MeeMix: A New Breed of Music Personalization is Born
Typically, there are two ways to perform such predictions:
- “Nature” (Pandora) – Decision making is based upon on the compatibility between one song and another—personal taste not being part of the equation.
- “Nurture” (Last.fm) – Decision making is based upon a environmental factors—song compatibility not being part of the equation
Steve Krause wrote a great post about Nature vs. Nurture in the music space.
Like Pandora, MeeMix breaks down each song, artist and user down to an elemental level. Then its recommendation engine analyzes the combination of a user’s profile, music selection and behavior to create a customized radio station. In my personal tests, MeeMix’s music selection was near perfect.
MeeMix was founded in mid-2006 by CEO Gilad Shlang, formerly the User Experience Manager at 888.com, and Dr. Ricardo Tarrasch (Chief Scientist), also formerly of 888.com with a doctorate degree in Biopsychology. The two possess in-depth experience in creating and leveraging user experience and communities based upon algorithmic forecasting. The company is funded by angels and currently employs a team of 9, including the former lead developer of ICQ, musicologists, and a community/SEO specialist.
The business model is currently based on album sales and will expand to songs, ringtones, and ads. Seeing as there’s so much algorithmic magic under the hood, I won’t be surprised if in a year or two licensing its engine will become MeeMix’s core business.
In the mean time, the team is focusing on learning its community and expanding its service both in scale and in application range (embeddable web widgets and a Facebook app).
Try MeeMix
We have 1000 invitations available to the MeeMix private beta here.
This guest post is written by Roi Carthy, an Israeli-based startup consultant. Roi will be covering the Israeli startup scene for TechCrunch.