Read Full Paper

Consumers facing increased time pressure have lower levels of attention, according to HKUST scholar Kohei Kawaguchi and colleagues. The researchers explored how consumers’ attention and utility (a measure of product usefulness) are influenced by time pressure and variables such as the number of recommended products. Ultimately, they concluded that “context-based marketing can increase sales more than traditional marketing in our setting.” This important finding may guide the design of product recommendations tailored to specific contexts.

As an alternative to traditional marketing based solely on demographic variables, marketing managers are increasingly interested in context-based marketing, wherein specific content (such as advertisements) is delivered to customers according to their location or behavior. “Foursquare, for example,” state the researchers, “sends messages to consumers when they are close to shops or restaurants that they are predicted to visit.” Still, context-based marketing and its optimization to increase revenue have not been well explored.

Kawaguchi and colleagues attempted to close this research gap by exploring how product recommendations affect consumers’ attention and utility under the influence of contextual factors such as time pressure, and how marketing interventions can be optimized accordingly. “One of our main interests is to examine the effects of contextual factors in designing recommendation systems,” the researchers explain. “Time pressure, however, is not usually measurable in a nonlaboratory environment.”

To circumvent this limitation, the researchers took a unique approach, using data from vending machines on the platforms of train stations in Tokyo. This setup enabled them to study time pressure outside the laboratory. “By utilizing the train schedule information,” they note, “we can precisely measure time pressure at the minute level.” The use of vending machine data also allowed the researchers to explore the effects of product recommendations and differences in the assortment of products offered.

The researchers then developed a consideration set model, wherein the consumer first considers a set of products and then selects the product with the highest utility. Next, they constructed a model in which both time pressure and recommendations could influence consumers’ attention and utility. Using this model and their unique dataset, they determined that “as time pressure increases, consumers pay less attention to each product but are more likely to purchase products.”

“Time pressure,” the researchers determined, “moderates the effect of recommendations.” Specifically, they observed that product recommendations promote both consumers’ attention and utility, although this relationship is weakened by time pressure. They further observed that by optimizing product recommendations to accommodate time pressure, the total sales volume increased by 3.7% and by 0.6 percentage points over the volumes achieved using uniform recommendations and a traditional strategy, respectively.

The researchers also studied the design of recommendations that depend on time pressure, finding that the number of total recommendations generally enhances consumers’ attention. However, this effect diminishes as the number of recommendations increases. “Each recommendation may increase the attention and choice probability of the recommended product,” they observed. Therefore, they conclude, “it may not be optimal to recommend too many products because it dilutes consumer attention.”