by Jon O’Hara

In the last days of the 20th century, Michael Robertson’s MP3.com became a lightning rod for the digital music revolution, shocking the music industry with its brazen attempt to change the way people access their music. While its notorious myMP3.com service allowed users to play music they purchased on CD from any computer, MP3.com—which otherwise had been a successful platform for the presentation of all kinds of independent music—failed to gain rights-holders’ permission to build the music database on which the service relied, and was soon virtually sued out of existence. Universal Music Group parent Vivendi Universal eventually bought the company for (ahem) a song.

VU itself later ran into some well-documented money problems, and as it went about disposing of numerous assets, it sold much of MP3.com, including the once-notorious domain name, to technology and entertainment information aggregator CNET. Looking to expand on the tremendous success of its premier video game information site Gamespot, CNET set about re-thinking what MP3.com should be, literally rebuilding it from the ground up based on Gamespot’s business model, which combines an ad-driven website with business intelligence products gleaned from site usage. Gamespot currently attracts 8.1 million visitors per month.

CNET Games and Entertainment Sr. Vice President Vince Broady, who founded Gamespot in 1996 with partners Pete Deemer and Jon Epstein, has spearheaded MP3.com’s rebirth. Launching side-by-side with CNET’s recently opened Music.Download.com, which now plays host to indie music much in the way the original MP3.com did, the new MP3.com is now what the company calls “a comprehensive, next-generation digital music discovery and content resource that caters to the way people learn about and listen to music today.”

Here, Broady talks about how MP3.com will help consumers find legitimate sources for online music, where the site stands today and where he and his team see it going.

CNET’s MP3.com officially launched May 10 after a short delay.
We had a “rolling thunder” kind of approach where we went out and talked to the press and then followed up a week later with the actual launch of the site.

It’s pretty deep with information for people who want to know more about music or music technology—but that’s CNET’s M.O., right?
CNET really came at it first thinking from the technology angle—we’ve been covering MP3 players and these services for years. And we’ve seen, in some cases, music becoming not just a peripheral category, but a core category, like with MP3 players emerging as a number one or two category of consumer electronics products on the site.

But to make MP3.com work, we realized that it couldn’t just be about the technology, it really also had to be about the content itself. A big part of what we were trying to do was help with the discovery of music, and that’s why we felt the need to have a deep, rich content base, and, uniquely, offer 30-second samples of millions of songs without requiring any kind of software download or registration.

It seems people were surprised that MP3.com was no longer going to be a hosting site for music, but rather an information site that would point you to music. You chose to create a different site, music.download.com, for hosting.
It was a measured, carefully thought-out decision. When we purchased MP3.com, we basically acquired the domain name. We weren’t really able to purchase anything else, such as the library of music, or any elements of the existing business—they weren’t available. So we had to rethink it from scratch, and one of the things we looked at was how people were actually using the site. It was a source for people to upload and download independent music. If you looked at what people actually searched for, it was Britney Spears, OutKast spelled four different ways, Eminem spelled nine ways. That’s just what they were doing. Unfortunately, by and large, those searches went unfulfilled, because the site didn’t really have any commercial music on it. And so we felt that there was this built-in audience of people who know the term “MP3.com” It shows up very high on all kinds of search engines, including Google. We thought we should make a site that, when people get there, actually delivers something for those popular artists.

At the same time, we triedto find a way to serve the independent artists. We have served the independent or self-published community for a long time in the software business at download.com, and we’ve made big hits of software that guys have developed independently—things like Winzip or ICQ. We felt like, we have a good infrastructure there and a massive audience, so if we offered a music service similar to our software service, it could actually work. And in fact it has worked quite well after just a few weeks.

What has the reaction been?
I guess we can judge it by the pipeline of submissions we’re seeing, which is significant both in terms of artists and songs. People seem to be responding well. People that use Download.com know that it can make a big deal out of something that’s independent and great. Now, people want to see the features that are already on the rest of Download.com—for instance, they want to see charts of the most popular things and so forth, and those things will be rolled out shortly. People want it to be as good for music as it is for software, and we’re well on our way to making that happen.

