For the moment, advises the Media Rating Council (MRC), served impressions may be an imperfect currency for buying and selling ads. But until viewable impression measurement improves, they'll have to do.
MRC has issued an advisory to members, pretty much telling them that viewable impressions measurement capability varies considerably by vendor, but stopped short of naming those vendors. Pointroll and Google are just two that offer viewable impressions measures. The advisory, penned by MRC CEO George W. Ivie, notes that "Several vendors – including ad servers, verification services and third-party syndicated measurers – have developed solutions for determining whether served ads become viewable and to count ads based on viewable parameters." But based on large-scale tests of real campaigns, measuring viewable impressions "is a challenge that must be overcome by most of the vendors" before finalizing a new online advertising currency definition. In short: they’re not getting it done.
Cross domain iFrames present a particular challenge to measuring viewability because they can obscure the capacity to determine whether or not an ad impression is viewable. iFrames have an important function in protecting content and protecting undisclosed data collection, but were behind 3/4 of unmeasured impressions in network placements in the pilot, and more than 1/3 of the unmeasured impressions in publisher placements. At present, MRC does not believe any vendor has overcome the challenge.
So while the capabilities improve, MRC recommends that buyers and sellers to seek validated viewability solutions (ones that have been audited and accredited/certified against applicable industry standards by an independent 3rd party) and ensure they clearly describe what can and cannot be measured.
MRC oversaw pilot tests of viewable impressions using 22 real ad campaigns running during May and June 2012. In total, twenty-two campaigns, representing more than three billion served impressions (the current industry standard used for selling and buying of digital display). The campaigns ran in a variety of environments and involved inventory purchased both directly from publishers as well as from network sales channels.
The results weren’t great (see below).
As a result, MRC believes it is "premature" to transact on viewability in advance of a proven measures, particularly in the case of cross-domain iFrames.
The MRC's position is aligned with the Making Measurement Make Sense (3MS) initiative's recommendation that "online advertising impressions, particularly for brand advertising campaigns, should be viewable in order to be more comparable to other media which afford the 'Opportunity to See' as the basis of currency." (3MS is joint initiative by the Interactive Advertisers Bureau, 4As and Association of National Advertisers.)
- The test campaigns showed a range of 0% to 77% of served impressions that were measured for viewability.
- Viewable rates for the live campaigns ranged from a high of 78.6% to a low of 7.8%. Viewable rates are calculated off of a base of impressions that could be measured for viewability.
- Overall measured impression rates across all pilot campaigns were 27% for publisher placements, and 19% across ad network placements.
- Viewable rates varied by ad sizes within placement types (network/publisher). Rates ranged from 12.6% to 63.7% for the most popular ad sizes deployed by the advertisers in the pilot test. Most popular was determined by the number of impressions served.