The Curious Case of Twitter’s #Music App

April 2013

Yesterday, Twitter announced #Music, their music discovery platform. It looks beautiful, but conceptually it’s a bit of a mess. (The Verge have a good #Music hands-on, but the idea is that #Music crawls Twitter for links to songs in order to calculate which artists and tracks are ‘trending’ on the network, then allows you to play through those tracks. It also allows you to view which musicians other Twitter users follow, and which tracks those Twitter users have played recently).

Twitter, the company, have a very real issue, which is that they’re bounded by the limitations of Twitter, the service. Twitter is a social network defined by resource scarcity: users are provided with 140 characters at a time, and it’s working out how to communicate within those constraints that makes the service compelling. Because the limitations of Twitter are so clearly defined and so fundamental to how the service is used, it’s difficult for Twitter HQ to mess with the core Twitter experience.

Twitter HQ have been moving in two directions recently. First, they’ve been trying to slowly shift our understanding of what a “tweet” means, by rolling out “expanded tweets” and “cards”, which allow some publishers to package rich media and long article summaries into tweets. In general, though, users still expect the core Twitter experience to revolve around 140 character bursts of content: it’s very difficult for Twitter HQ to innovate much around that limitation.

The other direction Twitter HQ are pushing is in launching “sister apps” that augment the Twitter experience without actually changing how Twitter works at its core. Instead of rolling innovative looping video sharing into Twitter directly, for example, Twitter HQ launched Vine. Now we have #Music.

The core issue with #Music is that it builds a music discovery layer on top of Twitter, which is not itself a music discovery service. There’s a kind of cognitive dissonance at play in #Music, which makes the service at times difficult to really understand. For example, when you click on an artist in #Music, you’re presented with a standard Twitter “Follow” button, which conflates “Liking a musician and wanting to hear more of their music” with “Wanting to subscribe to their stream of 140 character updates”. Similarly, all Twitter users, whether they’ve signed up for #Music or not, now have a new profile page at[username], which lists all the (verified) musicians they follow on Twitter. This makes some degree of sense, but not a great deal of it. After all, most of us don’t follow Twitter users to explicitly express our musical taste, and following a musician on Twitter is not necessarily an endorsement of the music they produce. Does wanting to follow the bizarre 140 character musings of Kanye West mean you’re necessarily a fan of Kayne West’s music? To #Music, yes.

In a similar vein, we don’t necessarily follow other users on Twitter because we share their musical taste. At this stage, though, there’s no way to stop seeing recommendations from a user you follow on #Music unless you unfollow them on Twitter, because the two services share the same underlying network. Twitter now provide two distinct reasons to follow users: to follow their updates (on Twitter), and to track their musical taste (on #Music). The “follow” action has been overloaded.

Because Twitter is not, at its core, a music discovery service, many of #Music’s features are necessarily clunky. The #nowplaying feature of #Music lets you track what your friends are listening to, but in order for the track you’re listening to to appear on #nowplaying, you must explicitly post a link to that song on Twitter. That’s great for Twitter HQ, because anything that encourages you to share more on Twitter is a win, but is it elegant? Platforms like Rdio and Spotify are already able to accurately assess what your friends are “#nowplaying”, because they can actually track what your friends are listening to… because that’s what they’re built to do. Twitter HQ’s #nowplaying solution is essentially a hack on top of Twitter, and an inelegant one at that, because it encourages users to junk up their Twitter streams with links to tracks on iTunes, Rdio, and Spotify. Again, it’s very clear why Twitter HQ would want to encourage this, but that doesn’t mean it’s a net win for users. Nor is it clear why #nowplaying should even exist: services like Rdio and Spotify already do the same thing much better.

What it comes down to is this: the only way for many users to experience #Music properly is to change how they use Twitter. If the only way to tweak your recommendations on #Music is to start following musicians or users with great musical taste, Twitter HQ have now complicated the act of “following”, and if the only way to share your musical taste on #Music is to tweet links to songs, Twitter HQ have now complicated the act of “tweeting”. Twitter HQ have built a music-sharing and artist-discovery service on top of a service for sharing short bursts of information. There’s a strange misalignment between the two services.

Ultimately, Twitter HQ are stuck between a rock and a hard place. It’s fairly clear they’re pretty sick of Twitter as a conduit for 140 character group text messages. 140 characters of text isn’t sexy. Music and videos are sexy, but Twitter HQ can’t revamp Twitter to make it sexier, because at least a sizeable minority of Twitter users are wedded to Twitter’s unsexy limitations (and, in fact, paradoxically find those limitations very attractive). In order to explore the hot and lucrative multimedia stuff, Twitter HQ must work out a way to build on top of Twitter while at the same time ensuring that the core Twitter experience remains unsullied. Twitter’s #Music experiment shows just how difficult that’s going to be.