Don't Blame Targeting

There is an industry conversation about targeting that’s juicy but laden with double-talk. That conversation starts with digital companies offering amazing targeting capabilities. They have remarketing, lifestyles, character traits, interests, intentions, psychographic, demographic, or … whatever you want. 

If they don’t have it, they can make it. For many situations, data is blended. It’s like a witch’s brew: five parts Auto-Intender blended with Hispanic, and the hair of a data scientist. Oh, and then there’s AI.  

Cynics in the media industry still say, “Where there’s mystery, there’s money.” Data, the key ingredient in modern targeting, is a case in point, but it does not have to be as mysterious as it seems. Targeting is cake, not frosting. 

There are hundreds of thousands of third-party segments available for sale. When you make them all usable across media with a device graph, messages can be thrust into virtually any receptive moment. 



Propping up the other side of the conversation are brands. Some have given the stink-eye, loudly and in public, to “over-targeting.” The narrative seems to be that they tried targeting and it didn’t work. Some say they “went too narrow.” That reflects a choice about using targeting, not about whether targeting is either good or bad. 

Anyway, this sounds like a reach problem. Specifically, you should not blame “targeting” for a poor targeting choice. That’s a little like blaming “driving” for a car wreck. 

All media are targeted one way or another. You could argue that some campaigns are “filtered” for the wrong people rather than targeted to the right people, but that’s just semantics. Basically, all media products have an audience with claimed characteristics, and methods for subsetting that audience according to extant targeting methods.  

Blame the Right Thing

So, if we want to get the blame for bad targeting right, we should at least acknowledge the two likely points of failure.  One is the target audience definition. The other is the audience delivery.

Audience definition is normally derived from well-established research about a brand.  Audience delivery, on the other hand, is normally not measured except as gross reach, “gross” being the operative concept. 

So, delivery is the prime suspect, but delivery itself has two elements: quantity and quality: Reach is quantity, easy to count, and often currency.  Quality is harder to define but matters as much as reach because if the impression is ineffective, it’s as though the reach never happened. 

Absent a perfect definition, we should agree that quality is simply the extent to which the intended audience was delivered. Quality is the hard part. 

Media companies have a nice solution to insufficient reach to a target: Buy more reach!  Yes, media can usually overcome a quality problem with quantity — and, by the way, make a lot more money in the process. 

The Time Dimension

The third side of the targeting Rubik’s Cube is the timedimension. Let’s take a quick look at that. 

There is always an optimal “when.” You are always targeting into a time window, whether you intended to or not. 

When we plan for a receptive moment, sometimes it’s having to do with adjacency to a purchase opportunity, sometimes with intention or interest, and sometimes with state of mind. Those are time-dependent per individual, as you might expect. It’s easy to miss on the time dimension, but it’s still a targeting requirement.    

If you don’t get a sale, you’ll probably get branding effects. These effects are difficult to tie back to an impression, but branding does not require much defense.  Branding has worked well for a hundred years, for thousands of brands. But even branding messages are wasted on a person unlikely to need the category. 

For long-purchase-cycle items (dishwasher, car), if you send an impression to someone not in the market, there is no soft edge. It’s a swing and a miss. Branding effects probably don’t last longer than your new dishwasher. 

For faster moving categories with high penetration (shampoo), everything’s a bunt. Few impressions miss a potential “who” (since most people buy the category), but almost all of them miss the optimal “when” (replenishment, near point of purchase). 

So, with CPG, the opportunity cost of an off-target impression is low because most people buy CPG categories. In six weeks, you’ll get another bite at the apple, and the branding effect from today’s impression could still be in play. 

Now, a midsummer moment of Zen: Sometimes we swing and miss, and sometimes we should have taken a swing, but didn’t. In the end, though, we win by making more good decisions than bad ones. The winners know the difference. 

1 comment about "Don't Blame Targeting".
Check to receive email when comments are posted.
  1. Gerard Broussard from Pre-Meditated Media, LLC, August 29, 2019 at 6:26 p.m.

    Nicely said, Ted!  The quality and accuracy of consumer target segments is a Pandora's box of issues that needs opening but marketers, media agencies and media companies (and let's not forget the data providers) might be too perplexed and frightened at what they would find.  Amazing that not more industry focus in concentrated here.  But I'm feeling that will change pretty soon, but will need the backing of advertisers.  Gerard  

Next story loading loading..