NBCUniversal has built a machine learning tool to target TV ads

NBCUniversal has a plan to use machine learning to drill down further into data about its programming so it can better target consumers and serve relevant television ads.The Comcast-owned broadcaster has cooked up something called the Context Intelligence Platform, which scours scripts, closed captions and visual descriptions to pick the right spots for the right ads.Josh Feldman, head of marketing and advertising creative for NBCUniversal, told Adweek that trials of the platform are underway and have produced 19% more brand memorability, 13% more likability and 64% more message memorability.“When my team created something for an advertiser that was a really heartfelt piece of marketing, we would make sure that the spot ran on heartfelt-type programming as opposed to, say, slapstick comedy,” Feldman told the publication, according to Engadget. “We've been doing this on a manual basis for a long time—but now we're going to be able to do it at scale.”Like so many other TV programmers and providers, NBCUniversal is working to make its advertising more addressable and relevant so it can better compete with digital giants like Facebook and Google that have eaten into the overall U.S. advertising market.Earlier this year, NBCUniversal joined OpenAP, a cross-audience ad platform founded by Fox, Turner and Viacom. In addition to joining OpenAP as an equity partner, NBC licenses to the group its Audience Graph, a data set that provides a centralized store of audience attributes.