Netflix Inc (NFLX) 2010 Q3 法說會逐字稿

Good day everyone and welcome to Netflix's third quarter 2010 earnings Q&A session. Today's call is being recorded. At this time, for opening remarks and introductions, I'll turn the call over to Deborah Crawford, Vice President of Investor Relations. Please go ahead.

Thank you and good afternoon. Welcome to the Netflix third-quarter 2010 earnings Q&A session. We released earnings for the third quarter at approximately 1:05 p.m. Pacific time today. These earnings press release, management's commentary on the quarter's results, and the webcast of this Q&A session are all available at the Company's Investor Relations website at ir.netflix.com. As is our standard practice, this call will consist solely of Q&A and we are going to conduct the Q&A via e-mail. Please e-mail your questions to ir@netflix.com.

00 p.m. Pacific time today. With that, let's begin the Q&A session.

The first question is from Youssef Squali of Jefferies. How important is exclusivity of content to Netflix when doing digital deals?

Youssef, it's Reed here. It is not inherently important to us. But many of the other buyers are exclusive and uninterested in non-exclusive content. So we end up doing exclusive sometimes, basically to play the way the market is today. And that has some benefits for us also. But, it is not yet a core strategy. Over a sufficient number of years, it could become a core strategy, but we have not made that determination.

Next question from Youssef. What is the percentage of the subscriber-base against Blu-ray today versus one year ago?

Youssef, it's up from a year ago but it's pretty flat in recent quarters. So, still a nice little part of the business and it is fine for us.

The next question is from Mike Olson of Piper Jaffray What are your current thoughts on pricing? I realize you are in grabbing share modes but our opinion is that gross adds and churn would not be significantly impacted by a $1 per month price increase and you could argue that higher pricing would allow you to spend more on new content, thereby attracting new subs and reducing churn. How are you thinking about this?

It's Barry. You know from previous calls, in response to questions about pricing, we frequently remind our listeners that we are testing all the time. And some of you will observe that we have some price tests in the marketplace currently. Absolutely the pricing in Canada is working well for us. And maybe that will or will not play a role for us in the US. And, as we continue to read test results we will continue to update our thinking about prices.

The next question is from Mark Mahaney at Citi Investment Research. Would you expect the relatively higher mix of TV show consumption via streaming versus DVD to continue going forward or would you expect it to come down as you build out more movie selections? Is there something about streaming usage that tends to lend itself more to quick TV show consumption?

Mark, I would say we are trying to figure out what makes consumers want as opposed to trying to guide it. We are trying to get a broad selection of content both in movies and in TV shows. And currently the subscribers are choosing those in about equal proportions. And the way we measure is not by stream starts or session starts but by total minutes watched. So, movies get a little bit of an advantage in minutes watched because they're longer. But, we are not trying to drive it one way or the other; we're just trying to build out a whole lot of content that our subscribers want to watch.

Second question from Mark Mahaney. Are your earlier indications that recent free subscribers convert to paid subscribers at the same rate as they have for you historically?

Sorry. Is the question is the P1 conversion rate holding constant with historical levels? The answer to that question is yes.

Next question is from Steve Frankel at Daugherty and Company. You commented that the increase in free subscribers was a reflection of both faster US growth and the initial push into Canada. Would you expect free subs to remain at this higher level relative to the historical norm or will it normalize in the next quarter or two?

Well, it may normalize eventually. But, particularly in -- Q4 tends to be seasonally -- we see fast subscriber growth around the holidays, which tends to be back-end loaded. And Q1 tends to be front-end loaded So, for at least in Q4 that trend likely continues.

Next question from Steve Frankel, Brigantine Advisors. Can you clarify for me what your current operating margin goals are and do you anticipate that expansion beyond North America in the back half of 2011 would, at least in the early going, drop operating margins materially below that level?