A lot of artists have expressed a desire to have more of a presence at MP3.com for their music. They’d like to see more of a linkage between the two sites. And that linkage is definitely going to happen. We think this is a really interesting time in terms of music evolving from its historical roots to a new sort of digital entertainment. But there’s a lot of noise and confusion out there. This is really the first time that popular music has been legally available in a broad range of formats on a broad range of services. And we just felt that for the time being, it would be really good to focus MP3.com on this new, emerging market and keep it focused on popular music that people are familiar with and are looking for. And then, over time, once people really understand the way MP3.com works and are maybe looking for something more, at that point to create the linkage with Download.com.

The emergence of digital music players has made it possible for people to listen to, literally, 10 or 12 hours of music a day, every single day. As a result of that, we think people could run through an entire catalog, in terms of a category of music or several categories, over a period of months or years. Which will give them a hunger for new music that’s in a style that they like, but that they haven’t already listened to. That’s one of the reasons why we share a “genre tree” that’s about 700 deep with MP3.com and Music.Download.com. At some point, we can sync up the two sites so that once a person has run through all the recorded or signed artists in a particular category, we can offer them truly independent music that they’ve probably never heard, but is some of the best in this genre.

Are you actively working with record labels to expose new music?
Obviously, we’re talking about a compressed time frame. We had to make decisions about where we wanted to start, so we began working with the digital music service providers, who in and of themselves had forged relationships of one kind or another with the labels through their ability to resell their music. We did that in a very proactive fashion. We could have gone and scraped all the data from these guys and put it up in a directory—but instead we decided to have legitimate partnerships. Let’s find out what makes each of these different business models tick, and let’s find a way to make it work with MP3.com.

Now, if you look at where our history is in the games market over at Gamespot, and these two properties are really connected in a lot of ways, our primary revenue driver, our primary advertisers, are the game companies themselves. That’s because we have developed a product that is one of discovery and transactions for users, and it’s very efficient for game companies to use that product as a vehicle to get out their message—whether that’s on the PR side, or on the pure advertising side. We think, over time, if we can establish ourselves—and we recognize that CNET still has to prove itself in the music space—we think we can build a great product around music that will bring in users. And as we do that, we believe interest from the labels will definitely increase, as far as wanting to use our property as a way to expose their new music, whether from a promotional or advertising standpoint.

Isn’t there also a B2B applicaton for market data you’re able to collect?
When we build products at CNET, we try to focus on three constituencies: the user, the industry that we’re covering and the marketer. On the industry side, we basically take all the activity that occurs on our websites—in the case of games, all the clicks and requests for information every day—and then we put that into an interface which allows a game company to see what has happened in the market. Which games are building momentum and increasing in terms of unique searches, people checking prices, message-board posts, and which games are starting to drift downward. Which games are becoming more tracked and more popular, and which games seem to not have the status they once did.

In addition to that raw activity data, we also map over it demographic and geographic data, so that we can say, among this particular demographic, say males 18-20 in Southern California and Texas, which games have the greatest interest. When we set up MP3.com, we did everything in an identical fashion, so that right now, today, we’re capturing this information. We know what music’s being searched for, what bands are being sampled, what stores are being used to purchase the music. We know if people who check out these bands are also checking out these other bands—we’re collecting all kinds of data.

We’ll let that collection process run probably for about 90 days, to ensure the data’s consistency and validity, and also to let the property establish what its normalized user base is. And then we will come out with a product that will be very similar to what we have in games, which we call Gamespot Trax. We’ll come out with something called Music Trax, or MP3.com Music Trax that will provide for those interested a view into what people are looking at and what the trends are.

Have you reached out to music marketers to gauge their interest in this information, or has anyone approached you?
The main people who have asked us about Music Trax are people who already know and use Gamespot Trax, and those are typically ad agencies, and more consumer-focused companies. We haven’t really had much dealing, to this point, with the music industry. We haven’t really dug in with them, because we haven’t had much to say until now. But again, our approach is to take our time, do a good job, establish credibility and not come out and say, “Hey, everyone pay attention to us!” We believe that if we take our time and do things well, then when we have the opportunity to show the labels what we can do—whether it’s in 30 days, or in three months, or in six months—we feel very confident that they’ll like what we can do. But we’re not going to trumpet from the mountaintops that this is something they have to spend a lot of time on until we’re both ready.

But you’ve proved the model already with Gamespot.
We have a model for doing this that we’ve already established. I would say it took really eight years to get all the pieces in place on Gamespot. We’ve had such a good run of success over the last couple of years with Gamespot that we began to really consider whether there were other places where a similar model might apply. Now, we know there are going to be huge differences between the games market and the music market, and again, we’re not assuming that we have all the answers, but we felt that there were enough similarities that we should try it. And when the opportunity came up to buy MP3.com, we jumped on it, and tried to get a high-quality product out there as quickly as we could.

