By Kyle Bylin | @sidewinderfm
In late 2010, Pandora set out to create a new Internet radio experience.
Pandora wanted to rebuild the popular service to be “faster, easier, and more social” while remaining familiar to longtime users. By July 2011, Pandora started to roll out its new site to subscribers, collect their initial feedback, fix programming bugs, and make further improvements.
Months later, the “New Pandora,” which now allowed users to share songs with friends and learn more about favorite artists, launched for everyone. After over a year of testing and perfecting the user experience, Pandora finally released brand-new mobile apps for iPhone and Android in October 2012.
The journey from “clean sheet of paper” to “entirely new Pandora” took over two years to pull off. For the first time, the Oakland company provided a consistent listening experience and feature set across every platform. Most recently, it also introduced a TV-centric service for gaming consoles.
It’s indisputable that Pandora is a product for the mainstream market. It has over 200 million registered users in the United States — 70 million of whom tune in every month. Its dead-simple interface and tightly refined stations have proved a winning combination for music listeners.
If you have a commute that you want to fill with music, Pandora will cue up 30 minutes of songs that sound like Imagine Dragons. The similarity may soon veer toward banality, but the odds are that you will enjoy the music.
By now, the criticisms of Pandora have become well worn: repeated songs, little variety, limited skips, and frequent ads. The upside to the popular service (or why listeners tolerate the latter) is that it does the work, tailors the mix, suggests new bands, and maintains a mood. The convenience of Pandora, in other words, seems to outweigh the trade-off of having a limited experience.
As time goes on, though, that strength has become a weakness.
Take the song “Odds Are” by Barenaked Ladies, for example. If you click around the social charts in Twitter #Music’s mobile app, you can find the catchy new rock single pretty easily, and if you’re paying subscription fees to Spotify or Rdio, you can listen to it repeatedly. But here’s the rub: It’s still very difficult for Pandora users to stumble upon “Odds Are” and fall in love.
The “filter bubble,” as author and activist Eli Praiser calls it, prevents users of Pandora and other such services from being exposed to songs outside their personal universe of recommendations. If you aren’t in the orbit of Barenaked Ladies or similar artists, the “chances are so small” (as the lyrics go) that you’ll find that song.
This lament about Pandora’s lack of serendipity has been leveled against it many times before, but its “musical cocoon” is becoming harder to ignore.
Erin McKean, founder of the online dictionary Wordnik.com, eloquently stated in her 2007 TED Talk that “Serendipity is when you find things you weren’t looking for, because finding what you are looking for is so damn difficult.” On Pandora, discovering songs that sound like “Odds Are” is the easy part, but uncovering the hidden gem itself is damn difficult.
More worrisome is that Pandora is also bad at serving up favorite music.
Popular songs by seed artists frequently appear on radio stations, latest singles emerge in relevant places, and “Quick Mix” offers a singular listening experience with greater variety. But that’s all. There’s no way to solely listen to favorite songs, discover music by unexplored artists, delve into a stream of loved tracks by relevant users, or hear the latest songs by favorite artists.
To be fair, these are ideas that may not appeal to the widest market; they may have failed to gain traction upon research and testing.
If that’s the case, that casual listeners don’t care about these things, it makes sense. But it’s still a missed opportunity, as helping listeners find out when favorite artists release music is a valuable utility that no one does perfectly.
Personal discovery should be integrated into the stations you engage with; it should not be work or create work; it should play out naturally and happily.
But the “real risk” for Pandora, writes Max Engel, VP of Product at Web publisher SpinMedia, is that “music recommendation is becoming increasingly commoditized.” When rival companies like iHeartRadio and Spotify can leverage the music intelligence platform built by The Echo Nest to develop comparable experiences without considerable investment, it makes differentiation for Pandora difficult while making its “Music Genome” database look quaint.
Mr. Engel also worries that Pandora “solves a feature problem, but not a service one.” Spotify and Rdio have radio features, he says, so users have “little reason” to leave their primary service to get a radio experience.
There is a belief that Pandora has cultivated deep loyalty among music listeners, and that the switching costs of leaving the popular service for a rival would be too high. But are they? A student will abandon Pandora once she notices that more of her friends are on Spotify Radio. A father will switch from Pandora to Spotify Radio because it plays fewer commercials. A gym-goer will leave Pandora for Spotify Radio because it has far more variety.
The new Pandora suffers from the same old complaints: It has blocked a few fatal blows, but it’s near a thousand paper cuts. Tim Westergren has built a single island in a bloody ocean, and it’s showing signs of deep erosion.