I think you can infer from our guidance forecast for Q4, based on the midpoint of earnings, that operating margin come in to the year roughly 12.5%. So, slightly higher than the 12% that we have spoken about. And in Reed's commentary, you know that for 2011, we are guiding toward 12% on the North American business. And that if we continue to track in Canada, on the current trend line, to move towards additional investments internationally, and in his commentary we spoke about a $50 million investment to support that expansion.

One final question from Steve Frankel. Reed commented that the company spends $500 million per year in postage. This is down from the $600 million figure the company has mentioned in the past. Has the increase in streaming already scaled back the postal expense by that amount?

Steve, no. Postage is still rising. It's between $500 and $600. So, my comment in today's script was over $500 million.

Next question is from Nat Schindler at B of A Merrill Lynch. Several years ago when the two out at a time was the most popular, you said that the average user rented about five DVDs per month on average. Now, with the one out at a time option becoming dominate, where has this number gone? Also have you seen a change to the average DVDs rented by customers as you add additional streaming content?

As it relates to DVD usage, we haven't for quite some time reported on average usage across the base, except to say that it's declined and there have been two principle drivers. One of course is the mix change in the business with the increased popularity of one unlimited subscribers and secondly, has been a growth in streaming and substitution behavior. Having said all of that, regardless of whether it is three unlimited, two unlimited, one unlimited plan, if you look at usage year-over-year for just DVD, we would see a decrease on average for paying subs.

Next questions are from Ryan Hunter at Wedge Partners. First, can you talk about the penetration rate for the Wii?

Ryan, it has been very strong with the Wii. It is a very popular device. It's often connected to the living room TV. It is more family centric than the more hard-core gaming consoles. It has been a great platform for us.

Second question, what is the subscriber count for Canada?

Well, in Canada we are not releasing subscriber count. What I can tell you, our experience has been great in Canada. We have been hitting and exceeding all of our numbers and we are on track to be profitable in Canada late next year. So, really just a super performance living up to our high expectations.

Next question is from Jason Cheu at ABR Investment Strategy. Would you be willing to offer some color on the video game console opportunity? With an installed base of 75 million consoles -- PS3, Xbox 360, Wii -- what is the net subscriber opportunity for Netflix? How often can you market to that console base and what has been the net subscriber yield?

Jason, most subscribers will watch content on a multiplicity of devices. They will watch on their PC. They'll also watch on an iPhone, they'll also watch on a game console. So, don't think of it is as a sub that is controlled or influenced by a single device, especially with multiple TVs in a household. There will be a Blu-ray player in one room that runs Netflix and a video game console in another.

That being said, to answer your question on marketing opportunities, with both PS3 and Xbox, were advertised on the screen in various ways. So, it is easy for subscribers to click and get a direct lead to Netflix or get directed to sign up right there on the screen.

He had one additional question. Are you experiencing acceleration in net sub adds from consoles now that the no disc is required?

That only -- the no disc requirement was only listed three or four days ago. So, we are just in the process of rolling that out. We'll be able to comment more on that next quarter.

The next couple questions are from George Askew at Stifel Nicolaus. The streaming-only test in the US offers a streaming-only plan for $7.99 per month and a streaming plus DVD plan at $9.99 per month. Other than price, is there a difference between the $9.99 plan you are testing and the existing $8.99 streaming plus DVD plan? Also, if you want the streaming-only service, does that mean that $8.99 price point will go away?

George, that is one test out of many that you are referring to, which is the $7.99 and $9.99 variants. And to answer your precise question, the $9.99 in that test sell is the same as the $8.99 today.

Second question from George Askew. An important driver of growth has been the consumer electronics partnership. At this point, with the exception of Android, Netflix seems to be on most major connected platforms. What platforms or devices are yet to be penetrated by Netflix?

George, you're right that we've had tremendous progress in getting integrated. The big push over the next two years, will be as more and more TVs are Internet TVs -- they have WiFi built in. And, we will march along with that.

