Thanks for taking the time to sit down with me today. I wanted to start off by asking you about Rhapsody To Go, your subscription service for portable devices. Other companies like Yahoo and Napster offer similar PlaysForSure-based subscription services, but with everyone using essentially the same DRM technology, how does Real stand out? Is Real doing enough to differentiate itself from Yahoo, which offers a similar service at a cut-rate price?
Well, we have differentiation on a couple of different levels. The first I will mention is the editorial. People say, "Hey, it's great that I can actually get not just unlimited access to a million and a half songs, but I can also get it in a relevant way where I go to a band like Aerosmith, and it tells me that the Rolling Stones were one of their influences, and it tells me who their contemporaries were. It shows me that link to Run-DMC/Aerosmith's 'Walk this Way,'" One of the things that people love about Rhapsody is this sense of connectedness; in fact, our full subscribers listen to over 200 songs a month per subscriber. We've never heard our competitors quote the numbers of songs their subscribers listen to because it'd be much less.
Do you mean they're downloading 200 songs per month?
That they play over 200 songs a month, whether it's streamed or downloaded. It doesn't matter because you get access to it either way. Initially, it was just streaming. Now, we support both modes, and, in fact, the usage has stayed strong. It's crept up a little bit, which is a bit of a surprise because the user base is growing. You'd think that because it might be an early adopter thing that the number per user would come down.
So that tells us that people like the experience and use it much more, on average, than any of our competitors. With regards to the technology piece, we actually do have a very differentiated technology in that we have a set of DRM technology, which we call Harmony, that allows for playback on a broader range of portable devices than anybody else.
That that includes supporting the Microsoft devices, both on a individual track download basis and a subscription basis. We do that is because there are a set of devices that have that support, but we also support other devices like Palm Treos that have the native support for our Helix digital rights manager, and we also continue to support the iPod for people that want to take tracks and move them to the iPod. We're the only people, other than Apple, that have that support, so we have quite a differentiation from the technology standpoint.
The last thing I'll say with the technology is to look at the work we've done with Sonos, which is to bring an integrative subscription service directly to a smart in home device. Your readers would probably know more than most about just how great a product it is. We have one in our home and everybody who has it says it's just changed the way they think about getting digital music through their home. With Rhapsody directly integrated into the service you can get access to any of your playlists, you can get access to any of the music that you've specified directly on the device, and we're talking to Sonos about doing an even deeper integration.
Really, we can do that because we have the end-to end technology, and aren't gonna be just a passive packager of other peoples' digital rights management technology. We integrate with other DRM's because we want to get consumers the richest experience, but there is quite a differentiation in the technology, and I think that will be increasingly visible to consumers.
Well, there's always a lot of engineering work to maintain compatibility with anything you do, and so we have a
standard level of quality assurance in the case of specifically making it compatible with the iPod because they haven't
announced a developer program. They've said some things publicly that suggested that they were anti-consumer and that
they actually wanted to break compatibility.
We've had to do some additional work there. Generally speaking, when the new model comes out of the iPod, and there's a new digital rights management implementation there, it will take us a couple of months to support it in our next release, so in that way we do maintain compatibility with the vast majority of the iPod product line when a new product comes out.
We've moved away from talking about Harmony, what happened was that it got so much attention when we first announced it — which was great in a sense that it gave us a lot of visibility — but it ended up having the dialogue become sort of a he said/she said, and we'd much rather focus on all of the great stuff that our products make possible, rather than just that one thing.
The other thing that I think is important to mention in all of this, because you brought it up, is the role of price in this whole creation.
The value of $10.00 a month on the PC or $15.00 a month on a portable device, for an active user, is a great value. In other words, if you are an active listener of music — and there are subscribers who listen to over 200 songs a month and over 100 different songs a month — if you were to buy those individually, you'd be paying ten times as much as you'd pay for a Rhapsody subscription.
And in a portable scenario, let's say people put 150 songs on a portable device each month, you'd be paying seven times as much. I think the value of the offer is great, and for people that aren't active users we're the only people with a free product.
If you look at what kind of Yahoo did, Yahoo came in and thought they were getting a lot of attention with the short term price promotion, and then they backed off and raised the price back up because I think they were finding that the small differential on economics versus really delivering the best service, well that was a criteria consumers used. We serve our customers and try to understand what drives them, and what they want is the best experience. Obviously, if we were charging $100.00 a month or something, then price would be more a consideration.
