Yet engagement has not been defined in any usable way except as a higher consumer score in IAB's Internet-based TV measurement. Engagement has not replaced reach/frequency, as was originally promised by the ARF. Nor has engagement significantly improved our planning models. How did so many smart people get led so far astray?

We were ready for a big idea. There was mounting evidence that advertising, especially TV advertising, was not working as well as it once did. The easy explanation was "consumers are not engaged." Then ARF offered its landmark definition and promised to make it right. Engagement, ARF explained, is simply "Turning on a prospect to a brand idea, enhanced by the surrounding context." If we can learn to do that, we'll be fine.

But those comforting words had not a hint of what or how to measure. For example, how do we quantify "turning on?" as in "turning on a prospect?" Do we use a blushometer?

What is "a brand idea?" Brands don't usually have ideas. People do. Do we simply mean the advertising's message? What is "surrounding context?" Is that good old "media environment?" And if it is media environment, how do we measure whether it is enhancing? Or do we assume media environment is enhancing if it helps turn on the prospect to a brand idea. As measured by a blushometer. Sorry. That's not research. It's Abbott and Costello.

Elevating engagement from measurement to "belief" is also part of the problem. Last year's invitation to the ARF's 53rd Annual Conference and Expo seemed to reach out to the afflicted with these words: "Many in our industry have become Engagement "believers," others are still holding out." Belief is a strange word for research. It is an act of faith not a proof of concept. Is the message "come and be saved?" Are those who doubt to be shunned as non-believers? In a way, that's what has happened.

"A lack of engagement" is a convoluted interpretation of what is really happening. The problem is dis-engagement. Advertising is losing the consumer connection it once could promise. The way to solve that is to define what we mean by "consumers are not engaged," determine under what circumstances they are not engaged, and seek a solution. The job in not to add engagement. It's to reduce dis-engagement. But that means we have to acknowledge that too much indiscriminate advertising is a big part of the problem.

Our traditional TV advertising model: programs attract audience which is delivered to the advertising. Here there is media engagement, which is people attracted by media as measured by things like audience size, targeting, duration, loyalty, all obtainable from audience data. Hidden here are measures of disengagement in tune-out.

And there is advertising engagement, people attracted by the advertising as measured by things like brand recall, likeability, intent to purchase, all from creative testing. These are two different kinds of engagement. And we have many useful measures for each. But engagement with a program need not spill over to the advertising it carries. Most evidence is that the frequent and long interruptions of programs breaks the spell.

The companion idea of "context" in the sense of program content that enhances or "reinforces" a product's advertising is just as unlikely. TV seldom offers contextual opportunities, if only because of the range of products advertised in a program and the near certainty that your commercial will not follow the program itself. (You can calculate that probability. Today it's less than 10%.)

We need to focus less on describing engagement and think more about how to identify and fix what is preventing it. Or am I preaching to the choir?

This commentary is insightful. I recommend it to others.