to turn to more operational highlights. Just after the end of Q3, we announced our new XD platform. XD represents a complete upgrade of the entire command and control software layer that operates our global network. Using patent-pending innovations built by the Limelight engineering team over the past two years, XD positions us well for the future in four key opportunity areas -- delivery, storage, analytics, and computing.

While today we compete mainly in the delivery market, XD provides us with a foundation to expand into other attractive value-added services. One of the key innovations the XD platform provides is the ability for our network to see past the edge server and get a real-time view of last mile network and end user device conditions.

The XD platform can also automatically tune object delivery on a per-connection basis in real time. This helps us ensure a brilliant user experience across a wide variety of devices, even under sub-optimal or changing network conditions.

We believe the capability we gain from the XD platform to actively manage delivery after an end user has connected to a Limelight edge server will provide us with a competitive advantage in the new world where myriad access devices are connecting to content sources over a myriad of dynamic Internet connections.

Additionally, just after the close of the quarter, we announced another key global telecom reseller relationship with the leading carrier in Canada -- Bell Canada. Global telecom companies increasingly want strong CDN capabilities in their portfolio, but this is a complex business that requires a great deal of capital investment and technical expertise. We believe that as a network-based CDN, Limelight can offer telecommunication providers a unique model for partnership. At the core of the relationship with Bell is deep technology integration between Bell's optical network within Canada and Limelight's global delivery platform and software.

This deep integration enables Bell to immediately enter the Canadian market with the proven industry-leading Limelight service running locally over their own network. Bell can immediately expand their product portfolio with higher-value CDN services as well as an array of applications from the Limelight Networks' ecosystem such as content hosting, ad insertion, and performance monitoring.

For Limelight, this close coupling with Bell's network gives us extended reach throughout Canada as well as the benefit of a sales team with deep local knowledge and existing relationships. We also gained greater insight into the Internet traffic patterns of Canadian broadband users, which lets us continue to refine and improve delivery.

I will now turn the call over to Doug, who will take you through the financials and other key points.

Thanks, Jeff. During the third quarter, we reported revenue of $32.5 million, up 1% compared to revenue from Q2 and down 2% from the same period last year. We reported third quarter adjusted EBITDA of $5.8 million compared to $6 million for Q2 and $5 million for the third quarter last year.

Our adjusted EBITDA increased to 18% of sales from 15% in the same period last year, and flat to last quarter. Our GAAP net loss was $5.2 million, or $0.06 per basic share compared to a net loss of $15.4 million, or $0.18 per basic share in the same period last year.

We also reported a third quarter non-GAAP net loss before stock-based compensation and litigation costs of approximately $600,000, or $0.01 per basic share compared to a non-GAAP net loss of approximately $700,000 and $0.01 per basic share last quarter. Please refer to the tables included in our press release for the reconciliation of GAAP measures to these non-GAAP measures.

During the third quarter, Limelight's international operations represented 22% of total revenue, which was an increase from 19% in Q2. Gross profit margin, which includes both depreciation and stock-based compensation was 36% during Q3, up from 35% last quarter and Q3 of 2008.

Cash gross margin was 56% for Q3, flat to last quarter, and down from 57% in the same period last year. As Jeff highlighted earlier, we once again set new records for peak traffic on our network, and we believe our ability to maintain our margins in a challenging economic and pricing environment is evidence that Limelight Networks has a global scale to continue to be one of the market leaders. Our gross margins did not decline to the extent previously guided, as we were successful at using our global scale and network-based architecture to reduce delivery costs.

However, at the mid-point of our revenue guidance for the fourth quarter of 2009, we anticipate gross margin compression of approximately 1 to 2 points from Q3. We expect to see a sequential decline in our gross margins primarily as a result of network expansion activities that took place during Q3 and will continue into Q4 partially offset by lower average delivery costs.

