Forcing the impression to be the "catch-all" metric to compare media ad performance has been a favorite crutch of the media industry, but it does more harm than good. Comparing TV spots to web banners to radio ads to newspaper ads is worse than comparing apples to oranges to pears.
But making Nielsen, ComScore, Arbitron and the Audit Bureau of Circulation numbers all tell the same story is even worse. Just because each media can be counted in audience impressions doesn't mean that they are comparable, particularly in the highly measurable digital media world we live in today.
How did we get here? This passage from the "The ARF Guidelines for the Audience Measurement of Yellow Pages" (adopted in 2004) is indicative:
The common currency for major media in the United States is the advertising "impression" (the third level of the ARF Media Model.) It is difficult, if not impossible, to measure "impressions" in a way that will be fair to all media. Is "watching" television equivalent to "looking into or reading" a magazine? However, since this quantity, labeled "impression," is often the actual basis for comparing media proposals from media sellers, it is imperative at this time for the Yellow Pages industry to embrace this terminology as a necessary component in their sales and purchases of advertising.
The Yellow Pages industry trying to define itself by impressions aside, here are the reasons why I believe that using impressions as the common currency of media is more deceptive than it is helpful:
Media Are Different
It makes sense to use impression metrics to compare one 30-second TV ad spot to another. It makes no sense to use impressions when comparing a TV spot to a website banner impression to a magazine ad page. Probably 90% of what makes each of those media placements special is unique to media and how their viewers, users and readers consume them. TV is all about sight, sound and motion. It is about shared experiences; it is passive, leaning back. The web is all about leaning forward, interactivity and declaring intentions. Radio is often just in the background. Magazines are glossy, sensual, personal. Defining the economic comparability of those various media by the 10% that is the same, distorts their relative value and creates a false sense of equivalency.
Exposures Are Different
Cross-media impressions comparisons are bad enough, but what about using impressions to create comparability among very different types of media exposures? Impressions are not the best way to compare interstitials on NYTimes.com to a corporate logo on a pro golfer's visor promotion to Facebook buttons. It doesn't matter which exposure is better; -- they are all clearly very different. The NYTimes.com interstitial owns your attention while you wait for your news. The Facebook button is one of a dozen more ad-like images fighting for your attention while you check our friends' status and photos. The pro golfer visor logo is one of hundreds of latent logo "impressions" that you see every 15 minutes when watching a golf tournament on TV. Who can argue that each of their impressions are comparable?
No More Scarcity
Focusing on impressions as the common currency of media made some sense when all media impressions were scarce and as a rule folks were not consuming three or four different media all at once as they often do today. Decades ago it required printing presses, broadcast licenses or transmission towers to deliver media, and their owners -- newspaper companies, TV and radio broadcasters -- made sure to create a limited number of ad units that they charged a lot for. Today, the internet and digital networks have made media distribution plentiful and cheap, and impressions have become plentiful and cheap as well. When impressions were scarce as gold, they had currency value. Not now, when they've become as plentiful as paper.
Back-end Overtaking Impressions on the Web
Web advertising is probably the most impression-standardized media of all. The industry adopted impression-based measurement and cost-per-impression measurement as its primary selling and buying metrics from almost the very beginning. However, as everyone who has bought or sold web display ads know, back end metrics, like cost-per-click and cost-per-action and cost-per-conversion are the real drivers of the economics in the category. Cost-per-impression campaigns are almost always "backed into" from results-based measurements. As more media channels become "web-ified," impression metrics will continue to lose ground to back-end results.
Let's stop deceiving ourselves and others by forcing impression metrics on everything that moves in media. It barely made sense in the past. It certainly doesn't anymore. Is this the time to revisit "engagement" as universal media metric? I think so. It won't be easy to develop a perfect, cross-media definition for it; but whatever we end up with will be better than where we are with impressions.
|ABOUT THE AUTHOR|