Why Google should buy Like.com

I watched the Blogosphere react to Like.com's launch this week. They were victims of their own hype and many influencers seem to walk away disappointed. However, look past the teething problems, lack of content and the need for user-experience tweaking, lies a very potentially compelling technology. The fact that many didn't get it might be a good indicator of either being still too early or being a potential market disruptor. Tagging pictures and video is a chore that is neither done well if done at all. That fact that the technology comes from face recognition technology probably means they have multiple patents sewn up. Take what they are doing and apply it to Google images and maybe this is where Google could expand its dominance in the search space. Tagging is not always efficient, how do you show that two handbags are similar because of their buckles in a group of several hundred handbags all tagged the same way. Other potential suitors maybe Amazon or eBay where there are millions of items potential items that do could with visual filtering in addition to textual filtering.

On a similar note, it is interesting to note that they were able to reinvent themselves after realizing that as a face recognition technology, the accuracy required to gain adoption was not acceptable. While we may laugh when presented with 5 people who are not my Uncle Bob but may look like him, I think consumers may react differently with 5 different dresses, watches or other items that we may also like. As a visual recommendation engine, it is one of the most exciting and potentially useful technologies in the Internet for some time.