• by Jim Meskauskas , April 17, 2006

It's a word heard regularly in the halls of media and ad agencies. It's the topic of conversation at industry gatherings. It's the subject of industry conference sessions. It's even made its way into the titles of senior marketing and agency personnel. And it's the focus of this column: engagement.

Engagement is the mot de la saison in the advertising industry. It's often referred to with the kind of enthusiasm that once accompanied the phrases "behavioral targeting" or "one-to-one marketing."

But in spite of frequent name-dropping, the meaning of engagement remains elusive. Like "love" or "branding," we seem to understand what it implies without being able to adequately define it.

The reason for this is that we aren't really sure what it should be. Is engagement a proxy for time spent with a vehicle or a medium? Is it supposed to be a measurement for effectiveness, replacing frequency in the reach/frequency equation? Is it a brand new measurement to quantify the level of involvement with the media one consumes?

The American Association of Advertising Agencies and the Advertising Research Foundation have been working together on defining engagement as a metric, with cooperation from the Interactive Advertising Bureau, among other stakeholders. The plan is that engagement will replace frequency as a metric.

At their annual conference on March 21, Advertising Research Foundation's Chief Research Officer Joe Plummer announced a definition for engagement as it's used by media and marketers: "Engagement is turning on a prospect to a brand idea enhanced by the surrounding context."

Though I'm happy to learn that there is movement toward determining just what is meant by engagement as a metric, right now it's like stirring stale coffee: There is motion, but it's not improving the flavor.

According to David Smith, president of Mediasmith, a San Francisco-based integrated media agency, the problem with engagement as a metric is that it's much like the Apollo project. "If all it does is rationalize TV and media we already have metrics for, it will not be really useful," Smith says. "About 40 percent of [all] media is not measured."

Is engagement going to simply be another way to get more granular about how television media is consumed? Will it be able to indicate effectiveness, or merely articulate an attention level that serves as an extrapolative surrogate for the potential influence of an ad message on an audience?

If engagement is going to be useful as a media metric, it needs to be applicable to all media, not just television. I ride the buses and subways in Manhattan every day, and I see teenagers listening to music on their cell phones. I see people on flights watching TV shows on their iPods. I have friends who check sports scores from their Treos while sitting in a pub having a pint. As Smith pointed out in an e-mail to me recently, "At the 4As, we talked about 14 new media or media extensions of existing media, all of which need metrics behind them."

There seems little doubt that, depending on what its definition ends up being, engagement as a metric is going to be more meaningful than merely total media weight or the average frequency of advertising exposure. But the challenge of its application is two-fold.

First, if it can't be universally applied, its usefulness is limited only to those media to which it's applied. If it's not serviceable as a normative means of valuation, then we haven't made any advances in articulating or predicting the marketing effectiveness of a given media plan. Consumers' media consumption is varied and lends itself to the cumulative effects of advertising communications. In order to be meaningful, engagement has to account for this.

Secondly, just how does one quantify a qualitative experience? Time spent isn't conclusive. It could mean any number of things, from couch-borne inertia due to a life of passive ennui that keeps the channel unchanged, to difficulty navigating a Web site. If it's merely attention levels, where is that data coming from? It would seem to me that the kinds of data necessary to support an engagement metric could only be drawn from extensive research and observation; it's not something that emerges from a couple of thousand recall-based surveys.

Conceptually, engagement is an improvement in the way we can indicate media's effectiveness. But unless it is categorical in its application and quantifiable in its indication, we're looking at a long engagement before the wedding.

Jim Meskauskas is vice president/director of online media at Omnicom Group's Icon International. (jmeskauskas@icon-intl.com)