(Determination of Airline Ticket Pricing Based on Demand Forecast and Dynamic Pricing Policy)

Abstract

Purpose – Revenue management maximizes revenue using the diversification of the sales price. The issue of dynamic pricing and revenue management have gained importance because of an increasing appreciation of the airline in recent years. Design/methodology/approach – The aim of this study was to find an optimal dynamic ticket price based on forecasting customer demand, which maximizes an airline company’s revenue under capacity constraints. The least squares method, Poisson regression equation and support vector machine were used to analyse the demand structure of the airline. Findings – Demand models were obtained by considering weekly and daily demand data, seat sales price and aircraft capacity parameters for two local and three global routes. Then, the demand results were compared according to the square root of mean error (mean value of error) and mean absolute deviation (MAD). So, revenue functions were created based on demand functions for each route. Thus, revenue functions expressed according to demand used as objective function of a nonlinear programming and capacity constraints were added. Optimal dynamic ticket sales price which maximize the airline company’s revenue was found. Discussion – In addition, a new approach was proposed in the price policy development by combining demand forecasting, statistics, machine learning and operations research within the scope of dynamic pricing of non-stocked products. In this way, we believe that we have done a considerable contribution on determination of the dynamic ticket pricing for airline companies.