Why has the subscription model had so much trouble getting traction?
The bigger issue, in my mind, and which I think is tied to why portable subscriptions haven't taken off, is that for most users the trade off is not between listening to a portable device scenario, paying $15.00 a month for unlimited access, or buying 25 tracks a month or 50 tracks a month. Most users that are filling their iPods are still doing it through piracy.
If you look at the four methods of getting songs, there's purchasing tracks digitally, and the average number of songs purchased per device has been staying pretty steady at about 25 per device, even though those people will have way more than 25 songs per device. Which means that they're getting them either from taking the CD's that they own and ripping them and moving them, which is a perfectly legal thing to do; or borrowing their friends' CD's and ripping them and moving them to the device, which is not a legal thing to do, and I think it's certainly not good for the music industry when people are doing that en masse; and then the last thing is that people are still using P2P services.
If you look at all the data from people like Big Champagne and Track List, it seems that the aggregate amount of use of illegal file sharing services has stayed pretty constant, even as specific ones get shut down. It's a little like that game of Whack-a-Mole, where you whack it one place and it pops up somewhere else, so the dialogue that we're having with all the people in the electronics industry and the music industry is that we think in order to change that dynamic you've gotta do something pretty dramatic.
Like what kinds of things?
We're doing an experiment right now where anybody can go in to Circuit City and if you buy a couple of different products from SanDisk, the Sansa product line of flash players, there's a coupon on there that says if you sign up for six months of Rhapsody To Go you get an $80.00 rebate.
So, in effect, depending on how you look at it, you either get a free device or you get a free subscription service. It's early yet, but based on what I'm hearing from the SanDisk guys the uptake that they've seen has been great.
It's one retail chain and it's one partner, but it's looking very encouraging that when you actually change the dynamic and you say to people, "Hey, we'll give you a deal if you give this a shot," it seems like it's very encouraging.
Now, after the holidays end and we get all the data in, we'll step back and we'll translate what that means, but I firmly believe that if we're gonna change the current spiral we're in there's gotta be a better way. The music industry is down in the U.S., it's down about 10 percent this year in physical, and digital is up about 2 percent in terms of the total basis. It's double from before, but it's nothing like 3 percent or 5 percent.
The music industry, as a whole, when you balance everything out, is down 7 or 8 percent, so that's clearly not a good thing, and not just for the music industry and for the artists, but also for fans, because at the end of the day the economic engine that gets new artists discovered and gets the money to make records and start touring, if that engine is getting disrupted, that's gonna have a negative impact on new-artist development.
We think that most consumers aren't dishonest, they just follow the path of least resistance, and what happened was, up until very recently, stealing was easier than getting legitimate stuff.
We're just now crossing over to a period where legitimate services are better, more reliable, more accessible, and easier than pirate services, and we now have to go in and encourage consumers to relearn behavior rather than following this default path that's kind of gotten burnished into the culture over the last eight or ten years, and that won't happen overnight.
What sort of reception have you gotten from the record labels for the To Go service?
They are generally supportive of the idea and they recognize that in the current market environment it's off to a slow start. Think about it from the standpoint of the consumer that has two devices side by side in the store. One supports subscription, one doesn't. Maybe it's an iPod and an iRiver or a SanDisk.
The ones that support subscriptions cost the same as the ones that don't support subscriptions. The most powerful brand is the iPod, and the one that has the coolness and hipness factor is the iPod; no question about it, and you give credit where credit is due.
Apple has done a very good job of building very good portable device hardware products, so the momentum right now is not on the side of portable subscriptions. Even when people are buying devices that are capable of doing portable subscriptions the attach rate per subscription is only a couple percent. There are a few reasons for this.
One is that, generally speaking, where Apple's market share is lowest is at the low end of the market, so the consumers that are buying something other than Apple products generally are buying the more price-sensitive or value-sensitive products - that's because Apple's sub hundred-dollar products are actually pretty crappy.
You can hype it all you want, but a portable music player without a screen is inferior to a portable music player with a screen. It's not rocket science. People want to see what song is playing. When you push the next button, you want to know what's next. It's clever marketing by Jobs to position serendipity as a virtue, but for most consumers it's obfuscation.
In that sub $200.00 category competitors oftentimes have better products than Apple, but those are the customers that are probably least interested in paying for subscription, so I think the worm will turn on this one, and I think 2006 may well be the year that it turns.
