It’s been four years since CBS paid $240 million to buy London-based Last.fm, the much-vaunted online radio service. Since then, so much has changed — the founders have left, other big social music sites like Myspace are slowly dying off, and a new breed of young online music services like Rdio and Spotify have arisen.
And yet at the same time, so little has changed. On the surface, Last.fm today doesn’t look drastically different from the product CBS bought. The site has continued to grow slowly and has a fervent user base — but it is fair to say that it hasn’t set the world on fire.
That’s precisely what Matthew Hawn wants to change. At the turn of the year, the former record label executive — a longtime user, former journalist and music junkie — was brought in after a decade working with Universal Music and Sony as head of product.
I caught up with him earlier this week to find out — and heard a pretty straightforward message: There’s nothing wrong with our service, but there’s a lot we can do with it.
The way Hawn outlined it, there are several things that Last.fm wants to do.
Right now, the site is pretty confusing for new users. There are things that everybody understands (charts, for example) but the idea of the service — that it learns what you like and plays you more of it — isn’t particularly well explained. Great web apps feel like magic, and Last.fm needs to recapture some of that.
Hawn admits that Last.fm has, in the past, been confused about its identity. A few years ago, before Facebook really cemented itself as the benchmark for social interaction online, the site introduced a number of features such as journals and groups. Now, he says, those elements look pretty crusty, and it’s time to focus. Some of these elements are likely to be shuttered over time in favor of smarter, leaner social links.
“We’re not a social network,” he says. “We’re a service with great social features.” His watchwords revolve around pushing the idea of the “interest graph”: his argument is that the things you like are not the same as the people you know (your social graph). So that, in turn, means Facebook can’t replicate the same taste-making and discovery around music of a site like Last.fm. There’s a gap that can still be exploited. But although he’s prepared to work with Facebook, he’s also cautious.
“Facebook is the Walmart of the Web,” suggests Hawn. Its size gives you distribution that you require; but you have to watch out that you don’t get “chewed up” along the way.
Focus on music
When CBS first purchased the site, it became part of the company’s broad interactive division. The hint seemed to be that the “scrobbling” feature — the mechanism that tracks what you listen to across many different devices — could be converted across other media, such as TV. That dream, which never really felt like the right move for Last.fm’s music-mad team, seems to have been pushed aside. Now CBS owns online video search service Clicker and has moved Last into CBS Radio.
It’s a better fit, but that doesn’t mean it’s a service aimed purely at music obsessives.
“We cannot just be for music geeks,” says Hawn. So the site is going to combine ease-of-use with improved recommendations and data crunching to make it more than just a haven for the hardcore. There will be more emphasis on live music, more recommendations for festivals and gigs, more ways to encourage people to use the service easily. Although he admits that Last.fm already concentrates on too many things, a broader approach is one that he hopes can help it grow.
Sometimes that can come from surprising directions: the Xbox implementation is now, he says, the biggest source of streaming radio usage for the site. That’s opening up new directions, and new audiences.
Focus on data
Ultimately, Hawn suggests, he’s torn between whether Last.fm is a music service that does data or a data service that does music. Whatever side of the fence he eventually falls, it’s clear that data is a hugely important part of what Last.fm has to work with. It is now able to record listening data from more than 600 different devices and software clients — something that has built up a vast database of listening habits, some 50 billion scrobbles.
The answer, he suggests, may be to see the site not as a competitor to Spotify, Rdio, MOG, Pandora or others — but as the connective tissue that brings them together. It can collect data no matter what service you use, it can feed that data back to make your listening more enjoyable. Tim O’Reilly has long argued that Web 2.0’s real power was data — that “Data is the Intel Inside”, and perhaps now we’re at a point where that is mainstream enough for people to reap the benefits.
These are still early days in the company’s attempt to find a second act. Is this the right way to go?