Optimal promises: A general framework with application to airline schedule times

Sellers in a wide range of industries frequently make promises about service performance (product quality) that can only be assessed after purchase. Making such a promise creates a tradeoff: A better promise (1) increases demand today, but (2) reduces the likelihood of achieving the promised performance, which reduces future demand. In this paper, we present a simple model of optimal promised performance. We then apply this model to the airline industry, and we develop hypotheses for how changes in market conditions -- the availability of on-time performance data, competition, and prevalence of connecting passengers -- lead to changes in promised performance via their impact on one or both components of the aforementioned promise tradeoff. Lastly, we test our hypotheses using airlines on-time performance data. Our results provide support for our hypotheses, and more broadly for our model of promises and the associated tradeoffs, indicating that airlines promise more when the cost of failing to attain the promised performance level is low: When passengers are less informed about airlines' on-time performance; when airlines face little competition; and, when there are relatively fewer connecting passengers.