Today, there's still a fair amount of TVs sold that don't have WiFi. So, that's the big push in terms of breadth. And then in terms of quality, we are upgrading the user interfaces in all of these devices, really at web speed in the search of what the best user interface is for TV and that will vary somewhat by screening an input device. So, we have a big investment program across all of those platforms.

The next question is from Ben Rose at Battle Road Research. How does Netflix think about the cable service providers? Are they long-term collaborators or adversaries?

Well, there are two parts to a cable service, or at least two parts -- three parts, voice, data, and video. To the video side, we are mostly competitors, except that we are at a very different price point with a very small fraction of their content. That is the analogy I use is that we're a motorcycle and they're a car. But, on balance we are competitive.

To data, we are a huge ally,drives a lot of data adoption, larger plans. So, very aligned. And voice, no interaction with them.

The next question is from Jason Helfstein at Oppenheimer. Among the content available on Watch Instantly, are you seeing greater demand for movies, broadcast TV shows or paid TV shows like Showtime? What is the average number of minutes per streaming session and do people tend to come back and finish a movie and show if interrupted?

Jason, with the content availability, we don't see greater demand for one or the other. We don't divide it up by broadcast or pay or cable network. We're just looking for great shows that we can license for our subscribers. We don't disclose the average number of minutes per streaming session.

And yes, lots of people come back to finish movies and TV shows. They will be watching on the television and then they're in the car, and they finish on the iPod and Touch, or the iPhone and there are all those different -- they move to a different room and they get to restart and of course our software is smart enough to restart them in the right place. So, that all works very well as you would expect an Internet-oriented service to work.

The next question from Daniel Ernst at Hudson Square Research. Do you have a sense of what percentage of subscribers are viewing instant watch on their televisions versus a PC?

Daniel, for us they are all just screens. There are four-inch screens that are on mobile phones, there's nine-inch screens on iPad, 12-inch screens on laptops, and then a wide variety of television-size screens from 20 to 100 inches. And, that is not a key determinant it for us and not something that I can give any more color on.

Second question from Dan Ernst. Do you have a sense for what percentage of subscribers in the quarter or perhaps year-to-date that are returning subscribers, ie previous subscribers who have come back to Netflix?

Actually, the percentage historically has been running about a third, slightly higher, and with the acceleration in growth, that percentage has fallen a little bit.

The next question is from Brian Fitzgerald at UBS. How do you think about streaming content costs per user in the future in light of the recent and upcoming content deals? Will it be pretty balanced in terms of streaming content costs being offset by analogous postal increases? Or will it be lumpy as certain deals come in or maybe build to an inflection point?

Let's see. There's likely to be some lumps. It's not going to be completely smooth. EPIX is a great example, and we're stepping forward on that one. And that's pressuring the margins in Q4. And as we grow into that, over the next year that will smooth out.

But, in terms of the long-term margin structure, we would like to try to keep it between 30% and 35% gross margin. You're replacing the postal cost with content cost.

Another question from Brian Fitzgerald at UBS.There are now 200 plus Netflix enabled products that will result in 60 million plus devices sold in 2010. Could you give a rough break of the percentage that may be exclusive and maybe rough estimates of what these numbers look like 2011? Any rough thoughts on TV plus set-top boxes versus tablets versus PCs versus mobile phones?

I don't think ... there is no material set of the devices that are exclusive. Basically, with software, it tends to be applications for architectures. And, so none of the devices -- at least there are no material ones that are exclusive.

In terms of the number of devices, that is probably not the relevant thing. That is the next 100 devices are the 100 smaller devices than the initial 200. So, really it's coverage -- again, the big change over the next two years is WiFi getting built in to more TVs. And as that happens, whether it's Google TV software built into it, or it's got our native client both work very well. That is the big change in the industry.

The next few questions are from Andy Hargreaves at Pacific Crest. Do any of the content deals give rights for global distribution or do you have to renegotiate for each geography?

