How retail companies can better interpret visitor flows

How do customers move around in stores and when do they buy something? Anna Ulrichshofer and Prof. Dr. Michael Jungbluth from THI Business School used Artificial Intelligence to identify characteristic movement patterns based on almost 200 million anonymous sensor data. During their analysis, the researchers found that visitor flows can be reduced to a small number of patterns. Based on these patterns, they can better assess who is buying something and which customers are more likely to come in to browse. The latter often end up leaving the store without having purchased anything. The research findings of the THI researchers enable retail companies, for example, to better adapt the design and service offerings of their stationary stores to customer flows. Ulrichshofer and Jungbluth have now presented their work at the international ISMS Marketing Science Conference in Chicago.

"In the e-commerce sector, data is constantly being collected to analyze the buying behavior of customers. Websites are programmed in such a way that users' attention is specifically directed to further products or the conclusion of a purchase," explains Michael Jungbluth. "We have transferred this principle from online to stationary offline retail with the help of state-of-the-art sensor technology." The data was collected in the retail store of a major retailer.

Michael Jungbluth is a professor of Artificial Intelligence and Consumer Commerce, Anna Ulrichshofer a research associate. Both have been teaching and conducting research at THI Business School since 2021.