A small Washington D.C. startup named Qloud is beta-launching today a collaborative filtering system for music along the lines of Last.fm. If it can achieve critical mass, the startup wants to be a social search engine for music, built on users tagging their iTunes tracks using an overlaid Qloud plug-in.
I had some technical difficulties testing this product, most of which were due to the fact that Qloud doesn’t have a Mac plug-in yet (it’s in the works). Still, this is so rough around the edges, much of our discussion will have to be around what Qloud founder Mike Lewis told me about the company concept.
Since iTunes doesn’t have tags itself, creating playlists means manually dragging songs together or sorting by name or genre. I’m not an avid tagger myself, but I do see the benefit of using tags to create mixes (for instance, a distinct “mellow” playlist, “female singer” playlist, and “workout” playlist could be composed from the same library of music, and there would be some overlap).
The plug-in sends user input and activity up to the Qloud web interface, where users can search by music, tag, or person, refining any search by demographic information — for instance, limiting the tag “dance” to when it is used by women under 20. Demographics might be a bit crude, but tracks tagged “dance” music by 40-year-old man is likely to be totally different, Lewis points out.
Qloud will make it simple to export any search to an RSS feed, an email, or most importantly, a playlist to get onto your iPod. (So, for instance, you could subscribe to Lewis’ top 50 tracks in the last 30 days.) But music discovery is awfully hard without being able to sample new tracks. Qloud can only do the same thing as everyone else — give you 30-second Real or Windows Media Player clips, and direct you to iTunes and Amazon. Lewis said he’s thinking about adding a subscription streaming service for about $3 per month, but he absolutely does not want to get into doing full downloads.
Lewis and cofounder Toby Murdock both worked with digital music at AOL and Ruckus. Their development is done by a full-time team of 10 in Romania. They raised a round of “under a million” in June from a source they wouldn’t disclose, but we have a pretty good hunch it’s Steve Case.
The competition for music recommendation is intense, with Last.fm, Pandora, MyStrands, and MOG among the funded companies. That’s not even including projects from established music outlets as well as mobile and consumer electronics players. And for now, this is definitely a niche market.
The big picture, asserts Lewis, is that a good, subjective search engine will jumpstart long-tail music discovery. I do think tagging is a big hassle, but sorting music by tags within iTunes using Qloud is actually pretty neat.
It makes sense in the way del.icio.us and Flickr do, because it’s personally useful. However, there’s a heck of a lot of competition; some future version of iTunes could very well include tags; and social search for music is a very fuzzy concept. The best possible option for Qloud would be for Yahoo to buy it to complete its tagging full house.