Press Release: Media Rating Council Announces Summary of Results of Phase 3 of Viewable Impression Reconciliation Testing

Testing Aimed to Identify and Reduce Discrepancies among Vendor Measurements

 

NEW YORK – July 8, 2015 – The Media Rating Council (MRC) recently completed its third phase of reconciliation testing aimed at aligning the results of MRC accredited measurement vendors’ viewable impression-related measurements. The organization conducted two prior phases of testing, which identified additional requirements for vendors to follow in order to minimize discrepancies in viewability measurement among the various vendors. This third and final stage of testing was conducted to gain additional assurance that all material causes of discrepancies have been identified and accounted for in the MRC’s requirements around viewability measurement.

In response to a call for additional data that directly compared the results from two or more viewability measurers for specific digital advertising campaigns, MRC received data from campaigns encompassing nearly four billion served ad impressions, covering a wide range of site, placement, and creative types (including both display ads and video ads).

An in-depth review of the reconciliation study results can be found here. Key findings include:

  • For nearly two-thirds (63%) of the campaigns (among those consisting of more than 100,000 viewable impressions measured), the differences observed between vendors’ viewable impression counts were within what was deemed an acceptable range (i.e., less than a 10% difference). The average (weighted) of the differences observed for these campaigns was 4.1%.
  • For the remaining campaigns (37%) of more than 100,000 viewable impressions, differences of more than 10% between/among the vendors’ viewable impression counts were observed. In these instances, the median difference noted between vendors’ viewable impression counts was 23%, and the average (weighted) difference was 34% per campaign. This was a greater level of differences than those noted in the two earlier phases of reconciliation testing.
  • Smaller campaigns of less than 100,000 viewable impressions tended to show greater variability in comparative results.
  • Closer analysis of the differences observed in all situations (i.e., test campaigns of all sizes) led us to determine the following reasons for the differences in reported viewable impression counts:
    • 54% of the difference was a result of differing treatments of mobile viewable impressions in vendors’ reporting.
      • MRC Observation: Discrepancies attributable to this reason have grown since our prior reconciliation phases. The MRC Viewable Impression Guidelines were written as a guide for measuring viewability in desktop environments. Viewable counts resulting from impressions served to mobile environments should be segregated in reporting from those served to desktop. Given the growth noted in this discrepancy area in this latest round of testing, we believe this segregation has a very high urgency.
    • 28% of the difference resulted from vendors treating multi-ad units in different fashions for measurement and reporting purposes.
      • MRC Observation: MRC addressed this issue in an earlier round of viewability reconciliation, and informed vendors that ads should be measured for viewability individually, regardless of whether they are packaged together as part of a multiple ad unit. This issue will be formally addressed in a forthcoming update to the Viewability Guidelines.
    • 13% of the difference was a result of differences in whether vendors opt to measure ad traffic served in a campaign by ad servers other than themselves.
      • MRC Observation: This was a new finding resulting from this round of reconciliation testing, and will be addressed in a forthcoming update to the Viewability Guidelines.
    • 2% of the difference was a result of differences noted in the application of certain ad verification processes, and the reporting of viewable impressions within those processes.
      • MRC Observation: This issue was addressed in the original Viewability Guidelines, and was reinforced by MRC to viewability measurers following the last round of reconciliation testing. It will be emphasized again in the forthcoming update to the Viewability Guidelines.
    • The remaining 3% of the differences noted resulted from other causes (including issues that were previously identified in earlier MRC reconciliation testing).

Based on the findings from this latest phase of reconciliation testing, MRC plans to update the MRC Viewable Impression Measurement Guidelines to address any specific areas identified through the reconciliation process that have not been already addressed, and to emphasize those issues that have been previously addressed but, remain sources of discrepancies in viewable impression measurement. These updates to the Guidelines will be issued by MRC within the next 30 days. Once issued, accredited measurers will have a limited amount of time to fully adopt these updates (to the extent they have not already been implemented), and compliance with these provisions will be verified through the MRC audit and accreditation process.

For more information please visit www.measurementnow.net or www.mediaratingcouncil.org.

 

About the Media Rating Council (MRC)

The MRC is a non-profit industry association established in 1963 composed of leading television, radio, print and Internet companies, as well as advertisers, advertising agencies and trade associations whose goal is to ensure measurement services that are valid, reliable and effective.  Measurement services desiring MRC Accreditation are required to disclose to their customers all methodological aspects of their service; comply with the MRC Minimum Standards for Media Rating Research and other standards MRC produces; and submit to MRC-designed audits to authenticate and illuminate their procedures.  In addition, the MRC membership actively pursues research issues they consider priorities in an effort to improve the quality of research in the marketplace.  Currently approximately 100 research products are audited by the MRC.