Two new start ups are developing recommendation engines that focus on relevant data - a gushing hose of it - instead of relying solely on algorithms. If they unfold as their advocates believe, they could step up functionality in this particular tech genre.

One is Matcha.TV, highlighted by Lost Remote, now in private beta.  "The key for social TV guides to offer smart, relevant recommendations is user data, Lost Data writes. "The more you know about a user’s favorites and viewing habits - and those of his/her friends - the more relevant TV shows you can recommend." Matcha.tv connects an account not only to Facebook, but with Netflix, Hulu and YouTube as well and thus is able to mine the appropriate data for smarter recommendations.

Matcha.TV points to an earlier iteration - the now defunct Netflix friends feature - as its model. "It's one of the more commented on posts over on Netflix's blog, and still gets comments from subscribers pining for the return of that feature over a year later. People liked browsing and interacting with their friend’s queues."

Rexly, a data-driven recommendation engine for iTunes, featured on Beet.TV, is another example. What counts are actions - such as a user spending time and money on a particular item - instead of merely liking it, says CEO and co-founder Joel Resnicow. The company started with iTunes, "because it is one of the few distributions channels in history without a long tail,” he says, but it eventually plans to cover all categories and distribution platforms.