The Perils of Survey Data in Cross-Platform Measurement
Cross-platform measurement is one of the most important topics on marketers’ agenda. Among the key objectives is gaining an understanding of how consumers spread their time and spending across devices so that media spending can be optimized. Some researchers have attempted to accomplish this measurement by conducting online surveys of consumers, asking them to report their own behavior. Unfortunately, this approach has now been shown to provide results that are materially biased toward the Internet, largely because consumers who agree to join online survey panels are two-to-three times heavier users of the Internet than the average person. This error is in addition to the recall error that experienced researchers know exists when consumers attempt to recall their own past behavior.
The following article identifies the sources and magnitude of these survey-based errors and shows how they can be eliminated through the use of behaviorally-measured panels. The article was recently published in the June 2014 issue of the Journal of Advertising Research and is reprinted with permission.