During the third quarter, our operating expenses were $17.1 million, an increase of approximately $500,000 from last quarter and down $11.3 million from the same period last year. Our operating expenses increased from last quarter as a result of having a full quarter of Kiptonic expenses and due to an increase in our sales and marketing costs.

Our operating expenses decreased from the same period last year due to a decrease in litigation costs, damage accrual, and bad-debt expenses. We anticipate that our fourth quarter operating expenses, excluding litigation costs will increase by approximately $1.7 million.

The forecasted increase is primarily the result of additional headcount in our sales organization, increased marketing program, and an increase in bad-debt expenses.

Total depreciation and amortization for the third quarter was $6.6 million, down from $6.7 million in the second quarter and down from $7 million in the same period last year. Depreciation and amortization in the current quarter includes $6 million of network-related depreciation.

Stock-based compensation expenses for the quarter were $4.4 million compared to $4.3 million last quarter and the same period last year. Third quarter interest earnings were approximately $300,000 compared to approximately $300,000 for Q2 and $1.2 million for the same quarter last year. The reduced interest income is associated with lower market interest rates and lower average cash balances from the same quarter last year.

Moving on to the balance sheet, our combined cash and marketable securities balance on September 30th was $152.8 million, down from $164.3 million in the second quarter. The decrease in cash and marketable securities is primarily related to our capital expenditures and a large contract prepayment we made to one of our global network providers. The prepayment was related to our backbone expansion as well as a long-term bandwidth purchase agreement.

Capital expenditures for the third quarter were $10.6 million compared to $5.3 million for Q2 and $7.9 million for the same quarter last year. As we previously guided, our capital expenditures increased in Q3 as we continued to expand the scale and capacity of our network.

Day sales outstanding for the quarter were 77 days, up from 73 days the previous quarter.

Regarding guidance for the fourth quarter of 2009, we expect to achieve revenues in the range of $32.5 million to $34 million. As I mentioned, we expect 1 to 2 points of gross margin compression at the mid-point of our guidance range, and we expect operating expenses, excluding litigation costs, to increase approximately $1.7 million.

Stock-based compensation expenses for Q4 are expected to be approximately $4.4 million. Capital expenditures are expected to be in the range of $5 million to $6 million.

With that, I will turn it back to Jeff.

Thank you, Doug. At this time, Operator, we will open the line for questions.

David Hilal, FBR.

Great, thank you. Jeff, first, the net customer is flat sequentially. Obviously, two drivers there -- the churn and gross customer adds. Can you talk about those two pieces that make up the net number?

Yes, sure. A little more color on bookings -- so we continue to work through what we'd call -- we'd characterize as higher-than-normal churn, smaller companies, primarily, that are just having difficulty with the tough market conditions. And we, however, saw, as I mentioned, a very solid bookings quarter. And two ways to think of bookings -- one is monthly minimum revenue, and then our market segment has also shifted to where folks are making annual bucket commitments, so we think about that as an annual contract value rather than a monthly contract value.

And we had a nice increase in the annual contract value from quarter-to-quarter, so we felt like, from a forward-looking indicator, bookings were positive, but we're still working through, a, raising the bar on the quality of customer, driving down DSO; and b, trying to focus on the higher-quality enterprises where we can cross-sell and upsell value-added services as we expand our product portfolio.

So maybe talk about that, Jeff, as it relates to guidance. So bookings, it sounds like, were pretty decent, and if you look at deferred, deferred is a nice sequential move. And then you look at CapEx, was more than you had planned to spend, which usually is a predictor of maybe what you guys are seeing in the pipeline. But yet guidance is just mediocre, at best. So maybe you can reconcile that.

Well, yes. So CapEx was pretty light in the first half of the year and, you know, our buildouts are somewhat lumpy. So I wouldn't read too much into just having a large Q3. If you kind of look at it for the first three quarters, it's probably about the right pace. And as we always do with our guidance, we go and end any quarter we have about, I guess, sort of 55% of our revenue under contract, and then the other 45%, 50% is variable based on usage. So this is not 90% plus visibility SAS business that you and I are used to, David. It's a little more opaque, and we're trying to provide the right level of forward guidance given what we have today.

