Last month, we pointed that the reason Netflix's recommendation engine is so accurate is because the company employs real people to watch every single piece of content and thoroughly tag each show and movie.
Most of these employees, dubbed "taggers," are based in the US, where Netflix is headquartered and where it has the largest library of content. There are more than 40 in total, and we've learned that four of them are Canadians tagging our country's smaller Netflix library.
According to one, Jordan Canning, who spoke with Russ Martin for Canada.com, Netflix requires her to note every detail imaginable for each piece of content. Each tagger inputs close to 100 data points per film.
These data points create the tags that power Netflix's suggestio engine, which leverages algorithms to recommend content based on user behaviour.
Netflix didn't always do it this way. In the beginning, it tested out tags provided by external companies. But the company didn't find them to be adequately accurate. Humans do it best, Netflix says.
Oh, and for those hoping to land the dream job of tagger, watching movies all day—there are qualifications.
“We’re looking for people who have knowledge of movies and TV shows,” says Todd Yellin, vice-president of product innovation at Netflix. There's also a tagging test potential candidates must pass, he adds.
And the pay? It won't make you a millionaire, but Todd says part-timers make "a few hundred dollars per week."