23 / 2024-04-12 10:59:05

Mitigating the Spiral Down Effect: Online Learning under Branded Fare Structure

Online decision making,Data-driven operations management,Airline revenue management

In the dynamic landscape of airline revenue management, accurately predicting customer behaviour and adjusting ticket-selling strategies in response to fluctuating demand is critical. Traditional models often overlook intricate customer buy-down patterns, resulting in a detrimental downward spiral of revenue. Furthermore, the assumption prevalent in much of the existing literature—that arrival rates and purchasing probabilities are known a priori—is rarely practical. This study combines the adjusted fare concept with the Bayes Selector algorithm to create an innovative approach for optimizing airline ticket-selling strategies under a mixed fare structure. By dynamically adjusting fare structures based on purchase probabilities and expected profits and utilizing the probabilistic estimates and adaptability of the Bayes Selector algorithm, this integrated methodology enables airlines to dynamically refine their seat allocation strategy for maximum profitability.

Our findings show that constant regret is attainable for the problem under the independent differentiated product demand setting. More interestingly, we establish a logarithmic regret bound in the case of a mixed fare structure, where tickets are priced based on the inclusion or exclusion of additional services.

This allows customers to pay only for the services they value but creates complex substitution patterns based on the available fare classes. The methodology encompasses two stages: an initial exploratory phase where the decision-maker (DM) experiments with various sales strategies to gauge customer reactions and deduce purchasing patterns, followed by an exploitation phase where, informed by insights from the exploratory stage, the DM employs a sequence of linear programming solutions to recalibrate sales tactics.

Furthermore, we illustrate the practical application of our approach through extensive simulation and a case study under a real-world airline revenue management scenario. This study concludes that by learning and adapting to the nuances of customer behaviour, airlines can significantly enhance their revenue management capabilities, leading to more accurate demand estimation and seat allocation strategies in alignment with the existing environment.