Andy, the general pattern is per geography because it mirrors the other distribution channels.

Next question from Andy Hargreaves at Pacific Crest. Are the majority of the streaming deals you are signing still fixed?

Andy, we like to think of them as fixed because they generally are in the shor term. But, then when it comes to renegotiation time, it ends up in some long-term basis, you know, sharing our upside with the content owners as we grow.

And the final question from Andy Hargreaves at Pacific Crest. Can you explain in more detail the music rights impact where, why were you paying $3.5 million for music rights before?

I am not sure if the question is... requires an explanation of what music rights are. They're the public performance rights for the notes and the words associated with the music that's synched in the TV and movie content that we distribute --- stream over the Internet.

There are public performance rights acquired from one of three right societies, ASCAP, SESAC, BMI. Two of those, ASCAP and BMI involved rate courts for the setting of rates, rates are in dispute. And so, as a consequence, we -- through time had made estimates that relate to our expectations for licenses associated with the public performance of the music, and some recent decisions in the courts which have caused us to rethink the money we set aside for the public performance rates. So, it is a monthly accrual. And then, eventually when licenses are struck, there will be an exchange of monies and we will move on. But, right now it is a hanging accrual in the balance.

Coming back to an earlier question I referred to, target 30% to 35% gross margin. That is, as new content deals replace the postage and that is true in North America. If we do global aggressively, in the short-term as we launch, that will be different. But that statement, the goal statement is true in the long-term globally, and on an ongoing basis in North America.

The next question is from Wayne Chang at Canaccord Genuity. Could you talk about the level of adoption or activation level you are seeing with a variety of Web-enabled devices in the marketplace, such as Web-enabled TVs, Apple products such as the iPad, or other tablet-like devices, and even gaming consoles? Are mobile handsets becoming a key driver to driving new subscriber activation or does the opportunity remain still relatively early?

Wayne, for music services, mobile is the core. And then, the more living room-based formats are auxiliary. And with video formats, it is the inverse. The living room formats and the laptop format and iPad formats are quite large and the mobiles, good supplements but it's relatively modest.

The next question is from Jim Friedland at Cowen regarding Canada streaming. In the comments you indicated that the Canadian business is on track to break even in the second half of next year. Does that imply that your Canadian content deals scale with subscriber growth or are you paying big upfront fees for streaming regardless of the number of subscribers?

Jim, that would be the latter.

And the second question from Jim Friedland at Cowen with regards to US streaming. Do you think the long-term operating margins for the digital-only business are similar to the margins in the business today around 12% plus?

You know, the long-term margins are really driven by competition. Because that really influences the pricing. So, you would have to tell me what the competitive environment is to guess at what the long-term operating margins are. It will be again, significantly influenced by the amount of competition and at this point we don't know what that will be.

The next question is from John Blackledge at Credit Suisse. What was change in disc shipments per sub per month in Q3 2010 on a year-over year basis? I think in Q2 2010 disc shipments were down about 20% year-over-year. Just wondering if you would expect the decline to accelerate in the near term or stay around Q2, Q3 levels?

John, as I said in answer to an earlier question we are not breaking out the disc usage per sub per month. The only color commentary we gave was in Reed remarks and he pointed out that total disc shipments had grown on a year-over-year basis by about 10%.

The next question is from Barton Crockett at Lazard Capital Market. The impact on other forms of entertainment consumption. Our young children love Netflix, watching instantly on the iPad. For that we are not inclined to buy more kids DVDs. They get more than enough content from the Pixar and Disney movies and PBS kids shows on Netflix. Are there other Netflix subscribers like us? Do you see any evidence that people are buying fewer DVDs as they stream more Netflix content? Could you also update us on the cord-cutting question? Do you still profess to see no evidence that Netflix subscribers are cutting the cord?

