NEW YORK--(BUSINESS WIRE)--The Council for Research Excellence (CRE) today presented findings from a new white paper, "Better TV Data for Modelers: A Research Study for the Council for Research Excellence," at the Advertising Research Foundation's (ARF) Audience Measurement Conference in Jersey City, NJ.
The white paper was written by Sequent Partners under the auspices of the CRE's ROI Committee, chaired by David Poltrack, Chief Research Officer, CBS Corp. Presenting the paper at the ARF event were Jim Spaeth and Alice K. Sylvester, Partners at Sequent Partners; Paul Donato, former Chief Research Director, Nielsen; and Richard Zackon, Facilitator, CRE.
The motivation for the paper was the demonstration over the last few years that overly aggregated, or estimated, media delivery data doesn’t line up well with weekly brand sales data, often resulting in under-estimates of ROI. The concern was that this common practice may not show television in a fair light. The paper explores whether the television data used in marketing mix models can be improved.
"Marketing mix models remain the primary method for advertisers to estimate marketing ROI for all elements of their marketing plans, including television," Poltrack said. "Yet the media data inputs for marketing mix models are frequently drawn from planning or buying systems that often are lacking in the models' requirements for granularity, precision and accuracy. Modelers need more disaggregated data, variability and exposures properly aligned by week in order to tease out the impact of different media on sales."
Nielsen's AdIntel service is the primary source of TV GRP (gross rating points) data for modelers. The white paper examines four potential opportunities to improve AdIntel data:
- Using precise minute ratings for each ad occurrence, rather than C3 or average quarter hour.
- Using precise day of viewing for playback, rather than day of original airing.
- Using actual cable network viewing by DMA, rather than estimates based on the distribution of cable subscribers by DMA.
- Using more precise viewing estimates in diary markets.
The study compared AdIntel TV GRP data to other Nielsen-sourced TV GRP data that would be more granular, precise and accurate. The objective was to determine whether there are significant differences, and if those differences would matter in a model's estimate of TV ROI.
Key findings included:
- Precise minute: This is an infrequent problem, but quite substantial the 10% of time it occurs. The average under-estimate of television sales lift is about -20 percent The study recommends that Nielsen use precise minute data in AdIntel.
- Precise day: Playback viewing is not a frequent enough phenomenon to require a change in AdIntel.
- More precise cable audience: Pending changes to AdIntel will solve this issue.
- More precise viewing: The retirement of the diary method and consolidation of television viewing around Local People Meter (LPM) and Return Path Data (RPD) and the inclusion of these data in AdIntel will substantially solve this issue.
The paper was the third in a series of ROI and marketing mix model studies conducted by CRE. The previous two studies were conducted in 2013 and 2015.
About the Council for Research Excellence
The Council for Research Excellence (CRE) is an independent research group created in 2005 and funded by Nielsen. The CRE is dedicated to advancing the knowledge and practice of audience measurement methodology and comprises senior-level industry researchers.
CRE members are ABC Owned Stations, Beasley Media Group, CBS, clypd, Comcast, Cox Media Group, CoxReps, ESPN, ITN Networks, Katz Media Group, the Media Rating Council, National Association of Broadcasters, NBC Universal, Nielsen, Omnicom Media Group, Radio Advertising Bureau, Raycom Media, Scripps Networks Interactive, Television Bureau of Advertising, Tribune Co., Turner Broadcasting, 20th Television, Twitter, Univision, Viacom Entertainment Group and Weigel Broadcasting.
For more information about the Council for Research Excellence, please visit: http://www.researchexcellence.com/