May 26, 2017
I’m seeing the term “targeted linear TV” crop up more and more (as in this Business Insider article about a Credit Suisse report on TV advertising’s evolution) to describe what has been variously called “audience-based,” “index-based,” “advanced,” or “data-driven” buying. Or even, perish the thought, “programmatic” — a term I’ve already pleaded with the industry not to use.
I write for a living, so words matter to me. In fact, a big part of my job is translating the hyped-up words that tech companies use into language that explains what the product, service, or tool really does. This is important because using words that connote or imply that a product/service/tool does something that it doesn’t do only leads to confusion, disappointment, and distrust.
“Targeted” TV does exist; it’s also known as “addressable” TV, where different commercials are delivered to different households, set-top boxes, or devices based on data specific to that viewer, a trend I’ve written about for over three years. Cablevision, DirecTV, Dish Network, Hulu, Sky Television, et al. deliver this truly “targeted” TV today. But this has somewhat plateaued as other multichannel video programming distributors (MVPDs) hold back on the infrastructure upgrades needed and programmers shy away from collaborating with MVPDs to open their inventory to dynamic ad insertion.
But what the Credit Suisse report and the industry increasingly calls “targeted” linear is not this — and in fact is so different that using the term “targeted” is misleading.
Why do I object to the term “targeted” so strongly? Because in linear TV, buyers are still buying spots in pods in programs. As long as they do that, they will always get a lot of “nontargeted” viewers — probably the majority of the people who see the ad. What is “targeted” about that amount of waste?
What this type of “audience-based” or “data-driven” buying does do is allow the advertiser to be more selective about the dayparts, networks, and programs that their ads run in by applying the explosion of data available today through more powerful analytic platforms. As a result, a higher percentage of the viewers are in the advertiser’s strategic target group (e.g., minivan prospects) than if it just bought “women ages 25 to 54.” In the end, the advertiser reaches more of its best prospects, and that is a good thing.
But, quite frankly, this is conceptually identical to what media planners have always done by using data sources like MRI and Simmons to understand the audience composition of an ad placement. Granted, this type of buying does bring two significant changes: 1) the scale of the data used — from research survey panels numbering in the tens of thousands to data from tens of millions of set-top boxes and 2) programmers’ growing willingness to guarantee delivery against these new audience definitions, not just age/gender metrics.
Updating decades-old planning and buying processes to take advantage of bigger data sets is a great advance. Retiring Excel in favor of more sophisticated analytic platforms and cloud applications that can churn out a TV schedule in minutes rather than weeks is an even bigger advance. And the shift from thinking only in terms of age/gender to more tightly defining to whom the ad is truly relevant is a step toward a future where more and more TV ads are addressable.
But for now, please, don’t call it targeted!