On the latter, Barton, on cord cutting we still see no evidence that our subs cut cords at greater rates than the general population, and it's something we survey now and then on. On your personal experience, you know, it is all over the map. We have tons of anecdotes of people who -- the streaming stimulates the uses of other formats for, like you said streaming meets most of their entertainment needs. So, quite a lot of variety with 17 million members.

The next two questions are from Michael Pachter at Wedbush. The press release says that in Q4, a majority of Netflix subscribers will watch more content streamed from Netflix than delivered on DVD.

There is quote-unquote more content mean more hour of content or greater numbers of streaming sessions than DVD transactions?

That is more hours. We are measuring everything through combined hours.

Second question from Michael Pachter. Also, the following sentence that the business would have transition from mostly DVDs to mostly streaming. If more than 50% of users consume more than 50% of their content via streaming I can see how you could say mostly streaming for those customers. But I presume that the 34% of customers that don't use streaming get 100% of their content via DVD. So, I don't see how you conclude that more 50% percent of content is streaming. Will you please clarify what you are talking about and give us the split in terms of hours spent. It doesn't make sense and the second sentence implies that over half of all hours spent, are spent watching streaming, but the metrics you provide don't support that.

It's just a math question here, Michael.

We track how many DVDs we ship and we track their run length, and make an assumption that each DVD is watched once. You could say some are never watched and some are watched twice but we assume watched once. Then, we look at the streaming hours that we are delivering and compare those. So, we are seeing two things, which are independent. One is that the total number of minutes of entertainment delivered by Netflix is higher on streaming than on DVD. Second, a majority of our members, so more than 50% in Q4, will watch a majority of their content by minutes on streaming than on DVD, which basically gives you a distribution measure. It says that it is not that, say, 10% of the streamers are watching an enormous amount of content and everyone else is majority DVD.

So, it's two views, total view, we're more streaming, and from a subscriber view a majority of our subscribers consume a majority of their content on streaming and DVD.

The next question is from Doug Nicholson at Deutsche Bank Securities. How much of your churn comes from subscribers that are on free promotions? Asked another way, how low is your churn rate on core Netflix subscribers, say those who have been subscribers for over three months or six months or one year, however you measure it? The obvious reason for the question, is that you do not typically see consumer subscriber businesses grow so fast while sustaining such a high churn rate. It should settle down to a lower level to the extent you are offering a good price value.

This is Barry, let me tackle that question. We don't normally see subscriber growth business grow this fast. There are two cohorts essentially to focus on with respect to churn, there are the free trial subscribers you referred to, and we see slightly lower -- we anticipate slower conversion rates in Canada than we'd see in the US. Let me just focus on the US where we have a rich history and have said repeatedly through time and it remains true today that about nine in 10 and subscribers who take a free trial become a paying subscriber. So, a great conversion rate, but that implies one in 10 free trial subscriber dropout and that is a 10% churn rate. So, point one, as it relates to the paying subscriber, churn declines over time and if you have been here for 12 months or longer, and more than half of subscribers have been here for more than 12 months, then your churn rate is in the 2% range or lower.

So, not terribly dissimilar from churn rate for basic cable, which is in, what, the 1.5% range. And of course whether or not that churn rate is sustainable from a financial perspective also depends on what the acquisition cost is. Of course, we have historical low acquisition costs because of investments we have made in quality streaming.

I suppose the follow-up question might be, how does retention rates for streaming customers compare with retention rate on DVD-only customers? And because of the improvements we have made in the quality of the streaming services we can say that the retention for streaming customers has just crossed over the retention for, I'm using the word retention, has just crossed over the retention for DVD-only customers and so it's actually better now. That is important, of course because the majority of the growth is streaming related.

The next question is from Edward Williams at BMO. As subscriber growth has grown due to streaming, how have the demographics of the Netflix subscriber changed?

Not significantly. We are still pretty much the broad Internet demographic that has credit cards.

