Doing the Research seriesHuman or machine—who do customers order more from?

16 August 2023, by Newsroom editorial office

Photo: pixabay/Fotorech

Whether when shopping or in the service sector, customers are increasingly coming into contact with automated processes first rather than with employees. In a project based on fast-food restaurants, Prof. Dr. Karen Gedenk at the Faculty of Business Administration (Hamburg Business School)—along with Dr. Kai Widdecke and Dr. Cord Otten—is examining whether this changes buying behavior.

Previously, people in fast-food restaurants were asked what they would like. In many places today, customers type their order into a self-service kiosk. How are you examining the effects of these changes on buying behavior?

We are working with a fast-food company that has introduced self-service kiosks in its restaurants. This means that customers can place an order either with this machine or with an employee.

For our research, the company provided us with transaction data from 20 of its restaurants, which we used to measure whether customers buy more or less when placing orders via a self-service kiosk rather than with a person. We also surveyed customers at 2 of the company’s restaurants to explore causes of the behavior.

What did you discover?

Actually, fast-food customers order significantly more at the self-service kiosk than at the counter: on average, they spent 14 and 16 percent more in our 2 studies. For the company, this is of course gratifying, as it generates significantly more revenue and increases the contribution margin. However, the effects on customers’ health are less pleasant, because customers order not only more food through self-service kiosks but also more unhealthy food. For example, orders through self-service kiosks contain 32 percent more sugar and 42 percent more trans fats.

Did you find out why orders via self-service kiosks are larger on average?

The evidence points to 2 causes. First, self-service kiosks are better at cross-selling—that is, they automatically offer other products and consistently ask, for example, whether you would like a coffee with your meal. Second, there is less social pressure when ordering at self-service kiosks. People mainly buy unhealthy products in fast food restaurants, and that can be embarrassing when facing employees at the counter. At the self-service kiosk, people feel less observed and judged. 

How can and should these findings be used practically?

Of course, the research is particularly interesting for companies, as they can derive important information from it for the use of self-service kiosks. Automation is often viewed particularly from the perspective of saving on staff costs. Our research shows that significant revenue effects can also occur. However, our research also has implications for society and policy with regard to promoting healthy eating. For example, restaurants could be required to provide clearly visible and comprehensible nutritional information on self-service kiosks to caution customers against unhealthy eating.