Oh, I believe portable subscriptions will become a lot more popular over the next two years than they are today, but I don't think it will happen automatically. I think it will happen as a function of better products coming to market and smarter marketing to introduce the concept of subscriptions on devices to consumers.
So what happens if Apple introduces a subscription service?
I think it would be the best thing in the world because I think it would change the debate. Today, you've got the guy who is No. 1 in the device business saying subscriptions are bad, and then you've got you Snow White and the Seven Dwarves.
Outside of Apple, nobody has established themselves in the portable music segment as having that dream product that your readers have to have. When you survey your readers, what's the second most popular portable device?
It's actually pretty muddy, there's a bunch of stuff out there, but nothing that stands out as a clear competitor.
Right. It's kind of none of the above.
Yeah, there's iRiver, there's Samsung, there's Cowon, there's Creative —
You know, SanDisk has a nice new product -
Everyone is squabbling over their two percent of the market. Right, so therefore, there's no thought leader device that is getting people to think, "Man, if I want to try a subscription, that's the thing that would be like the really cool embodiment of it."
Why has it been so difficult for a competitor or an arch-rival to emerge to the iPod? How much are you guys engaging with the Samsungs and the Creatives and the iRivers of the world and talking to them about their products?
I think the answer there is that we've reached a conclusion that we have to get much more active in partnering with device guys in general.
It seems like that's what's really holding Real back as a business.
Yeah, the original strategy - it was probably about nine months since we've realized there were flaws in this — was to assume that the rest of the market, outside of Apple, would see Apple's closeness as an opportunity and would do good innovation.
I actually think that will happen, but I think what has happened in this interim period is that it's been much more the Seven Dwarves than anybody really stepping up to the plate. And - to give responsibility where responsibility is due — for a bunch of reasons the software that Microsoft created for portable subscriptions didn't result in a great out-of-the-box experience, so we've worked hard to do integrative implementations with companies that were doing portable subscriptions support on these various devices.
What we found was that there were a lot of issues where the out-of-box-experience simply wasn't as good as it should be, be it firmware — if the first thing you're told is that you have to upgrade your firmware, that's not a very good experience - or the transfer speeds which, frankly, aren't as good, or at least weren't as good, in the first generation of products as it was for the iPod.
That kind of seamlessness is our goal. We're very proud of work we've done with SanDisk and Sonos, we're getting great feedback from people for the work we've done and we'd love to hear feedback from your readers, the SanDisk is probably the best thing on the market in terms of portable device subscriptions implementations in 2005.
We will work very hard in the next year to get stuff that's even better, and I think by the middle of next year there will be integrative end-to-end implementations of Rhapsody and portable devices that are world-class and second-to-none in terms of the quality and the experience, and that will have that — I don't know what you call it at Engadget, that sort of device envy, that gotta have feeling.
I believe that in 2006 there will be really, really cool, gotta have portable music devices that have integrative subscription support that will get people excited. We are in the development stages of partnerships that will bear fruit not that far into 2006.
So you think that one of these companies out there will sort of emerge as a counterweight over the next year?
Well, whether it's one or a couple is of less importance to me than it's at least one.
But you think there'll be at least one?
How do musicphones fit into your strategy? When I sat down with Bill Gates earlier this year and I asked him about portable media he wasn't very concerned about finding someone to build an iPod killer that uses Microsoft DRM. He seemed pretty sure that cellphones would eventually dominate and be a huge part of the portable media business.
Well, I think phones have a very important role to play. The reason that I've been focusing on the dedicated portable devices discussion is because I think there will always be a segment of the market that's best served by a dedicated device. It's the difference between digital cameras and cameraphones. I don't know very many people that are serious about digital photography that don't have a digital camera, but lots and lots of people take pictures on their cellphones as well.
Similarly, I think there'll be a lot of music and music activity that happens on mobile phones and having services that span mobile phones and dedicated devices is a key part of our strategy. If you want to summarize our strategy with Rhapsody in a single phrase, we want to create the jukebox in the sky that people use wherever they are, whether they're on their own PC, whether they're traveling so they access us via the web, whether they use it on a mobile phone, or whether they use it on a dedicated portable device.
In a highly heterogeneous world where people have multiple devices, we want to offer the most popular services for those devices, and that's a very different strategy than Apple's strategy, where they only support one model of one phone and even then they create this crazy artificial limit where you only get 100 tracks on it.
