The Behavioral Economics of Consumer Brand Choice: Establishing a Methodology

Abstract

It is well established by marketing research that only comparatively few buyers of a product category (such as baked beans or breakfast cereals) are entirely loyal to a single brand (Heinz or Kellogg’s Frosties, for instance). Most buyers practice multi-brand purchasing over a period of, say, three months, selecting apparently randomly among a small subset (“repertoire”) of tried and tested brands (Ehrenberg, 1988). Data on buyer behavior for ready-to-eat breakfast cereals in the US, reported by consumer panel members are typical: “The average Shredded Wheat buyer in the year buys it about 4 times in that year, and buys other brands about 37 times. About 12 million US households [buy] Nabisco’s Shredded Wheat in the year. But it is not obvious whether these households are to be thought of as Shredded Wheat customers, or as other brands’ customers who/occasionally bought Shredded Wheat” (Ehrenberg & Goodhardt, 1977). Similar patterns are apparent for the vast majority of fast-moving consumer goods in steady-state markets (i.e., those with only a slight upward year-on-year trend in sales) in most affluent, consumer-oriented economies. Each brand attracts a relatively small proportion of the buyers of the product category who purchase that brand exclusively during the period under review: as that period lengthens, this proportion declines. Multi-brand purchasing is the norm to the extent that even the heaviest purchasers of a given brand buy other brands within the category much more than they buy their favorite brand over the course of say a year.

I would like to issue a challenge to current and future researchers to find new, creative ways to investigate, outside the laboratory, the quantitative models arising from the Harvard Pigeon Lab… The researchers — as opposed to the clinicians — who have followed in [Skinner’s] theoretical footsteps…have very often focused on tight, and then tighter, and then even tighter controls of variables in the laboratory. There is no question that such research is extremely important to our understanding of the principles of behavior. It should not be, however, the only type of research being conducted. There are a great many benefits to an outside-the-laboratory, empirical approach. (Logue, 2002)