Santa Clara University

Offering the Best Selection

Consumer-Preference Data Can Help Retailers Decide

Every year in June an established retailer holds a women’s sweater promotion as a marketing test. The styles and colors under consideration for the fall line are laid out, and those that sell the best are featured in the store’s assortment a few months later as cold weather draws near.

It’s a classic example of the sort of hunch-and-observation technique that retailers have used over the years to try to determine what will sell the best. But it’s far from perfect, and it illustrates a business issue that two professors in the Marketing Department at SCU’s Leavey School of Business are investigating.

“The basic problem is that retailers have always had a challenge figuring out what products to carry in their store assortments,” says Dale Achabal, L.J. Skaggs Distinguished Professor. “They know what sells, but they don’t know what they could have sold if they’d offered consumers a better-chosen assortment of products.”

Achabal, who is also executive director of Leavey’s Retail Management Institute, is co-author of a paper that has pioneered the use of information drawn from an interactive web site to analyze consumer preferences more scientifically. It is called “Choosing Robust Retail Assortments for Infrequently Purchased Products,” and was co-authored by fellow marketing professor Shelby H. McIntyre, their Operations and MIS Department colleague Professor Stephen Smith, and Christopher H. Miller, former assistant professor of marketing at SCU, and now at Bond University, Queensland, Australia.

To narrow the focus of their research, they concentrated on DVD players, which represent a product for which consumers make a “considered” decision. Achabal and his co-authors worked with the principals of Active Decisions, developers of, in which over 2,000 visitors completed an interactive survey to get specific product choice recommendations from a choice of 117 DVD players.

Tracking the information in the surveys, they found that consumers broke out into a variety of groups, depending on what they were looking for in the product. For example, some might want a specific brand at a certain price; others might want the best quality regardless of cost; and others might be adamant about having certain key features in the product. Knowing this could enable retailers to determine which groups their customers were likely to fall into and help companies provide an assortment of goods that would satisfy the greatest possible number of customers in their target market.

“When you get down to it, any product is a bundle of attributes,” Achabal notes. “This sort of consumer preference data can give retailers an idea of what an alternative assortment of products could do for them.”

Achabal suggests that based on this understanding of consumer preferences, one of the most common mistakes a retailer might make would be to replace the product that has the lowest volume of sales.

“The product with the fewest sales could still be satisfying a certain segment of consumers,” he argues. “If you take that off the shelf and replace it with something that duplicates what you already have, you could just be cannibalizing the sales of your existing products.” In addition, as McIntyre points out, a reduced level of choice is a fundamental consumer issue. “If a store doesn’t have what I want, they’re not giving good customer service,” he said.

The paper develops an analytical approach and decision support system for tracking consumer preferences. Achabal said he sees this work as “adding science to the art of retailing” in a way that could be quite positive.

“If a retailer builds better assortments of products in the store, revenues should increase and the customers should be more loyal. It’s a win-win situation.”

Printer-friendly format