About AMA

Executive Summary
Consumers tend to overpredict how much variety they will want in the future. Specifically, consumers choosing now (for later consumption) tend to select a wider variety of items than if they had made those choices later, at the time of consumption, a phenomenon referred to as the “diversification effect.” This effect has been widely construed as one of choice. The authors examine whether choosing multiple items for later consumption—a consumer decision ubiquitous in retail environments—affects the number of available alternatives considered before choosing. In other words, the (unobservable) subset of alternatives that a consumer considers immediately before choosing (the “choice set”) differs when choosing now versus choosing later. Prior research has examined observed choice alone, precluding the possibility that diversification is driven primarily by choice set formation, wherein consumers alter the breadth of choice set options.

Using a controlled choice experiment and a hybrid econometric framework that accounts for both choice set formation and choice-conditional-on-set-formation, the findings demonstrate that set formation plays a critical role in diversification: Consumers discount previously chosen options, but only for choices made for future consumption and only in set formation. Furthermore, the expected number of choice set items is substantially greater in multiple-item, versus single-item, choice. Specifically, when choosing all at once for future consumption, choice set sizes appear relatively larger overall, but the previously chosen item is less likely to be in a consumer’s (latent) choice set.

This research has important implications for brand managers, especially in product categories prone to diversification, such as food and beverages, travel and tourism, and entertainment. Small-share and niche brands tend to be the beneficiaries of diversification, so their brand managers should make efforts to design tactics that encourage multi-item choice. The reverse is true for large-share brands, which are more likely to lose share when consumers diversify. In particular, small-share brands will benefit more than large-share brands from tactics that increase their products’ likelihood of entering consumers’ (latent) choice set, such as investing in point-of-purchase promotions, especially for items such as groceries that tend to be purchased for future consumption.

In addition, managers of smaller-share or niche brands should consider how “change of pace” persuasive messages or positioning may influence consumers. For example, it is commonly assumed that consumers will trade off a less satisfactory attribute, such as (for a variety-seeker) recent purchase, for a good price. However, if variety seeking operates nearly entirely in choice set formation, this assumption may be incorrect: No matter how good the deal offered for a product, recent consumers may not notice or care, because they have already ruled that product out. In principle, it should be possible to test for such promotion-dampening effects, given sufficient field data on advertising, promotions, and longitudinal purchase histories.

The findings have implications for marketing research practitioners as well. Variety seeking has been incorporated into models of consumer choice; however, most modeling efforts assume that consumers consider all available items. This research suggests that is not the case and points to the need for marketing researchers to incorporate choice set formation into their choice analyses, especially when examining multi-item choice and variety-seeking behavior.

Biography
Linda Court Salisbury is Assistant Professor of Marketing at Boston College’s Carroll School of Management. Her primary research interest is in temporal aspects of consumer decision making and their effects on consumer expectations, preference, and choice. Professor Salisbury has examined phenomena such as choice diversification, preference uncertainty, customer expectations, and debt repayment decisions. Her research has been published in Journal of Consumer Research, Journal of Marketing Research, and Marketing Science. She received her PhD from the University of Michigan’s Ross School of Business and her MS and MBA degrees from Rensselaer Polytechnic Institute. Before her academic career, she worked in management consulting and the consumer packaged goods industry.

Fred M. Feinberg is D. Maynard Phelps Professor of Marketing at the Ross School of Business, University of Michigan. He previously taught at the University of Toronto and Duke University, having completed his PhD in the Sloan School of Management at the Massachusetts Institute of Technology. His prior research has addressed control theoretic issues in advertising response, statistical models of individual-level choice processes, and the interface between marketing and other disciplines, primarily engineering and psychology. He is associate editor at Journal of Marketing Research and Marketing Science, senior editor for Marketing at Production and Operations Management Society journal, and the coauthor (with Tom Kinnear and Jim Taylor) of Modern Marketing Research: Concepts, Methods, and Cases.