Okay. And then, finally, Microsoft -- in your Qs you always kind of give us a little more detail, and I wanted to understand, maybe you can tell us now what the Microsoft contribution was in the quarter and, specifically, that trended down last quarter, and I'm wondering how that is and will trend, going forward.

They were 15% of revenue during the quarter.

Okay, and then, going forward --

Which was flat to last quarter.

How do you think that trends, going forward?

Well, on a macro multi-year basis, Microsoft -- we helped them build an in-house CDN. And so you should assume -- and they've been publicly talking about shifting. I think the last thing they said was they want 60% of their traffic to run on their own CDN in 2010. I believe that was the number, but don't quote me on that. You need to fact-check that.

And so, over time, Microsoft will decrease in size as a percentage of Limelight's revenue, and that's been the plan, by design, all along. Now, we are in there as a very, we think, solid partner of Microsoft's, working with them on multiple fronts and are going to do everything we can to continue to be a primary service provider to them for the traffic that they do not deliver over their own CDN.

Mike Olson, Piper Jaffray.

Just a couple of quick ones here -- you said international is 22% of revenue, up from 19% in Q2. Do you have what the international revenue growth rate was on a year-over-year basis?

Year-over-year is about 33%.

Okay. And then on the expense side, G&A has been running in the low 20s as a percent of revenue for the last couple of quarters after being in the high 30s to 40s in '08. Is this the new run rate for G&A, is in the low 20s? Is that fair to assume?

Yes. What I would look at is we're only guiding to next quarter, and the one thing that I put in my prepared remarks talked about increasing in Q4 potentially for higher bad-debt expenses. So we've had two quarters in a row where I believe our team, across the board, has done a really good job working with our customers in collecting on receivables. So as we've been able to do that, where we are booking some very large bad-debt expenses, Q4 of last year and Q1, we feel that we had a good reserve, or an allowance for bad debts and therefore didn't require a lot of additional bad-debt expense. So we are forecasting that to increase in Q4 just to be back to more of a normalized run rate.

So, other than that, we do plan, over the long term, to leverage our G&A as we continue to grow, so we should see continued leverage and G&A expenses not growing as a percentage of sales.

Okay, and then the last one -- could you talk generally about the pricing environment and how you're expecting pricing to trend over the next couple of quarters?

We saw the same kind of year-over-year price range that we saw last quarter. I think we said between 30% and 35%. And that's if you take our traffic revenue and divide it by the bits or the GBs delivered, gigabytes delivered.

So we believe -- we continue to believe that there has been a structural re-pricing in the industry, and that in 2010 and potentially even in this quarter, we should see a slow and gradual return to the more historic norm of 20% to 25% year-over-year price declines rather than the 30% to 35% that we're working through today.

But that's another one of those that's reasonably opaque, and when you look at price lists in our industry, as folks' volume grows, their unit price -- they get a lot of benefit of scale, and so it's really hard to just look at the data set, because when you factor in the growth of each customer, one customer might get 60% better pricing the next year. But if they're 200 times larger that doesn't mean that prices in the industry went down 60%.

Kerry Rice, Wedbush.

Just a couple of questions as I think about your guidance. One is, what's changed from last year Q4? Because you saw a nice ramp up in Q4 in revenue, and, arguably, in a deteriorating economy, and in this quarter are in Q4 -- well, maybe -- we'll start with that, and then the second part is when you announced your Q2 earnings, you said that pricing had reverted back to this historical norm of 20%, 25%. So I was going to see what happened between July and the end of the quarter, now that you're saying there is a 30% to 35% decline in Q3.

