Better to deliver tomorrow morning than late today? The time-of-day effect in customer perception of online sellers' logistics service quality

Drawing upon energy depletion theory and expectancy disconfirmation theory, this study aims to zoom in on the effect of delivery time on customer perception of online seller logistics service quality (LSQ). We conjecture that customer rating of LSQ will vary depending on the delivery time in a day (i.e., the time-of-day effect). With a large sample consisting of more than 42 million orders from Alibaba, the results from mixed-effects ordered logit model corroborate the time-of-day effect and that promised delivery service interacts with the time-of-day effect by strengthening it. Following that, machine learning techniques are employed to quantify the importance of the different predictors and results show that the time-of-day effect is the most important predictor. The study reveals a new time-related attribute, contributes to the LSQ framework, and has important managerial implications for practitioners.