The next couple of questions are from Scott Devitt at Morgan Stanley. Given the number of deals you made on the content side in Q3, are you still seeing opportunities to continue to aggressively invest in digital content?

Scott, we are pretty spent up for Q4. But we are three weeks into the quarter already, so that probably makes sense. We have a lot of opportunity next year as the subscriber base continues to grow.

Second question from Scott Devitt. With the rollout of the new disc-free applications on Blu-ray and PS3 you seem to be focusing a lot on reducing friction, improving user the interface and increasing search functionality. Are there any other areas you hope to focus on in order to differentiate your streaming products from other video applications from a technology standpoint?

Not particularly, Scott, I would say any technology advantage, quickly gets copied. So, we try to do great work and we're improving in the UI, but obviously all the great work we do there would be quickly copied within a year or two by anybody who wants to.

The next question are from Eric Wold at MCF. In the prepared remarks, you state that Q4 will experience year-over-year reductions in marketing spending. Are you referring to the overall dollar amount or SAC?

Both.

Second question from Eric Wold. Should DVD shipments continue to flow or even decline in certain markets like the Bay Area, at what point would you consider closing distribution centers or would they always need to stay in place for the next-day delivery benefit?

Mostly, the latter, Eric. Over a sufficient number of years, if the volume comes down by 50%, there might be some small closures. That is probably many, many years out.

The next question is from Tony Wible at Janney Montgomery Scott. Do you believe MSOs will move to usage-based billing as part of the FCC debate over Title Two? Do you see an opportunity to use compression techniques to minimize bandwidth if that is the case.

Yes, you know, that is a key question, Tony and a good one. We have some vulnerability depending on cap usage and what happens. Comcast has a cap but it is 250 gigabytes so most users feel they have unlimited experience and it gives us plenty of room to deliver high def streams. On the other hand, AT&T mobile data an iPad is now capped at two gigabytes, so, not enough room to deliver hours and hours of high def. There is not a lot of improvement in compression techniques but what we can do is to just deliver and lower bit stream and lower quality video experience. So, for example not too high def. So, that's one possible way to partially mitigate the impact. We are definitely sensitive as you point out to, in the long-term, at most of the industry end up 250 gigabytes or two at the other extreme.

The next question is from Arthur Mangriotis at Fox Point Capital. Please elaborate on your comments that the international competitive window will only be open temporarily.

Arthur, broadband is growing rapidly across the world. That creates an opportunity for us and potentially for others. And five or 10 years from now, you would have to believe that if we are not there, other people will be, so that is what we are referring to.

The next question is from Doug Anmuth at Barclays. Is SAC down because you deliberately push marketing dollars towards content cost or as a result of the strong subscriber growth. i.edid you go into the quarter intending to spend a bit less on acquisition?

We spent to our plan, marketing spending as a percent of revenue, was up slightly on a Q-over-Q basis and a year-over-year basis. The decrease in SAC reflects the increase in the organic growth which is a function of the increase invests we have been making in the licensing of streaming content. Historically, you have heard me talk about the inter-relationship between gross margin, SAC, and churn and the way in which we play one off against the other in order to increase the overall growth rate and enterprise value associated with growth in subscribers, and that's what you see here.

Ands Barry mentioned in his written comments as part of absorbing the EPIX lump, we will be spending lighter on marketing in Q4 than historically we would. And, we will have lower gross margins and lower SAC than you are used to seeing from us.

Another question from Doug Anmuth at Barclays. What are you doing differently in terms of distribution centers as you prepare for overall disc shipments to decline? I know you are focused on continued automation and efficiency. But do you have broader plans on how you ultimately scale back these facilities over time?

Nationwide, our DVD shipments are still growing about 10% year over year. Right now we are trying to figure out how to handle the increased load and get our new automation deployed quickly. So, I'm sure someday we will have to think about the issues you referred to, but not in the near term.

