A Mathematical Model Groups Like-Minded Consumers In Geographic Areas – Close and Far
- For national retailers, a new model shows how to satisfy consumers based on their psychological preferences and geographic location.
- Managers can use this model for better efficiency in their efforts to meet customer needs.
- The model can also give insight into the preferences of like-minded individuals with shared socioeconomic or demographic characteristics.
Will the Jetsons be the reality of the Millennials’ middle age? After all, videoconferencing is already the norm thanks to Skype and Facetime. Drones deliver our packages. Amazon owns Whole Foods, so pressing a button for breakfast might not be far off. As science turns fiction into reality, major retailers from Amazon to Wal-Mart to Kroger need to stay attuned to consumers if they want to stay relevant.
From Sydney to Pittsburgh to Houston, researchers are developing new tools to adapt their strategies to local demands. Rice Business Professor Vikas Mittal collaborated with colleagues Rahul Govind and Rabikar Chatterjee to create such a tool, which they call Spatially Dependent Segmentation (SDS).
Over the years, marketers have segmented customers either based on geography or their psychological beliefs. In one segmentation model, people living in the same zip code are classified as belonging to one segment (e.g., urban or rural). In another approach, customers are classified according to similar beliefs (such as prioritizing price or prioritizing service). The SDS approach does something different: it simultaneously considers consumer psychology and geographic location.
First, the scholars gathered multiple observations about consumer attitudes and needs in each area under analysis. Next, they incorporated estimates that group together similar units into spatially contiguous market segments. This allowed them to estimate consumer satisfaction levels in market segments that were spatially contiguous.
Developing market segments around concentrations of consumers with similar values, culture, and socioeconomics helps big retailers pinpoint ways to satisfy consumers at individual stores, even in very different neighborhoods. For example, big retailers no longer need to view all rural customers through the same lens. By clustering areas with similar consumer profiles, retailers of all kinds can more accurately choose products, services, and pricing, as well as provide training to employees and enough parking for customers at individual stores. For larger retailers, the improved precision makes managing a range of stores less cumbersome. The scholars’ model neatly improves efficiency while permitting better response to consumer demand.
While Mittal and his fellow researchers measured satisfaction of like-minded, neighboring consumers, they noted that their new methodology could apply to shopping scenarios beyond brick and mortar. As shopping moves online, for example, retailers can use this information to make recommendations during live online searches. Their model can also shed light on the tastes of like-minded individuals who live in very different communities.
As the reach of retailers continues to globalize and technologies to evolve, it’s easy to imagine the SDS model helping map regions for drone deliveries. It could also help big retailers spot interpersonal and other links between consumers in given areas. And it could shed light on how and why electoral voting patterns differ across regions and socio-economic groups. Adapting marketing and management strategies to specific stores and their customer base can helps ensure that each person who puts in a breakfast order finishes their bacon and eggs (or kale and quinoa) fully satisfied. Satisfied consumers are loyal customers – the secret to long-term success.
Vikas Mittal is the J. Hugh Liedtke Professor of Marketing and Management at the Jones Graduate School of Business at Rice University.
To learn more, please see: Govind, R., Chatterjee, R., & Mittal, V. (2017). Segmentation of Spatially Dependent Geographical Units: Model and Application. Management Science.