What’s Still Missing From the Measurement Conversation

Last week, I attended IAB’s Annual Leadership Meeting in New York, with “The Great Measurement Debate” session at the top of my list to see. The discussion, while lively, mostly centered on the pluses and minuses of sample-based measurement versus census-like measurement—a well-worn topic. This conversation, however, was missing a key element that must be considered: the difference between content measurement and ad measurement.

Historically, TV programming was measured with audience estimates or ratings, which provided a proxy for TV ad exposure; for years, these two things were one and the same. Today, now that second-by-second viewing data insights are available at scale, the industry should be moving (faster) to census-like reporting of household-level TV ad exposure. Anything short of that is not giving advertisers the reportable record/accounting of their TV investments.  

Content also needs to be measured, especially in the current, fragmented marketplace. Media companies that develop and distribute content want to understand who is watching (and where, when, etc.) – that is different than measuring who saw an ad.

The discussion of sample-based versus census-like measurement should be finished. TV is a data-rich industry now, and we should use that data to the fullest.  Of course, whether we need a sample for calibration or to translate household-level impressions to persons-level impressions is still to be determined.