Brand Images Can Be Stable

John Rossiter
John Rossiter
  • Speaker
Rotterdam School of Management (RSM), Erasmus University Rotterdam

Event Information

Research Seminar
Mon. 4 Jun. 2007
12:30-14:00 hours
Mandeville Building T3-42


Why is it that, within a brand image reported by an individual consumer, some brand-attribute associations are stably repeated when the consumer is reinterviewed, whereas other associations are not repeated and new associations are made?  Stochastic modelers C. Rungie, G. Laurent, F. Dall’Olmo Riley, D.G. Morrison, and T. Roy (2005) propose that individuals’ answers are “unreliable” but this merely describes instability rather than explaining it.  The explanation, we suggest, is that many associations are constructed “on the spot” by consumers faced with the task of completing the survey (J.A. Krosnick, S. Narayan, & W.R. Smith, 1996).  By means of an experiment embedded within a longitudinal survey, we show that consumers’ tendency to construct temporary associations is exacerbated by explicit instructions encouraging guessing, by fatigue induced by being asked to make many ratings, by being asked to rate brands they have never heard of, or by surveying consumers whose first language is not the same as in the questionnaire and who may therefore misinterpret or variably interpret attribute descriptions.  On the positive side, we identify disaggregated data in the original studies meta-analyzed by Rungie et al. (2005), as well as in our own data, which show that some brand associations remain stable among many individuals.  The building of targeted, lasting associations by advertising is also demonstrated in numerous tracking studies of advertising campaigns.  Generally, we conclude that the key assumption in brand management that marketers can induce stable brand images in consumers, an assumption questioned by Rungie et al.’s (2005) stochastic model and aggregate data, remains viable.  Specifically, we conclude that brand images will be stable if validly formed but not if they are made up on the spot by consumers trying to be helpful on the survey.
Contact information:
Dr. B. Donkers
Bas Donkers
Professor of Marketing Research
  • Coordinator