ThinkAnalytics Pushes Deployments Past 70 Million Mark

20 major global service providers are now using company's search and recommendation engine

By George Winslow -- Broadcasting & Cable, 6/28/2012 3:02:10 PM

ThinkAnalytics has announced that its multi-platform television search and recommendations engine has now been deployed in by operators serving over 70 million viewers in 16 countries in eight languages.

The company also noted that "20 of the largest service providers in the world rely on ThinkAnalytics' technology," including two major ones in North American that it was unable to name.

Peter Docherty founder and CTO of the company noted in an interview that the explosion of the amount of content and the proliferation of devices allowing people to access that content was prompting significant interest among operators in deploying more advanced search and discovery engines.

"We can't expect consumers to recognize value in programming that they don't know exists," he said. "Nor do we believe they should have to work so hard at getting better recommendations by filling out forms or relying on a ‘lean-forward', Web-type search."

The problem applies to all devices but is particularly difficult with settop boxes and other devices that serve entire households, where individuals have different viewing habits. "No one is happy with the current results," he noted.

To overcome that problem, ThinkAnalytics has pre-tagged a library of about half a million movies with metadata to improve its recommendations. It is also applying a variety of different approaches to search and recommendations that go far behind one single approach, such as the recommendations engines that are simply based on past purchases or the Amazon-type recommendations based on suggesting additional movies that had been watched by other users who had viewed a specific film.

Better recommendation also had a big payoff for operators both in increased content purchases and reduced subscriber churn, Docherty argued. He noted that customers receiving their recommendations purchased 50% to 60% more content and that their recommendation engine has shown itself to "improve customer loyalty and retention."