While the glory days of reaching 75% of television households with just one commercial are long gone, automatic content recognition (ACR) stands to help advertisers calculate incremental reach. “That can certainly be valuable, says Jonathan Bokor, SVP, Precision Video, Publicis Media.

Bokor discusses the challenges posed by using disparate datasets depending on TV/video tactics in this interview with Beet.TV.

Where ACR can come into play is determining specific household reach—something not available with Nielsen’s panel-based data. If a marketer runs a general market campaign on broadcast and cable and also wants to use addressable or connected TV, it would like to know people if the latter have already been reached with the general market campaign.

“If we could distinguish between what is reach and what is frequency, that would be very powerful,” Bokor says. “It would allow us to be more integrated in the way that we reach those audiences and reduce the amount of frequency that we have and be able to re-aggregate reach.

At this point, the ACR market “is a fragmented one,” Bokor adds. He cites as players in the space Inscape, Samba TV and Samsung.

An up-and-comer in the ACR category is Alphonso, which was founded in 2013. Having initially focused on apps, the company has broadened its focus to gathering viewing data from companies like TiVo and directly from TVs and streaming devices through deals with manufacturers, as The New York Times reports.

“It would be really great if we had integrated, census-level ACR data across the whole range of households, but we’re not quite there yet,” says Bokor.

Right now, addressable TV campaigns have as their basis names and addresses that are matched to operators’ subscriber lists. “In connected TV, we’re typically using a combination of IP address and device ID. Each different tactic carries a different data consideration,” Bokor says. “But what we really want to get to is common data across all of our marketing. We don’t really want to have to stitch together all these different datasets that use different identifiers.”

In addressable TV, it’s “challenging” to use a digital target that we might be employed for online video, which is based on cookies or device ID’s, and convert that into an addressable TV target based on name and address, according to Bokor. “We want to get to the point where we can interchangeably use a single target definition or audience definition across multiple tactics. We’re not there yet, but that’s where we really want to be.”

This video is part of a series The New Marketplace for Television Advertising, presented by dataxu. Please find more videos from the series here.