Understanding Competition When Brands Belong to Multiple Categories
Imagine deciding where to eat in a shopping mall food court. Even for such a simple choice, there is a lot of information for consumers to juggle. Among many attributes, they can think about the type of foods offered, the service quality, the cost of the meal, or the healthiness of the options. Research has shown that consumers simplify their choices by automatically making inferences, and ultimately decisions, based on the categories they perceive restaurants belong to. The issue is complicated because different consumers can use different category structures to organize restaurants and because some restaurants may belong to multiple categories simultaneously.
Consider Chipotle; it can be characterized as Mexican, fast-food, fast casual, and its meals are consumed for both lunch and dinner. Chipotle, and other multi-category restaurants, must understand these structures to best market themselves. This begs the question: How can marketers identify consumer category structures when some brands are seen as belonging to multiple categories?
Blanchard and colleagues have developed a quantitative technique to help managers empirically identify the latent categories consumers hold. Unlike traditional clustering analysis, the technique allows each consumer to see a single object (e.g. a restaurant) as representative of multiple categories. The model also depicts how representative a restaurant is of a latent (unobserved) category. Together, the latent category data and representativeness data can be used to evaluate market competition.
To illustrate the capabilities of the technique, Blanchard and colleagues applied their model to consumer data for 25 restaurants in the State College Pennsylvania area (home of Penn State University). The model identified 9 latent categories that consumers were using to keep the restaurants organized in their minds. Some of these were pretty standard categories, such as burger restaurants (e.g. McDonalds, Wendy’s, Back Yard Burgers), Mexican restaurants (e.g. Mad Mex, Chipotle, Taco Bell), delivery pizza chains (e.g. Dominos, Papa John’s), and casual sit-down restaurants (e.g. Applebee’s, Chili’s, Olive Garden, Red Lobster, Ruby Tuesdays, Texas Roadhouse, TGI Fridays). However, other categories were less traditional (see the figure), like the sandwiches and subs category that included many hand-held options (e.g. Chipotle, Mad Mex, Panera Bread, Pita Pit, Qdoba, Quiznos), brunch (e.g. Denny’s Waffle Shop, Applebee’s), and dominant fast-food chains (e.g. Burger King, KFC, McDonalds, Taco Bell, Wendy’s).
These categories reveal the intricacies of the restaurant market when it comes to how consumers categorize brands. For example, many restaurants are found in multiple categories and have varying degrees of membership. There are also levels of abstraction revealing inferred hierarchical structure to categories (i.e. pizza and more specifically delivery pizza). Further, categories are created depending on specific items offered and contextual factors. Traditional clustering analysis would fail to depict these details and only found the sample to hold a category structure of 7, as opposed to 9, latent categories.
IMPLICATIONS & CONCLUSIONS
Blanchard’s modeling technique lends itself particularly well to showing how a consumer perceives a given restaurant, especially those that cross multiple categories according to consumers. For example, Mad Mex is a local, Mexican themed, waiter service restaurant known for its cheap, large burritos and margaritas. Consumers naturally see the restaurant as being very representative of the Mexican food category and very representative of the sandwich and subs category, likely due to its burritos. They also see Mad Mex as representative of the sit-down category, and the quick service category. Knowledge of the wide range of categories in which consumers place the brand allows marketers to identify the full swath of competition and target their messages to appeal to the many possible interpretations dependent on the consumer context. Conversely, some brands placed in a multitude of categories may see the evaluation as a problem and choose to refocus their image.
The technique also lends itself to identifying differing categorical structures used by diverse populations in assorted locations. It is reasonable to expect that different populations would have different latent category structures. For example, an older and wealthier group in NYC may exhibit a category structure that places greater emphasis on segmenting sit-down options than casual options. Smaller chains with less defined images may find they are differently categorized across markets, while they may compete in many categories in Penn State’s Happy Valley in bigger cities they may face a limited perception as a Mexican eatery. For businesses and market strategists, understanding consumers’ evolving categorizations of ever-changing
marketplaces is integral to gaining a competitive edge.