So my take on this is that we'll see a very different world in the next couple years as these mobile phones get more active. The thing we've done there, which is kind of a down payment is something we're doing with Sprint and Rhapsody Radio which supports, I think at last count, seven or eight different headsets. That's an example of a service, and of course a service has to work on as many phones as possible. When you take a service-centric view as apposed to device-centric view you get very different end-results.
So do I think mobile phones are going to matter a great deal? Well, absolutely, and I think that interoperability is gonna matter even more and we've put the flag down. We think consumers are going to want freedom of choice, but rather than try to win the debate with the abstract, we're going to go out there and deliver freedom of choice on the widest range of devices possible.
We think that once there's a second cool device and once people start doing stuff on mobile phones, we'll have millions of consumers who will have an "a-ha" moment where they'll say, "Wow, I didn't realize that I had joined a cult. I thought this was a cool product, but now that I want to use one of those other ten products, I'm kind of landlocked, and I don't like that." We think our strategy takes a little longer to play out it, but it is absolutely the thing that delivers the best experience for consumers long-term.
You're trying to offer as much flexibility to customers as possible within the limits of DRM, but is the complexity added by all this flexibility and interoperability part of why it's been so difficult to convince consumers to switch? With Apple your options are more limited, but it's also simpler and less confusing.
There's no question that when you look at the history of the car industry, Ford's market share was much higher in the early days than General Motors because of Henry Ford's "You can have whatever color you want as long as it's black." It took General Motors a long time, probably ten years, for their market share to surpass Ford's. Nowadays, since both GM and Ford are struggling, people forget that analogy.
A closer analogy is Apple, which had 42 percent of the PC market in 1982. Because of Apple's insistence on end-to-end control, their market share diminished to where it's in the low single digits today. I have little doubt that if Apple continues to be as proprietary as they are, their market share will follow the same arc, even though I give them credit for, in this phase of the business, doing a better job of innovating in portable devices than anybody else.
I don't fault consumers for buying the best products at a given point in time, but I'm a big believer in Bill Joy's Law, which is that the vast majority of the smart people in the world don't work for your company, no matter what company you work for. If you believe that this is the case, then over time, in any business, the vast majority of innovation is not going to come from one company, no matter what that company is.
So if that's the case, then there will come a time. Maybe it will take a couple years. I understand that people are metabolically focused on what's the here and now, but I am absolutely convinced about this. More convinced than ever, in fact. And not even that long of a term, either, I think it's more like two years. I don't think it's ten years.
The benefits of openness, the benefits of interoperability, the benefits of customer choice will actually seem meaningful to consumers in a way that can drive very rapid shifts in market share. Since we're not a portable device company, we don't have it out for Apple in any abstract sense because Apple has already said: "Hey, they want end-to-end control of everything. It's not worth it for them to support compatibility with other people." Go talk to the people at HP on how their partnership with Apple worked out. Go talk to people at Motorola, if you can, because they still work with them, and talk to their director about how their partnership with Apple worked out.
I understand that they feel a little burned.
Well, it's very clear that two things are the case. One is that Apple is doing an excellent job of innovating today, and their strategy is one that inevitably will fail. When you try to have that much control by one company over everything and you don't let other people innovate with their devices and when you don't have interoperability then you lose once the rest of the industry gets its act together. I certainly would welcome an opportunity to create interoperability with those guys, but we recognize that the reality is they don't want to play, so great, so we'll work with the rest of the industry.
It might take a little longer, but we think we'll deliver a superior solution, not just in terms of the quality of the experience, but the benefit of the consumer, and the consumer gets to choose from the widest range of options.
DRM in the first place then no one would have a problem selling tracks that could be played on an iPod"?
I think that's a factually correct statement, but I'll give you another set of facts to swirl around. In the first seven years of the CD, from about 1983 through 1990, for every $1.00 that consumers spent on CD hardware, consumers spent $1.50 buying music, buying CD's. In the first seven years of digital music players, MP3 players, for every $1.00 consumers have spent on buying hardware, consumers have spent - and this is the most recent date which is the ramp-up; it's ramped all the way out to $0.20, so 1/7 the amount on a relative basis. If you're the music industry, and you're only getting 1/7 of the revenue per capita, the first thing that comes to mind isn't, "Hey, let's get rid of copy protection."
Now, you can argue that if they had and they had done a bunch of other stuff, maybe they would have gotten a better outcome, but I understand why they are where they are. I even understand why people like Sony and BMG can misfire with things like XCP because they definitely view themselves as needing to change consumer behavior.