I don't believe we had said that it had reverted back on our last earnings call. We had talked about that we saw during Q2, the 30% to 35% year-over-year type of price declines. So I'm not sure where you're getting that. I think we have been seeing those similar types of price declines and price competition within the market, certainly, over the last six months.

Yes, I think we did give the color commentary the last earnings call about we felt like, in the back half, that we would hopefully see a return to the more historical norms, but we did not see that this quarter, if that's what your question refers to.

Somewhat. Yes, I'll have to go back and pull out where I saw that. But maybe, then, what's changed between this Q4 and last Q4 that you're not seeing that same pop that you did? Is it simply a pricing issue, because it sounds like traffic is growing pretty good.

Well, I think what I'd say is last Q4 we had three or four things that happened in the quarter that were nice over-performance events. If you look back to our guidance last Q4, it wasn't -- I think our results were nicely ahead of the guidance range, and there were three or four pretty substantial -- $0.5 million to $1 million revenue events that happened in the quarter that led us to over-perform in Q4 of 2008.

Okay, and then one final question -- you talked about churning some of the less attractive customers -- when do you think you'll get through that process? Because you've been churning these less attractive customers for a few -- at least for a couple quarters now. Do you see an end to that in the near -- next quarter or in the near term?

I think it's hard -- that's sort of like asking us to divine on the economy, and there's just a higher percentage than normal of businesses that are struggling. A lot of these online guys who raised some capital when markets were hot and now can't get their B&C rounds of funding -- folks like that. And it's just hard to say. What we've definitely done is raised the bar of the quality of customer we'll accept, to begin with. And tighten the reins on payment requirements and just trying to focus on being disciplined in that manner.

Derek Bingham, Goldman Sachs.

A question on traffic -- when you think about it, year-over-year, traffic growth may be on a per-customer basis. Are you seeing those on your network accelerate, or have things been pretty stable?

I would say stable year-over-year traffic growth. You have to take it customer-by-customer. We have some customers where you can clearly see their business model as gaining traction, and they are showing very solid growth. We also have, as you know, Derek, most of our larger customers use us in a multi-CDN architecture, so sometimes we could see traffic growth, and it's just because we're getting more share of their traffic allocation. But, all in all, we are still seeing general growth across our customer base.

Okay, all right, but there's not early signs of an inflection upward that you see, so far, associated with increasing bit rates or something like that? It's been kind of -- just kind of steady pace of high growth?

Yes.

In terms of how your unit pricing for bandwidth has trended versus the unit price declines that you see on the revenue line -- your cash gross margin has been stable, so maybe that's the answer there, but is there any reason to believe this relationship will change, looking ahead, one way or the other?

Well, we think at scale, as part of our investor presentation, we have our target model at about a $300 million revenue run rate. We think we can be running at about 45% gross margins, which would be 7 or 8 points higher than today. And that's really the benefit of scale -- as you get a broader customer base, you fill more valleys of inactivity with traffic and thus paying customers, and you really don't have any incremental costs across the platform for those customers. And so you do see -- the ability to run the network at a higher overall level utilization, which helps drive up margins. And then also launching and getting traction with these value-added services, which are higher margin than the traditional core CDN business, should help us there as well, Derek.

Okay, and just one more, if I could -- on the service provider relationships or relations with hosters or other kinds of channels, could you talk about how that's trending, maybe as a percentage of your revenue or any other color you could give on anticipating more of these types of relationships and becoming a more important part of your business?

Yes. So channels are definitely growing as a percentage of revenue. We don't really break it out as a separate reporting line, but we've announced a couple of large telecommunications providers and, as we said in our comments, there's a pretty natural fit there. Most of these folks don't want to go off and try to get into this business themselves. It's a very complex software-oriented business, and so there's just very good synergies, especially Limelight being a network CDN. So that's clearly a channel that's promising for us, and we do have relationships with hosters as well. And that area is a good leverage point for a company like Limelight that's looking for greater distribution opportunities.

Chad Bartley, Pacific Crest.

