Data-Driven Future Of TV: Will Spots Be Cherry-Picked?

Audience cherry-picking and cherry-packaging is coming to TV this upfront season. Two weeks ago at NATPE in Miami, both Group M's Irwin Gotlieb and Initiative's Kris Magel told us that their media buying shops are moving aggressively to leverage consumer viewing and purchase databases to buy and re-aggregate fragmented audiences on TV -- just as their agencies do on the Web.

Today, two-thirds of television viewing in the U.S. occurs on shows with a rating of 0.5 or lower. Since audiences this small are hard to measure accurately beyond sex/age demographics with current TV measurement, the vast majority of these spots are sold in bulk at low rates as rotators or for direct response. Rather than commanding the $15-$35 CPM rates of the best prime-time shows, these spots are challenged to command $1-$4 CPMs.

As Gotlieb and Magel foreshadowed, that is about to change. For example, if an agency knows which spots among the rotators they have bought are likely to index at a 200 for Big Box's retail shoppers, they might schedule an ad for Target in that spot. For spots likely to index at 175 for working moms, they might schedule a spot for Papa John's Pizza that promotes the high quality of its ingredients.

Basically, the buying agencies will use data to find high-value viewers in low-cost rotator inventory and use data-driven scheduling at the spot level to turn $3 CPM inventory into spots that perform like $20 CPM premium spots. To do this, all they need are highly granular targeting databases, which virtually all of the big agencies are building today, and the right to pool and proactively manage scheduling for rotator spots among multiple different clients -- which will probably be their quid-pro-quo for rate increases for networks in this coming upfront.

TV media owners, meanwhile, who see this trend coming, are starting to build their own audience databases to become cherry packagers rather than be cherry-picked. Rather than delivering under-appreciated inventory in bulk to buyers, they might use data to identify and package high-value audience spots for their own, charging a significant premium for the big box shopper spot or the health-conscious working mom, capturing as much of the $3 to $20 uplift for themselves.

This will be fun to watch. Will media buyers and their clients get most of the uplift? Or will TV media owners get the bulk of this through price increases? Does one of the sides even have to lose at the expense of the other? Maybe they both win? What do you think?

Whoever has more and better data will have an advantage over those who don't, whether it's relative to other buyers or relative to sellers, who without data won't know how much value they might be giving away.

On the buy side, I'd certainly be placing bets on Group M and Gotlieb, given his and their history in successfully pioneering winning data-driven buying strategies and systems for TV media in the past. And, it would be hard not to bet on Magel and Mediabrands' Matt Seiler, given their prominent and vocal evangelism in this area over the past year. Also, I'd look to smaller, more digitally focused agencies that also do TV, since they might be more nimble.

While concepts like programmatic buying and addressability have been getting a lot of headlines lately as folks speculate whether the television industry is ready to finally adopt Web-born technologies at real scale, this upfront might be much more about data and old-fashioned spot scheduling. What do you think? The application of massive volumes of highly granular data to the TV ad buying process has picked up a lot of steam lately.