Our view is that you have to do much more with carrots. If you look at the DVD case, DVD's come with DRM, but it hasn't stopped them from being what you'd have to say is the hit product of the last ten years of consumer electronics. That's because there was industry-wide adoption and support and even though you might say that CSS got hacked by DVD Jon and there are region-codes, the reality is that for the vast majority of usage cases you buy a DVD from whatever store you want to buy it and you plug it in to whatever brand of DVD player you have and it just works.
So where DRM fits into this is the interaction between the desire for DRM and the lack of cross-industry collaboration with regard to interoperability. As far as I'm concerned, if I were a music company I would demand that anybody I'm licensing my content to only put it into a DRM that's interoperable. I think when you talk to executives at record companies they wish they had done that with Apple, but they feel like given that they didn't, the cows are out of the barn.
There's a growing sense, particularly with our readers, that the record labels are so overly concerned with piracy and DRM that they're going down with the ship.
Today they are going down with the ship, and the question is, Are there techniques that you can use to make DRM more transparent? If I were running a record company, I would say, look, I'm limiting the licenses to people that use interoperable to DRM's, and I'm literally not going to license to people that don't, and if that means that there's a period of time where my catalog is not available on iTunes, so be it.
Even though digital is not that big of a business and it's not growing fast enough to compensate for lost CD sales, the problem is that Apple is a large part of what sales are there in digital, so they would be loath to cut that off. The back-up thing to do would be what kind of you're suggesting, which is, "Hey, while there's not interoperability, I'll give people waivers so they can sell naked MP3's, so they can flow anywhere." But, of course, the music industry is loath to establish precedent.
Now, you'll see individual artists selling straight MP3s, for instance, if you go to Pearl Jam's website and you buy CD's from their live concerts, what gets downloaded are straight MP3's, so there are individual artists whose attitude is, "Look, I'd rather reach the fan base in the most convenient way possible rather than go the DRM route."
Do I think that the right thing to do would be for the majors to recognize that the current environment where there's not DRM operability is part of the problem and to offer, for a couple of years, DRM-free music?
Yes, I do.
Do I think they're likely to do it? No, I think it's very unlikely, and since I think tilting windmills is different than what I'm supposed to do, I'd rather focus on getting them to do things that are good and reasonable things and that they're willing to do.
That's why we're focused on marketing ideas and approaches to get subscriptions, because from a standpoint of a consumer, if you're on a subscription path and your device supports subscriptions, you get access to everything, and it's kind of win/win. You can buy stuff if you want to own permanently, but you get access to the full kit and caboodle, so it's potentially a very compelling offer to consumers. The key is to package it in a way that allows them to open their arms up to it.
Well, that was my next question. How do you educate consumers about this? How do you explain to consumers that it is compelling? What do you need to do?
I think first off, you have to have a phenomenal end to end experience, and I think what we have today is a good end to end experience, and I think when we get to something we think is phenomenal, we will shout from the rooftops.
Is that partly about the failures of the manufacturers and the hardware?
It's partially about that, but it's really about the custody chain that you have. We did a bunch of stuff to the PC. You had a software layer that was written by Microsoft to plug in, and then you had device guys take the last piece.
So what we've done in the last six months, and what we're doing with Sonos and SanDisk, who were the first two examples of the fruits of that, is we've said, "Let's step back and just think about the end to end. How do we deliver the best possible experience?" The good news is that we're making progress.
I also wanted to ask you about the big announcements you made recently related to Rhapsody.
There were two related announcements. The first one was Rhapsody.com, which is about bringing the Rhapsody you know in as rich a form as we can to the web, the second announcement is an initiative to turn Rhapsody into a set of web services. We call that the Rhapsody Web Services Initiative, creatively enough, and in that case what we came out with, I would say, is a down payment on the full initiative which we call the 0.1 version. It's useful, but unlike the case with Rhapsody.com, which is a full-fledged version 1.0, this is really more of a start. I think they're important because they're the next step in taking the jukebox in the sky and making it available to everybody wherever they are, whatever they're doing.
The specifics of Rhapsody.com is that for the first time Rhapsody is available to Mac users, to Linux users, and it's available to people when they're on the go and not in a situation where they can download client software. We have a small plug in that gets downloaded automatically and it's a very low friction way for people to experience the service.
