A dynamic pricing engine for multiple substitutable flights
As enhancements in airline IT begin to expand pricing and revenue management (RM) capabilities, airlines are starting to develop dynamic pricing engines (DPEs) to dynamically adjust the fares that would normally be offered by existing pricing and RM systems. In past work, simulations have found that DPEs can lead to revenue gains for airlines over traditional pricing and RM. However, these algorithms typically price each itinerary independently without directly considering the attributes and availability of other alternatives. In this paper, we introduce a dynamic pricing engine that simultaneously prices multiple substitutable itineraries that depart at different times. Using a Hotelling line (also called a locational choice model) to represent customer tradeoffs between departure times and price, the DPE dynamically suggests increments or decrements to the prices of pre-determined fare products as a function of booking request characteristics, departure time preferences, and the airline’s estimates of customer willingness-to-pay. Simulations in the Passenger Origin–Destination Simulator (PODS) show that simultaneous dynamic pricing can result in revenue gains of between 5 and 7% over traditional RM when used in a simple network with one airline and two flights. The heuristic produces revenue gains by stimulating new bookings, encouraging business passenger buy-up, and leading to spiral-up of forecast demand. However, simultaneous dynamic pricing produces marginal gains of less than 1% over a DPE that prices each itinerary independently. Given the complexity of specifying and implementing a simultaneous pricing model in practice, practitioners may prefer to use a flight-by-flight approach when developing DPEs.
KeywordsDynamic pricing Airline revenue management Substitutable flights Dynamic pricing engine New Distribution Capability
Michael Wittman would like to thank Craig Hopperstad for excellent development assistance with PODS; members of the MIT PODS Consortium for funding and helpful suggestions; participants of the 2017 AGIFORS RM Study Group Meeting; and Jan Vilhelmsen, Robin Adelving, and Jean-Michel Sauvage for their hospitality during the author’s research visit at Amadeus in August 2016.
- Aydin, G. and J.K. Ryan. 2000. Product Line Selection and Pricing Under the Multinomial Logit Model. In Proceedings of the 2000 MSOM Conference, Ann Arbor, MI, 1–49.Google Scholar
- Belobaba, P.P. 1992. Optimal Vs. Heuristic Methods for Nested Seat Allocation. Presentation to AGIFORS Reservations Control Study Group Meeting, Brussels, Belgium.Google Scholar
- Boeing. 1996. The Decision Window Path Preference Model: Summary Discussion. Marketing and Business Strategy, Boeing Commercial AirplanesGoogle Scholar
- Bratu, S. 1998. Network Value Concept in Airline Revenue Management. Unpublished Master’s Thesis, Massachusetts Institute of Technology, Cambridge, MA.Google Scholar
- Mumbower, S., L.A. Garrow, and M.J. Higgins. 2014. Estimating Flight-Level Price Elasticities Using Online Airline Data: A First Step Toward Integrating Pricing, Demand, and Revenue Optimization. Transportation Research Part A: Policy and Practice 66: 196–212.Google Scholar
- Ratliff, R. 2017. Industry-Standard Specifications for Air Dynamic Pricing Engines: Progress Update. In Proceedings of the 2017 AGIFORS Revenue Management Study Group Meeting, San Francisco, CA.Google Scholar
- Seelhorst, M., and Y. Liu. 2015. Latent Air Travel Preferences: Understanding the Role of Frequent Flyer Programs on Itinerary Choice. Transportation Research Part A: Policy and Practice 80: 49–61.Google Scholar
- Smith, B.C. and C.W. Penn. 1988. Analysis of Alternative Origin-Destination Control Strategies. In AGIFORS Symposium Proceedings, vol. 28, New Seabury, MA.Google Scholar
- U.S. Department of Transportation. 2016. The Value of Travel Time Savings: Departmental Guidance for Conducting Economic Evaluations Revision 2 (2016 Update). 27 September 2016.Google Scholar