I wanted to ask a follow-up on pricing -- I appreciate you guys quantifying the declines you're seeing. Jeff, last quarter you talked about the telcos, specifically, driving that and being more aggressive on price. Obviously, Akamai has talked about a slightly different strategy on price. So can you give us an update on the competitive landscape and telcos being more aggressive, Akamai being more aggressive -- who is driving that acceleration and that decline?

You really have to take it account-by-account, Chad, because we -- certain accounts have certain needs and gravitate towards baking us off against whether it's a telecom provider or a competitor like Akamai. And if you look at our gross margins, they've maintained in the mid-30s here, and the buy side in our business is sophisticated, especially the larger buyers, and they understand our costs, and they know we're building faster servers, improving software efficiency, and driving down our own bandwidth costs, because these guys also buy raw transit.

And so I think it's more of an industry thing, and then the comment on the telecommunications providers is that they actually get to have the ability to monetize delivery in two ways, whereas we just get paid for the -- let's call it the on ramp to the Internet -- they actually get paid by an ISP that might be buying backbone connectivity from them if they're a tier 1 provider. So they have a way to subsidize their CDN offerings slightly, and that's the systemic repricing that we talked about -- have talked about for a couple of quarters here -- that we think has flowed through the market and should, we think, subside because we're now comparing our year-over-year numbers to pricing that was affected already by the entry of telcos into the market, which really happened about two years ago.

A quick follow-up then, thank you, Jeff, for that. Specific to Akamai, they talked about a change in their strategy last quarter. I don't know if it exactly played out on the pricing side as they thought or that many investors thought. Did you notice a change in their pricing strategy in Q2 or in Q3, or have they been pretty consistent?

We think that we've had fierce competition with Akamai for eight years, and there was nothing dramatic that changed from our perception about their behavior last quarter.

Donna Jaegers, D.A. Davidson.

Just a few quick ones -- on international, the 33% year-over-year gain -- what were the drivers for that, and how much of that was of currency benefit?

The currency piece was very little. The majority of our international revenue today is denominated in US dollars, so it's a small piece on the currency side.

Great, so the 33% growth, then, anything particular driving that?

No, I would just say continued good customer expansion throughout both Europe and Asia. So the market has been performing well. We've expanded there some of our network expansion activities earlier in the year was bringing out more servers in locations in Europe, and I think with better network performance drives more traffic and, ultimately, more revenue.

Great, great, and then on the XD platform, I was curious if you could talk about the number of beta customers that are testing that and those normal sales cycles that you would expect for that product?

Sure, we've had very -- between 5 and 10 -- I don't have the exact numbers of customers trying various components of the XD platform. XD is an entire platform, and this is, as you know, a complicated system. And so we might have some customers testing one feature and other customers testing another feature of the platform.

And the sales cycles are 90 days to six months in this business. Internet kinds of delivery has become mission critical, their customers expect either flawless experience, if it's streaming video or music, just like they get over the television. So our customers are very careful when they're shifting traffic to a new platform. And if it's an eCommerce customer, any kind of transaction disruption obviously is very serious. So this is now a mission-critical business, and platform shifts take a little bit of time. People test them thoroughly, and then when they see something they like, they will migrate traffic to it, over time.

Great, and then just one last quick one -- legal expenses -- Akamai, I think, appealed last quarter, although I haven't seen the details in their Q yet. Can you talk a little about that appeal and any sort of spending increases that we should expect on your part?

We can only point to dates on the calendar, and we think the hearing will be midyear 2010, possibly a little earlier than that. We're just not sure. And costs around the hearing will, of course, go up a little bit. But the good news, if there is any good news with litigation, is that the appeals process isn't as expensive as the lower court litigation process.

Sri Anantha, Oppenheimer.

Jeff, when I'm looking at your domestic revenue, it seems to be declining, and it's still more than a sequential on a year-over-year basis. If you were to break down that into [three] pieces, how much of that decline is being driven by churn, and how much is it just repricing your base when contracts come up for renewal?

