What refinements to TV data will improve ROI modeling?
This research study by the CRE’s ROI committee examined how different approaches to four TV measures used in models affected granularity, precision and accuracy. Findings showed the impact of varying some factors and noted others that are being changed to address concerns.
The four measurement areas studied were:
- Precise minute – Using average quarter hours (AQH) GRPs vs. Precise Minute GRPs derived from specific ad minute data for each occurrence in the schedule
- Playback date – Assigning the date of original program airing vs. the playback date on which viewing occurred
- Actual cable network viewing – Estimated national cable network’s viewing in DMAs vs. actual local network viewing to national cable networks.
- More precise viewing – Employing diary AQH GRPs vs. exact minute GRPs from a combination of people meter and diary data.
Among the findings from the study:
- For precise minute, most of the time there is only a small difference between the data sets. However, the impact on ROI can be quite significant in the 10% of times it occurs.
- For precise day, the effect of misattributing playback across weeks is small, because playback on the same day and within the same week playback is very high
- For the remaining two areas, new approaches are expected to enhance modeling accuracy:
- Nielsen announced that actual cable viewing measurement is being changed as a part of the re-engineering of local television audience measurement. This major step forward will provide a passive measure of household tuning that is relatively consistent across DMA’s
- Nielsen said that it is re-engineering local market measures, so that broadcast and cable network data by DMA will be measured and reported the same way, using passive set-tuning data. This will address modelers’ concerns about the accuracy of diary measures and the potential bias introduced by three different measurement systems in different DMA strata.