Another question from Michael Olson of Piper Jaffray. Question for Barry. Does your comment in the prepared remarks regarding Q4 guidance, where you say year-over-year reductions in marketing spending will mostly fund the increased content spending, suggest that Q4 marketing spend a dollar basis will be down year-over-year in Q4 or are you saying it will be down year-over-year as a percentage of revenue?

It will be down as a percentage of revenue. And, it will be down Q-over Q.

The next question is from Doug Robinson, a private investor. When will the Android cell phone be supported for streaming apps?

Doug, we are working on the Android now. As soon as we have something to announce we will announce that.

I would like to supplement my previous answer as it relates to marketing spending to make a slightly different point. I want to remind everyone on the call that earnings are a managed outcome. So, there have been many quarters in the past by way of example the fourth quarter of last year where, if with memory serves me, during the quarter because the business was much more profitable than we forecast it was going to be, we allocated during the quarter end additional spending on the marketing. Something like $6 million.

We delivered the earnings numbers that we forecast we were going to deliver and we will also able to deliver more growth in part because we were able to invest in more marketing dollars. So, if as Q4 unfolds, similar dynamics occur relative to our forecast, and we have the wind at our back, we would be anxious to invest additional money in marketing because it would further accelerate our growth. But based on our plan, we're expecting the answer I gave, which is a decrease both on a percentage basis and a decrease Q-over-Q in terms of absolute dollars spent on marketing.

Another question from Barton Crockett at Lazard with regards to the Starz deal. Starz receives net affiliate fees of close to $2 per subscriber per month for providing its flagship Starz-Encore package to cable. In my observation, the lineup of Starz content on Netflix is better than on cable due to more title availability, easier navigation, etc. In you renewal talks with Starz, is it conceivable that Netflix could pay rates per sub for Starz content comparable to what cable pays?

Can you comment generally on your optimism or skepticism about a successful renewal of the Starz deal?

Well, Barton, obviously I cannot comment on those specifics. I am optimistic in the long-term. That is, we have money to pay and they are in the business of collecting money. So, there is no reason in the long-term, strategically, it shouldn't work out.

How soon and whether there are hiccups along the way, like you see in so many content-supplier relationships, that's TBD. With the breadth of our content offered, you know, we will be fine. We look forward to working on that deal.

The final question is from Michael Pachter at Wedbush Since postage hasn't declined precipitously, you are really saying that overall consumption of entertainment is up a lot, DVD usage down a little, and streaming up a lot. That is a high class problem and suggest a higher value proposition. Could you confirm that I am reading the response correctly?

Michael, the only thing I heard in the question was the DVDs down, but DVDs are actually up. It emphasized your theme that DVDs are growing slowly for us, DVD shipments, and streaming is growing rapidly. I am not sure, I heard the question.

That was the question.

You are really saying that overall consumption is up a lot. Let's see, usage per sub down a little and streaming up a lot.

DVD usage.

DVD usage on a per sub basis down.

But a 50% sub growth with only 10% shipment growth.

The question is on a per sub basis are you seeing greater consumption? Are you seeing greater consumption of entertainment? And I think the answer to that is yes. Said differently, subscribers are extracting more value from the service than they have historically, which I think explains in part one, the increased organic growth and the decline in subscriber acquisition costs and, two, the continued decrease in churn.

Great. That is the last question I have. Before we conclude the call I would like to turn the call back over to Reed for a few brief closing remarks.

Thanks for joining us today; hopefully we answered all your questions efficiently. There is a lot to talk about this quarter, the year-over-year subscriber growth at 52% and accelerating, more content, more device partnerships, our entry into Canada. By any measure, our evolution to a streaming company has just been phenomenal. I think you'll see that evolution even more clearly when we report our fourth quarter results and look forward to talking with you then. Thank you.

Thanks, everyone.

Ladies and gentlemen thank you for your participation in today's conference. This does conclude the program and you may now disconnect.