Israeli startup Jinni, which calls itself a “Pandora for movies” in reference to popular music-recommendation service Pandora, has raised $1.6M from DFJ Tamir Fishman in a first round of institutional funding. Where Pandora uses subject experts for its music recommendations, Jinni’s technology crawls the web and picks up movie reviews and metadata and uses natural language processing tools to tag films and tv shows according to what’s being said about them.Users click on movie-image tiles to search Jinni’s “Movie Genome” by mood (pictured above), title, genre, audience, time/period (pictured below), place, and praise. It also has an audience categorization option, with subcategories such as “boys’ night,” “girls’ night,” and “date night,” and offers suggestions based on each. Each result provides an overview of the film as well as the option to view a list of other “more like it.”
As with most recommendation engines, the more you use it, the more tailored the results will be. Jinni will create a taste profile for a returning user and users can also create various “taste personalities,” depending on their mood, said CEO Mike Pohl. Using community features, users can share their picks and see what others are doing. Users can also power their Netflix account using Jinni.
Pohl says Jinni is the only player in the movie space to offer recommendations and a bundled solution — browsing, recommendation, and taste profile in one service.
The recommendation results link to film providers, such as Netflix, Blockbuster, Amazon, and Lovefilm, but there are no formal relationships other than APIs yet in place. Pohl says he hopes to begin to power specialty sites for such potential partners. He sees three potential business/revenue opportunities for Jinni: a destination site to expose users to assets, a tool for helping retailers in sales or downloads of film or tv assets for a fee, and as a technology to be licensed to providers to allow people to find content based on their personal tastes.
Jinni opened the service up to the public about six weeks ago and had over 40 million unique visitors in November, a result of viral growth, according to Pohl. [Update: a spokesperson from Jinni tells me that figure is incorrect and that unique visitors for November were instead 4 million.] The company, which launched in 2007, competes with sites such as Tastekid, which recommends music, movies, and books based on artists/bands, movies, or books you tell it you like, Clerkdogs, which bases recommendations on movies users say they like, Criticker, which makes recommendations based on other users’ tastes, Nanocrowd, which makes recommendations based on viewer comments on the web, and IMDB, which is more a database than a recommendation engine.
With this latest round of funding, Jinni plans to expand the development team and catalog choices and plans to conclude some agreements in the first quarter of 2010 with customers who would use Jinni’s API. It will decide at that point on the need and timeline for pursuing additional funding. Jinni received previous funding from angels, friends and family.