“We’re not optimizing against an addressable human audience,” said David Nyurenberg, associate director of digital video at Rain the Growth Agency. “We’re looking at all the signals and [seeing] which signals correlate to conversion.”

More precise and quantitative targeting

Of course, targeting based on contextual signals is not new. Marketers can currently target off the IAB Tech Lab’s content taxonomy, which distinguishes content like sports from news.

But metadata allows for more precise and quantitative targeting. Sincera can compare the ads-to-content ratio on television shows like 2 Broke Girls, where 11% to 12% of the program’s runtime has ads, versus Cops, which is 17% ads, said O’Sullivan.

The signals can help brands still buy inventory on the web amid signal loss. Metadata can also be used to buy better media in CTV, which doesn’t typically use cookies, but where finding quality programmatic inventory can be difficult since most ads are bought directly.

“Demand-side platform algorithms will naturally optimize to long-tail publishers,” Nyurenberg said. “By removing the addressable audience piece, we were able to optimize against metadata [toward] premium inventory.”

But targeting with metadata in CTV has its limitations because not enough inventory is available programmatically to use a purely metadata strategy, said Robin Cohen, executive vice president of integrated media investment and planning at Rain the Growth Agency.

Nonetheless, metadata strategies represent the paradigm shift in programmatic advertising.

“Instead of knowing who the users are,” Manning said, “you have to profile the information you have about the impression.”

An earlier version of this article incorrectly stated that Rain the Growth Agency used algorithms from Chalice Custom Algorithms for a CTV campaign.