But it's definitely a more truncated version of the experience than you would get with the client, correct?
Yeah. What you get is access to the full jukebox in the sky; you get all the search; you get access to playlists that are entered. What you don't get are any of the things that you can do with music when you take it offline. You don't get CD burning. You don't get the portable device support. You don't get all the different kinds of ways to organize your offline library, but we think that what we have in the web service, on the web side, is a very good start. We make it very easy to use, you just click and play. Because of the Rhapsody 25 rights that we negotiated earlier in the year, Rhapsody is the first service that doesn't require you to be a paying subscriber to get access to the 1.5 million songs in our library.
Anybody who is not a paying subscriber can listen to up to 25 songs a month, that's the kind of service that makes a lot of sense to bring to the web because we can give people access without them having to go through the step of being paying subscribers, and a lot of people want to have their access directly on the web. It's the richest experience of its kind, but if you really want the richest form of Rhapsody experience, you download the Rhapsody client. We came out with a new version, client 3.1, that has some nice enhancements in it as well. In terms of the web services piece, I described it as a 0.1 release. The features that we've added are — and again, this a good start, but there's a lot more to do — is the ability for anybody to query the database and automatically get back the XML necessary to put that query result in a web page.
For instance, if you were running a Lynyrd Skynyrd web site and you want to have a list of all the Lynyrd Skynyrd songs that we have — like if you want to have a Lynyrd Skynyrd discography —then you can click the HTML and go to a direct play within the little, the mini web Rhapsody player.
Which would be actually embedded in the site itself?
Well, that's my other point — it depends on the form of it. The first form of it you get because of the way we do registration; you get brought over when you click onto the Rhapsody.com log in. One of the things we plan to add in the future is to take the log in and not require a log-in account and just allow a machine to have that Rhapsody 25 associated with it.
In that scenario we'll allow somebody to actually have something that's deeply embedded in their web site so the person that has the Lynyrd Skynyrd web site, to stick with that example, could actually - you could click, and you play "Free Bird" directly on that site, as long as that machine hadn't had 25 songs plays in a given month. That's the fuller form of web services where it's a purely integrated part of the web site that people have integrated it with.
The other thing that we have today is the ability to take a link; that we just use a link that would be a link to any playlist that you can create, and you can cut and paste that and put that directly in as well, but again, with the same playback behavior where, as a consumer, you would go to the Lynyrd Skynyrd web site, — and let's say somebody had put together a playlist of "Free Bird," "Give Me Back My Bullets," "Sweet Home Alabama," and that was their sort of best of Lynyrd Skynyrd — and click on that and get that playlist directly to the mini Rhapsody player.
The last thing that's part of the web service initiative is that we're actually creating a blog on the site to get feedback from developers and from web site people on what they actually want.
One of the things that's very clear in this web services world is that it's very organic. People like doing all kinds of mash-ups that you wouldn't necessarily have predicted, and therefore, they'll want not just general flexibility, but access to the interface and all kinds of things we couldn't have thought of, so I'm sure someone will have some notion of, "Hey, click on a Google map to find music that was created in a given studio." I don't know if we have embedded data for the studio where the album was made - I'm sure we have the name of the artist and the discography, but I don't know if we have geographic information, but somebody might want to have the ability to tag the database with geographic information, for instance, so we might have to create an extensive interface because. Right now the way the database is created is we do all the ingestion centrally, but you can imagine that people would want to tag this stuff in all kinds of different ways.
One of the things we might learn is that people think it's a cool service, but they want to have arbitrary extensible tagging for any artist. Then, of course, if you're not careful you get the Wikipedia problem, where somebody claims that a given album was created in Alabama when it was really created in Atlanta, so that while one of the things we could allow for is inbound extensibility of the metadata, the problem is how to do that in a way that results in maintaining accuracy.
That's one of the things people like about Rhapsody. Unlike what's happened with the pirate P2P services, when you click on a link for "Sweet Home Alabama," you get "Sweet Home Alabama." You don't get an Easter egg and you don't get a mislabeled version of it or a crappy recording, you get a super high-quality recording of it. So like a lot of these web services initiatives, we will, in a success scenario, have to confront how much to enable extensbility versus how much to focus on the accuracy of what people get, and part of why we're taking the approach of, "Hey, let's open up a dialogue with the development community" is because we'll learn from people the best ideas on how to build on that generality without bringing down the reliability and quality of what we're delivering.