I think you'd see this across the industry, and I think it's pretty much pricing, not so much churn. When you look at the customer count number, most of the churn is small customers that aren't real material. We haven't lost major customers in a continued pattern here at all. We preserved the bulk of our major customers and continue to grow market share within them. But you're seeing the bulk of the price pressure in the CDN market is in the large object half of the market that I talked about earlier, which is really 90% of Limelight's business. And Limelight's business was 85% in the US. So when you compare us to others, you'll see the same kind of year-over-year domestic revenue decline. And other folks, where if you had visibility into their businesses, and it's very much driven by the year-over-year price compression.

Got it. And on CapEx, last time we saw such a huge increase in CapEx, we saw a substantially nice increase in revenue in the subsequent quarter. So is this increase in CapEx in anticipation of some traffic commitments or traffic levels that you are expecting in 4Q and beyond?

I think you might have dropped off when we answered David Hilal's question to that effect and, really, you should just look at the blended first three quarters of the year. We were lighter on CapEx than normal in Q1 and Q2, and so Q3 was higher than normal. But if you look at the first three quarters, it's more like a normalized run rate. Our CapEx just tends to be a little bit lumpy. So I'd really look to the revenue guidance to think about how we're viewing Q4, not to the CapEx.

Got it, and in your prepared remarks, I think you talked about success you're seeing with your value-added services. Could you quantify what percentage of revenue is coming from value-added services today and from what kind of verticals are you getting traction with those services? Thanks.

Sure. So it's not small enough to be meaningful today. We will -- in 2010, we are contemplating breaking it out so you can see some color there, because we do think that that's something investors will want to know about -- the growth rate and the different margin profile of these services.

And the second question was on what types of markets? Well, the mobility and monetization solutions are largely large media focused, helping large media companies and enterprises -- actually take content, deliver it over a mobile device wrapped in ads so that it can be monetized. And so we're seeing good traction there with large media companies around the globe, most of whom are already Limelight customers. But a few of them actually were mobility and monetization customers when we acquired that business.

And I think another interesting area is virtualized storage, which is more of a core offering we've already been operating but just selling to our existing customers. And we're seeing growing interest in that as well. And then the consulting business, we have very talented folks who have a very specialized set of skills, and as quickly as we can hire more talent, they get booked up. So that's a very promising area of the business also.

(Operator Instructions)

I think you had Michael Turits next, Operator?

Michael Turits, Raymond James.

Last quarter you said you thought that in the back half you'd return to sequential growth. You did this quarter -- I think it was up 1%, and it seems like at the midpoint of your guidance, it's about 2% sequentially. Is that about what you were looking for, is that -- it's growth, but it's still pretty modest. Is that about what you expected in terms of the back half recovery? Or is it slightly less and, if so, why?

Well, we guided -- we came out, I guess, in Q3 at the midpoint of our guidance range, so I'd have to say that's what we expected, because we guided to that range. And we'd certainly like to be seeing greater growth, Michael, and, again, it's really a function of the pricing dynamic. We continue to see record traffic levels each quarter, and that -- I think you've seen our investor slide where we show the graph of the traffic growth. It's phenomenal. And so we think the return to higher revenue growth is going to be driven by, a, when the pricing dynamic changes and reverts back to historical norms and then, b, as we get better penetration with these value-added services -- those two things will allow us to get the top-line growth back to where we'd like to see it. Of course, 1% sequential is not why we show up to work every day. We'd certainly like it to be higher.

What's the dynamic that makes it return to historical norms? What do you see that suggests that should stabilize at some point?

Well, it's the discussion of the telecommunications providers and the impact they had on the market, which is a one-time impact. It just takes a while for that to flow through all your customer contracts.

And then I guess you said that -- and you're right -- when you would return to sequential growth. When do you expect the domestic actually returns to sequential growth?

