Before I joined comScore, I operated my own media research consultancy, Warp Speed Marketing. (Perhaps you’ve heard of us. By which I mean, me.) One of my projects as a consultant was working with Arbitron Outdoor, circa 2002-2004, on the development of a ratings service for billboards and other Out-of-Home inventory; we rolled out a test service in Atlanta in 2003, measuring and reporting on, if memory serves, over 7,000 distinct pieces of inventory.
On that assignment, I got to learn a lot about the outdoor advertising business. One thing I learned was that the then-prevailing audience metric was the Daily Effective Circulation, or DEC. Now, see if any of this sounds familiar… the DEC was based on empirical site-centric traffic counts, but did not provide information at the persons level, and did not disentangle reach from frequency, and advertisers and agencies believed that these counts dramatically overstated actual audience delivery. I remember meeting with the leading out-of-home shops, and having media directors tell us that they routinely discounted the DEC metric by 50%. This lack of trust in the raw traffic metric was a limiter on advertiser confidence in the medium.
Long story short, the industry’s measurement arm, the Traffic Audit Bureau (TAB), engineered and commissioned a new measurement service, called EYES ON, which integrates multiple factors including vehicle traffic but also Viewability factors for each piece of inventory, to provide buyers and sellers with counts of persons who actually saw -- laid eyes on -- the ad. The counts produced via the new EYES ON measurement are lower than the old DEC numbers. And what impact has this had on the business? Well, in a soft advertising economy, Outdoor is up 4.5% in the second quarter of 2011.
In a lot of ways and largely under the radar, in the U.S. the outdoor industry actually pioneered concepts that are now accepted as state of the art in digital measurement. One of these is the the integration of panel data with census traffic counts (and in outdoor, when they say, traffic… they really mean, traffic!)
In June, the IAB’s Making Measurement Make Sense initiative (3MS) declared 5 Guiding Principles for Digital measurement. The first of these is an important one, and worth pondering; “Move to a “viewable impressions” standard and count real exposures online.” About this, the MMMS principles document elaborates thusly: “Today we count ‘served impressions’ as recorded by ad servers. Often, ad units are not in a viewable space to the end-user or fail to fully load on the screen – potentially resulting in substantial over-counting of impressions. Viewable exposures are increasingly the norm across other media and better address the needs of brand marketers.”
Indeed, the trend across media measurement -- from TV’s average commercial minute rating to out-of-home’s EYES ON -- is toward counting the actual ad or campaign audience. I sometimes worry that in digital advertising, our business has moved so quickly in the direction of efficient delivery of impressions to cookies, that we’ve sort of forgotten that the objective of the game is actually to put ads in front of consumers (actually, in front of engaged consumers.) And frankly, there are far too many opportunities in our space for bad actors to knowingly jack up traffic and impression counts, without actually putting more ads in front of more consumers. (Here I’d recommend my colleague Brian Pugh’s blog post on fraudulent traffic.)
If you look at the five 3MS guiding principles, each of them may be seen as providing clarification and insight that helps demystify the digital morass and make it more hospitable to brand advertisers. Aligning around the viewable impression is an important first step in this demystification. When advertisers buy a spot on a network prime-time show, they have a pretty good idea what they are going to get. But spend the same money online -- through networks and exchanges, letting cookies decide who gets your ad and when and where -- and, let’s face it, there is far less certainty.
The consultants from Bain who did the industry interviewing for 3MS did a great job in collecting input from all sectors of the industry, and assimilating that input into meaningful findings. (I had the good fortune to meet with them more than once and was duly impressed each time.) Clearly, viewability—or perhaps more broadly, ad verification-- is an important advertiser hot button, and we as an industry need to respond.
comScore acquired AdXpose over the summer, so we’ve learned more about campaign performance with respect to viewability. Kirby Winfield of AdXpose tells me that in a given campaign, as many as 50% of the impressions served might not be viewable. That’s not an average, and performance will vary depending on the mix of vehicle types included. But it underscores the legitimacy of the advertiser’s concern.
Does this mean we’re off target with all the algorithmic cookie targeting? Not necessarily. Presumably we are collectively clever enough to figure out how to incorporate viewability measurement even into exchange inventory, with a premium price for validated viewable units. And if that devalues ad inventory that generally produces low visibility? Well, my experience is that more information about inventory quality tends to increase demand, and thus advertising revenue overall, even if it devalues some specific inventory units.
But if we want to grow the pie and draw greater shares of advertiser budget into digital, maybe we do need to take a step back from our fixation on efficient delivery of impressions to cookies in real time, and focus more on the actual thing advertisers want: to deliver messages to consumers. Viewability is an important component in securing long-term advertiser confidence, and we need to bake it into our measurement. Just ask our friends in out-of-home.