Pandora’s Guide to Mobile App Platforms

Pandora has had a charmed history with mobile platforms — the company’s music streaming app is one of the most popular on just about every type of phone out there. It just makes a lot of sense to have personalized radio in your pocket. But with a small developer team and a confusing emerging mobile platform world, it wasn’t as if Pandora always knew what to do or even why it was succeeding.

In a recent talk at Digg, Pandora CTO Tom Conrad traced the history of his company’s experience with mobile platforms, making a number of counterintuitive observations about what worked and what didn’t.

Conrad somewhat infamously said two years ago — before the rise of the Android (s GOOG) platform — that: “I need Android like I need a hole in my head.” He was worried about the fragmentation between carriers and handsets that an open mobile platform would bring. And yes, that’s largely happened, but it’s not that bad, he said while at Digg. In his experience other mobile platforms are just as fragmented, but in different ways. For instance, it’s not enough to make Pandora compatible with every type of BlackBerry (s RIMM) phone; you also have to make it compatible with all the different corporate deployment processes that control the management of so many of the phones in the market. And today, Android usage of Pandora is set to pass BlackBerry and is on pace to rival the iPhone (s AAPL).

Digg Technical Talks – Tom Conrad from Digg Development on Vimeo.

Probably the single most annoying compatibility problem Pandora had was back before the ascendance of smartphone apps, when it slaved over making an app for the Motorola (s MOT) RAZR despite the phone’s “tiny amount of RAM and mediocre processor,” as Conrad put it. The app was running fine on the black and silver editions, but then Motorola released a pink option on which Pandora didn’t work. It turned out that pink wasn’t just a color change, but a new version with cost reductions on the chipset that introduced new bugs in the app.

What doesn’t matter to mobile app success? In Conrad’s experience, he thinks first-mover advantage is overrated. That’s easy for him to say, given Pandora has never really failed and has gotten very early access and advice about new platforms from Apple and Palm (s palm). But as Conrad pointed out, Pandora has overtaken apps like Slacker that worked to get on Android early. Plus, as many as two-thirds of mobile Pandora users aren’t associated with online accounts — “so it does feel like there’s still customer acquisition opportunity,” Conrad said.

“Each platform has its charms,” said Conrad — and of course its particular screen resolutions, app approval processes and other factors that make it challenging to launch and update apps. Pandora’s mobile users tend to be male and young, he said. They listen to 13 hours of music a month, on average, but in shorter bursts as compared to 17 hours across longer sessions on the web. So what should you do to ensure your apps succeed? Conrad advised:

Know who your users are. Sprint (s s) included Pandora with mobile plans in 2007, meaning the cost to buy and run a Pandora-capable phone could be much lower than buying a new iPhone. But people who want to buy entertainment phones turned out to be more likely Pandora users, Conrad said. iPhone buyers were often historical iPod owners, conditioned to use their portable devices to listen to music.

The device matters. When Pandora launched on BlackBerry, it came out on every device but the Storm. Conrad thought the Storm was weird, and that the touch interface would benefit from a dedicated implementation at some point later down the line. But a Pandora developer in Canada (where iPhone data plans were ridiculously priced) took it upon himself to write a Storm app on nights and weekends, and after it was released the Storm quickly accounted for 60 percent of all Pandora installs on BlackBerry. It turned out Storm buyers were very similar to iPhone users.

Conversely, Pandora on the iPad has only had modest success. So far the iPad isn’t a great mobile music environment, despite new user experience improvements enabled by the large touchscreen. Without multitasking, turning on Pandora makes the device a “pound-and-a-half paperweight,” Conrad said.

Accept offers of friendship. Palm (s palm) wooed Pandora as a webOS launch partner. Conrad was skeptical — and webOS didn’t end up succeeding — but it gave Pandora the ability to play with great features like multitasking. Plus, it led to Pandora’s first TV commercial.

User experience is key. Pandora’s apps are designed in-house, so it has been able to control and perfect them.

Plus, some things are obvious. Phones with headphone jacks are more likely to be used for listening to music. Pandora is free — people like free stuff.

Related content from GigaOM Pro (sub req’d):

The App Developer’s Guide to Choosing a Mobile Platform