NBCUniversal Taps Machine Learning to Tie Ads to Relevant Moments on TV

The tool identifies places for brands to buy spots in more opportune times

NBCUniversal's new tool will better match ads to relevant scenes. Getty

The next time you see, say, a wedding scene on a USA Network show followed by a champagne commercial, it may not be a coincidence.

NBCUniversal announced a new machine learning tool today that helps brands place ads around scenes relevant to their product across any of the media giant’s broadcast and cable properties. The Contextual Intelligence Platform analyzes programming scripts, closed captioning data and visual descriptors of both ads and shows to find opportune moments for a given advertiser to appear as well as an emotional gauge for each scene determined by proprietary algorithms.

Focus groups for ads placed with the platform thus far have shown an average bump of 19 percent in brand memorability, 13 percent in likability and 64 percent in message memorability, according to Josh Feldman, evp and head of marketing and advertising creative, NBCU.

The announcement comes as linear television providers continue to grapple with how to bring digital targeting practices to a medium that still largely operates on traditional phone-call media buying and manual ad placements.

NBCU is now working with three to five advertisers for the system’s beta-test, and is aiming for an official release date early next year.

Feldman declined to reveal which brands were participating in the trials, saying only that they were selected from a range of consumer categories that includes retailers, studios and packaged goods.

“Before there was machine learning, there was common sense,” Feldman said. “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. 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.”

The difference now is that NBC’s ad sales team will be able to use machine learning to make that approach more granular and tie together all of the company’s media offerings with the exception of live content like news and sports. The platform will also cover repeat programming like syndication and movie libraries to squeeze more money out of reruns.

“This is going to be a big part of our conversations with advertisers for next year’s Upfront,” Feldman said. “Our sales leads are fully up-to-speed on this; this is a project that Linda Yaccarino has personally blessed and thinks is a game-changer for us.”

While the platform will simply provide advertisers with a media buying plan based on relevant scenes, NBCU’s product team is already working on ways to expand it to the point where it will be able to place ads automatically. Denise Colella, svp of advanced advertising products and strategy, NBCU, said the Contextual Intelligence Platform is part of a broader push to better integrate linear television with digital advertising techniques.

“I don’t like to think of the products individually but think of them as a suite that builds on one another,” Colella said. “When you look at digital premium sites, you’re looking at thematic advertising in context. The goal is to bring that to linear television and to make sure the ads match the audience involved, the audience matches the context involved and that we’re able to carry that experience across platforms. That’s really where we’re going.”

To that end, NBCU also announced today that it would be expanding its lucrative Prime Pod targeted advertising format and CFlight advertising metric for live, on-demand and time-shifted commercial impressions to all of its properties next year.

@patrickkulp patrick.kulp@adweek.com Patrick Kulp is an emerging tech reporter at Adweek.