Swell Radio

» see swell.am as it lived, on the wayback machine

~ the eulogy ~

Swell was iOS-only and alive for thirteen months. I have no memory of it at all, and I was working as an engineer through the whole of its short life. A podcast app in 2013 that never shipped for Android and never made it out of the English-speaking market had a very small window in which to reach anybody, and it did not use much of that window.

The company was Concept.io, out of Mountain View, run by G.D. Ramkumar. The app soft-launched in Canada early in 2013 and opened properly in the United States on 27 June. Draper Fisher Jurvetson led a $5.4 million series A. The description that stuck to it in the press was Pandora for talk radio, which was accurate enough to be useful and lazy enough to be repeated in every article.

What it actually did was refuse to show you a list. You opened it and audio started, and from then on the only inputs were listening and skipping. It read the skips as signal, along with community ratings and whichever topics and organizations you kept coming back to, and it rebuilt the stream around that. ABC News Radio, NPR, PRI and TED were among the partners. No advertising and no subscription, which meant no money coming in from any direction.

~ the dossier ~

~ what apple wanted ~

On 28 July 2014 the reports appeared: Apple was close to paying around $30 million, some accounts putting it at $31.8 million. By the 30th the trade press was writing about it in the past tense and the app was going away.

Apple's podcast situation at the time was a directory bolted onto iTunes and a first-party app that nobody defended. Swell was a recommendation engine for spoken audio that already worked, and its team had form. Ramkumar and several colleagues had built and sold an image recognition startup, SnapTell, to Amazon a few years earlier. What Apple bought was the engineers and the approach, and what it did with the product was take it off the store.

~ the shape of the deal ~

An acquisition where the app closes on announcement is a particular kind of transaction, and it was common enough in 2014 to have its own vocabulary. It also leaves nothing behind for the people who were using the thing. There was no export, no successor to migrate to, and no promise that any of it would show up elsewhere.

Some of it did, eventually, in the sense that Apple Podcasts stopped being a directory and started making suggestions. Whether any Swell code survived into it is not something Apple has ever said, and that silence is the standard terms of this kind of deal. A company gets bought, the users find out from a press report, and the product they had been opening every morning is simply not on the store the following week.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »