Should I stay or should I go? An empirical analysis of consumer behavior using airline web-traffic data

Abstract

We analyze consumer search and purchase behavior in response to airline revenue-management practices using data from a major carrier’s website and Google Flights. We first describe patterns in search timing, purchase decisions, and paid fares. Then we estimate a multinomial logistic regression to identify factors driving search timing, finding that single adults with loyalty status, especially booking one-way nonstop itineraries, tend to search closer to departure. Next, we use a binary logistic model of conversions of searches to sales, showing that competitors’ prices and changing customer composition explain rising conversion probabilities as departure nears. Finally, using a fixed-effects regression, we reveal how search and booking patterns affect prices paid. Late-arriving travelers, particularly single adults with loyalty status, pay substantially more, consistent with the airline’s pricing strategies that segment more inelastic customers. Overall, our findings underscore how revenue-management, competitor fares, and consumer characteristics jointly shape online search and purchase behavior. • We analyze consumer search and purchase behavior in response to airline revenue-management practices using data from a major carrier’s website and Google Flights. • We first provide a descriptive analysis of patterns in search timing, purchase decisions, and paid fares. • We then estimate a series of logistic and fixed-effects regressions to reveal drivers of the timing of searches and sales and how search and booking patterns affect prices paid. • Overall, our findings underscore how revenue-management, competitor fares, and consumer characteristics jointly shape online search and purchase behavior.