Commentary

You Can't Do Programmatic TV Without Good Data

Programmatic ad buying technology has gained so much traction online that it was only a matter of time before it came to the biggest of all ad channels, TV. And what do you know, Cox just announced that it will sell TV inventory programmatically.

This should be good news for TV advertisers, as it signals progress for an industry that is still stuck in the stone age. However, the truth is that simply overlaying current metrics and data within a more automated system doesn’t give advertisers a better result – it just makes it easier to make poor buying choices.   

TV requires much bigger investments than digital, and therefore the risk is much larger. To put it another way, successful and accountable TV advertising is impossible, even with programmatic technology, if we insist on using traditional metrics and data. Keeping the metrics and data used to buy and sell TV up to date is far more important in today’s market than relying on cutting-edge technology to transact the deal. 

This technology supposedly relies on “big data,” but much of what is passed off as “big” in the TV space could more accurately be called “medium data.” Advertisers are sold ineffective, modeled data, often derived from small sub-sections of the population (or worse, surveys).

For data in the TV ad world to qualify as “big,” it must be first-party data, but that term creates it’s own swirl of confusion. First-party data does not come from outside sources, such as set top boxes, data brokers, etc., no matter how they label it. True first-party data is the advertiser’s customer data, information that they own, amassed from  sales transactions, online traffic, call center data, loyalty clubs, abandoned shopping carts -- you name it.

Every advertiser is sitting on a heap of valuable data that can help them understand audience and consumer intent, and this data is the most effective way to execute programmatic TV ad buying and targeting.

Unfortunately, many marketers are still in the dark ages when it comes to understanding that data and turning it into insights. It is all encompassing, which is why they call it “big data.” Advertisers have all the information, but don’t have the resources or time to turn it into anything. They don’t even know what it’s supposed to look like when it comes out the other end.

Savvy marketers buying TV ad time should be looking for their product’s buyers,” and that’s what first-party data gives you. Online advertising’s downfall is that it’s still caught up in targeting “consumers, even when using programmatic buying and targeting. When it comes to TV, advertisers still only chase viewers,” which might as well be interchangeable with “impression” because it offers such little value.

It doesn’t matter if we’re talking TV, print, online, or out-of-home – demographics are poor surrogates for consumer targeting.

Advertisers already have the information about their actual customers. As TV finally moves toward a programmatic buying model, advertisers need their first-party customer data to identify product buyers who are very similar to their existing customer base, then find the programs that attract those likely buyers. Otherwise, the targeting is just TV done the same old way, with the same poor modeled data.

The programmatic TV renaissance (or possibly the reformation) won’t happen until everyone understands they need good data to realize the value of programmatic buying. Real buyers are the people advertisers want, and the only way to get there is first party data. Once an advertiser understands who those people are, the insights are applicable across the board. TV, direct mail, online – every channel feeds another, and it could change an agency’s entire infrastructure if you can deliver across every channel programmatically.

 

 

Recommend (3) Print RSS
1 comment about "You Can't Do Programmatic TV Without Good Data".
  1. James Nail from Forrester Research , March 19, 2014 at 12:19 p.m.
    Jon, I agree with and like where you are going with this. How do you feel about overlaying Polk, grocery loyalty data, etc.? This data is important for defining the next ring of prospects outside the first party current customers -- and TV is still the best medium for growing your market share. And because we don't have perfect data on 100% of the population at this level of granularity, what role does look-alike modeling play?