When Kimberly Foerster was helping her au pair fill out a profile on a dating website in 2009, she had an epiphany: finding the right movie or TV show was similar to finding a new love.
“I thought, ‘That’s how you should decide on entertainment — just like Match.com — because for me, characters are always the reason why I like something,'” said Foerster, a former investment banker based in Chicago.
This notion sparked her to build Charactour, a website launching Tuesday based on a character-based algorithm intended to suggest not just TV series or movies, but books and video games. The site supplies links to online stores or streaming services where the character’s title is sold.
A collection of 4,500 profiles are devoted to characters from a wide range of entertainment, each detailing everything from their childhood upbringing to relationship status. Charactour offers multiple ways to use that database to connect consumers with content; one in particular has the user take a personality quiz that deciphers which characters are most like them.
For example, users can rank their maturity level on a scale of 1-5, with Ron Burgundy from “Anchorman” as the lowest option and Gandalf from the “The Lord of the Rings” franchise as the highest.
Using a continuum of 26 traits — such as brainy or homicidal — the algorithm assesses how similar a character’s traits are in relation to another character. “We have all these different personality traits that are reflecting a character in a similar way that Pandora goes through all of the different elements of a song,” said Foerster, CEO and co-founder of Charactour.
Other ways Charactour matches users with entertainment is by letting them select a collection of traits that they want their character to have and another that blends together two characters in the site’s “mashup” machine. “Some people think of it as if two characters had a kid, who would it turn out to be?” said Pete McEntegart, the other founder of the website and its editor-in-chief.
A fourth matching method is the “CharacTournado,” which randomly generates character ideas.
The site also includes a natural language processing machine, which analyzes a character’s dialogue to provide a character’s automatic traits assessment. CharacTour’s algorithm is the first of its kind, prompting the startup to file for a provisional patent for the character-based search engine last year.
Having raised $700,000 from friends, family and angel investors, Charactour aims to put a more user-friendly face on content discovery than what is typically available from popular streaming services. “We did hear from our beta users that this was a fun experience for them,” said Foerster. “It’s not like Netflix where they were just staring at DVD covers.”