Editor's note: Steve Cohen is partner and co-founder of in4mation insights. He can be reached at firstname.lastname@example.org.
When developing new products or services, brands often obsess about the attributes they hope will influence consumer behavior, often using choice models to identify characteristics that are most persuasive.
Big surprise? Only some of the attributes you bake into your products command enough consumer attention in the moment of purchase to influence choice in meaningful ways.
Discrete choice models work best to identify the attributes most likely to influence consumer choice when they accommodate attribute attention as a component of choice criteria. In this article, we share what we believe is the best way to address attribute attention, using an improved approach to choice modeling.
Classic choice models assume that people maximize their utility by processing all attributes when making their choices. For various reasons, people may not process all attributes but rather select important ones and/or ignore unimportant ones.
Choice-based conjoint analysis (CBCA) has been a bedrock of marketing research for many years. A basic assumption is that by displaying a full set of product attributes to the consumer, she will use all information in her decision-making. A challenge is the interpretation of a small importance coefficient for an attribute. Is it small because the attribute was processed but she doesn’t care about it? Or was the attribute ignored and not processed at all? Choice models assume that people use a trade-off decision process that takes into account all the information available.
But is that true? Are all attributes used? Does everyone use the same ones? Are predictions from CBCA the same when we assume everyone pays attention to all attributes versus every person only pays attention to attributes of salience to them personally?
If you could estima...