The MP3.com you see today is really only the first phase. There will be two additional phases that will come out between now and the end of the year. The second phase will be a community element, which is pretty non-existent right now. We have message boards, but only in the most bare-bones fashion. This community element will allow users to create a profile and store their musical tastes and interests, enable them to share playlists and just do a lot of things around their own set of interests in terms of genres or bands. Not only to contribute to the site, but also to communicate with each other. These are all general concepts, but we have done all this on the Gamespot site just recently, which has been extremely well received, so we look to bring a really strong, interesting, innovative community element to MP3.com.

The third phase will be a truly personalized service, a service specific to one individual. The artists that you like, the digital services that you use, the hardware you have—all those things being taken into account every time you look at a page on the site. For example, if we determine—either because you tell us, or because we infer it based on your activity—that you’re using Napster and Rhapsody, then when you click “check downloads and streams,” we can make sure that Napster and Rhapsody always appear at the top of your screen.

I hope you’re not going to call that “My MP3.com.”
Probably not. We’re not gluttons for punishment. But I will say that it’s been surprising just how much rancor there has been over the history of MP3.com and what is it going to be and what it should be. We’ve just done everything we can to try to make a service that will stand the test of time and that people will use.

What is MusicVine?
MusicVine is an attempt to say that there are multiple ways to look at the information we have, and that maybe a visual search interface would be fun in the context of music. The way it works, basically, is it takes a lot of data points into account—for instance, it takes into account the musical category, or the time the music was recorded, or other filtering that allows us to say, “Users who like this also tend to like that.” It takes the data points that might help you find similarities, and then displays them in this interface. I would say that this iteration is very much the first step. It’s not, in my opinion, as clear as to how these things are related, why these things are the same color or different colors, why this is big and why this is small.

Can you describe the Gamespot/MP3.com business model?
The CNET business model that has actually succeeded is a mix of advertising, business intelligence or services, and also leads—people going from our property to someone else and conducting a transaction. The vast majority of our revenue is from advertising—75% or so. We’re real confident in our ability to sell ads on MP3.com, because we have had six months of call reports and discussions with consumer, technology and advertisers, all of whom recognize this sea-change taking place in terms of people moving to digital music. Yet there have not been that many vehicles to reach users in a digital-music mode with their marketing message. If you think about it, iTunes, Napster and all those places, are really designed to sell music or expose the iPod or what have you—they’re not really designed as media vehicles, and we are. So we’re very confident about that and we know the response is there.

The second thing, something I’m actually kind of proud of, is our adoption of a very customized approach to developing lead models with each of the independent providers. Because nobody really knows what the lifetime value of a customer is. We all know that the margin on a song isn’t very high, so we couldn’t really go in and say, “Hey, give us all your margin.” So in some cases, we have a one-time bounty for signing up a new customer; in other cases we have a percentage of the transaction; and in other cases we have a percentage of the lifetime revenue that new user generates. We kind of just said, “What works for you, Mr. Digital Music Service Provider? We’ll make that work for us.” Because the most important thing right now is to let consumers know that there’s a choice, there are great services out there with the music that they want to listen to, and the most important thing for us is to make that available.

On the business-intelligence side, basically the way that we look at this is, again, we’re here for the long haul. In the case of Gamespot Trax, we actually allowed people we thought should see product to access that information basically for free for a year, so that they could understand it, ask questions, help us find problems, see it first-hand and decide whether or not they believed in it. We anticipate doing something similar in the music space.

Who will that information be targeted toward?
We hope that MP3.com will act as sort of a “halfway house” for people on the Internet engaged in file-sharing and stealing music. We’re hoping that by presenting an easy-to-use, functional alternative, we can help people realize, “You know what? I might be able to find this song somewhere on one of these other networks, but the alternative of being able to just go to MP3.com, quickly find it, cue it up in iTunes and buy it works for me.” Right now, there aren’t too many options that are easy to use and don’t require users to set up an account and download software. By making MP3.com a web-based thing that anyone can access and see where the music they want is, click it and get it, this can really help to, hopefully, change the tide a little bit. And that’s really what we needs to happen for the music industry.