It's really the same answer because the bulk of the things I described were still -- 78% of our revenue is domestic, and we're still primarily in the large object space while we're growing into the small object enterprise and whole site business that's growing from a small base. So the answer I just gave you really applies to the domestic part of the business.

Okay. And then on the gross margins, you'd said that you were looking for 1 to 2 points down this quarter. You kept it flat on cash, and I guess it was up slightly on -- slightly up on both cash and noncash gross margins. Did you make less of an investment that would have had that kind of step up in COGS this quarter than you anticipated? I know you got some efficiencies, also, but did you also invest less and spend less on a COGS side than you expected?

Some of it was timing. When we do some of our expansions, when we're bringing up a new location, a new colo, we order the equipment. The equipment then gets deployed to whichever location, and some of it's just timing from where we think when equipment is going to first come in here to our warehousing operation for configuration and then get distributed out to whichever location it's going. So oftentimes it's just a timing issue, and so we did think we were going to have more deployed and incur some additional colo and rack-type costs than we did in the quarter. So a lot of the equipment was here and got deployed towards the end of the quarter, so we didn't have as much of the stepup in COGS that we initially had thought, going into the quarter.

So hard to say, but if you'd had the same anticipated investment, would you have had still that 1 to 2 points down?

Yes, it's tough to say because, again, it's just on timing, and so it's really -- that's why we put it into the 1 to 2 points for Q4. So it is a little bit difficult, but I think that's about right.

I'll squeeze in a last question -- I wasn't sure what your guide was on this -- is your sense that then after the opex moves up for the bad-debt expense in 4Q does opex go down after that, or does opex kind of hold flat at the dollar level for a while?

It will, as a dollar amount, it will increase, over time, as we grow. But I would expect it to not be growing as a percentage of sales on a yearly basis. There will be times when in a quarter it may bounce around a little bit but more of a trend, we expect to see it to decline as a percentage of sales.

Operator, at this time, we have time for one more question.

Sameet Sinha, JMP Securities.

Could you discuss pricing trends, domestically and internationally? Do these tend to go lockstep, or are the dynamics different? That's the first question, then I have a couple of follow-ups.

Well, on that question, I'd say that the international markets have a slightly less acute price pressure simply because different providers focus on different geographies and have different strengths, and so it's easier to distinguish yourself and value-sell in an international market.

In terms of fourth quarter, just thinking about visibility into fourth quarter, do you get the general sense that most of your customers have their plans in place for the fourth quarter, and they won't want to mess around with changing CDN services or making any sort of changes to their Internet delivery systems in the fourth quarter?

That's definitely true of the eCommerce vertical. It is not so true of software providers or -- unless they're an eCommerce software provider, but software providers, in general, for software updates in game software and media companies, but definitely the eCommerce guys lock down usually around October 1st.

Okay. A final question -- can you speak about trends that you see out there in terms of video delivery and eCommerce going into the fourth quarter? Are these trends -- generally what we've been hearing have been really positive in nature. Maybe it's consumer sentiment coming back, which benefits business sentiment. When you speak to your customers, what's the sense that you're getting from them? And maybe because your guidance sequential growth does not really exhibit that.

Well, our business is driven by consumers, consuming content over the Internet, not so much by any business sentiment. Of course, our contract negotiations are driven by business sentiment where customers are feeling bullish about their business, they'll step up and commit to more. And, again, I would tell you that our guidance reflects the facts as we have them and the visibility as we have it today regarding the fourth quarter.

Could you just remind us what were those three or four events that happened in fourth quarter of '08 that provides these tough comparisons?

I don't have those off the top of my head. I think we had a couple of large consulting projects delivered, and we might have had one or two lab events that happened in the quarter.

Operator, at this time, we thank everyone for attending, and the call is concluded. Appreciate you joining.

Thank you for your participation in today's conference. This concludes the presentation. You may now disconnect. Good day.