Although it’s been a while since we heard anything from DJ-focused music service Turntable, the team has kept busy. Yesterday it debuted a new smart-radio service called Piki that offers yet another spin on music discovery.
No one has quite cracked the code when it comes to finding a perfect balance of hearing music you already like and discovering new songs you probably will end up liking. Piki’s answer is to create algorithms based on what your friends are listening to rather than grouping songs together because they sound similar or belong to a particular genre. It’s structured like Pandora, rather than an on-demand music service like Spotify, and it’s iOS app is available for free. A good comparison for Piki would be a mobile version of “Twitter for Music” service Blip.fm — except with more of a focus on passive listening and discovery.
“Pandora is really good at matching songs that sound like other songs, and it creates this very narrow path of music. That works to some degree because Pandora has hundreds of millions of users,” Turntable and Piki co-founder Billy Chasen said in an interview with VentureBeat. “What we’re saying is that it can be better.”
How? Well for starters, you stop matching songs and start attempting to match up people, Chasen told me, adding, “The way I’ve always discovered new bands is when a friend will tell me, [for example] ‘look, there’s this guy out of Montreal that does crazy shit with his violin and it’s great.'”
But matching people with the same musical tastes is easier said than done.
In fact, I can name several services (Last.fm, 8tracks, Rdio, Fuzz, Playground.fm) that made the same claim while providing dismal music discovery, at best. Chasen explained this is because other services make the social portion of their services more of an afterthought. For instance, adding social network sharing buttons and deep Facebook integration isn’t actually social, he said. I’d lump sharing playlists with strangers in the same boat.
“A lot of music apps have this idea of powering music [discovery] with just your friends, but that’s not the whole recipe,” Chasen said. “You probably have a few friends that have good taste in music, but most of them you won’t really like. It’s time to make some musical friends, which is something we saw happen on Turntable.”
And Turntable was a legitimately cool and genuinely original idea for a streaming music service when it debuted in 2011 — quickly spawning a handful of copycats that have all either hit the deadpool or pivoted. That said, Piki has my attention, especially since Turntable isn’t disappearing or transforming into Piki.
“[We were] continuing to see that there’s this opportunity to insert a social layer into music,” Chasen said. “Turntable only touched the surface of that — being in a room and playing music with people is social, but there’s room for a lean-back music service that does more with those social connections.”
When you sign up for Piki, it asks you to “pick” a favorite song, then it prompts you to connect with friends on Facebook, Twitter, Turntable, and your iOS contact list. The service is still new, and I didn’t have many real-life friends to follow, so Piki suggested users from the community with similar tastes. The service then populates a radio station based on a few things: your song picks, who you follow, and how often things are picked or played across the community. Unlike the Pandora method of creating custom smart radio stations, Piki wants you to “tune” a single station. You can mashup genres from the sidebar controls, which pulls in songs from those particular genres that were “picked” by those you’re following.
The more social side of Piki involves adding notes to songs that you pick, or dedicating a song pick to another Piki user. All your picks pop up within a feed on your profile, which others can check to see if their musical tastes match up with yours.
Initially I wasn’t impressed with Piki, but after playing with it for a while I can say it has a lot of potential to replace some of the other music services out there. In addition to “repicking,” a song played on your own personal station, you’re basically asked to add your “picks” whenever you have the urge to listen to a song. There’s even a Shazam-like listening feature that lets you identify songs when you’re out and about. And since it plugs those songs into a smart radio algorithm, I’m more likely to use Piki over Shazam, which more or less just panders to the top 10 most popular songs/albums/artists. (If I wanted that, I’d just go to the iTunes music store.)
Piki also has the benefit of learning from some of the challenges Turntable faced when launching. Chasen told me that launching Piki via mobile devices first was a conscious decision, not only because that’s where people want to listen to music, but also because it’s a lot more useful in terms of what you can do. It’s also a really smart move considering how long it took Turntable to make an appearance on mobile devices. The development team at Stickybits, (the company that produces both Turntable and Piki), has also been working on the service for well over a year and has the benefit of already having music licensing deals in place that carry over from Turntable.
That doesn’t mean Piki won’t have challenges, namely rumored competition from Twitter as well as several others trying to crack the social music discovery problem. There’s also the matter of DMCA licensing, which doesn’t allow you to play full songs from other users’ profile pages/feeds. (Doing so would make Piki an on-demand music service, which is much more costly.) You’re also limited to whatever is in Piki/Turntable’s 13-million-track music library or your own music library. So, if I hear an unsigned band’s performance on SoundCloud or YouTube, I won’t be able to add that song as a pick.
Chasen said he’d like to add support for YouTube, SoundCloud, and other services in the future, but for now the development team is focused on Piki’s initial launch. The team is planning to release a Piki web app and Android app in the near future.
VB's research team is studying web-personalization... Chime in here, and we’ll share the results.