An application of demand forecasting and fare management for an airline company

In this thesis, we aim to propose a model for estimating demand and obtaining ticket fare that maximizes revenue, based on real market data. In this context, eight different single leg local market routes are examined, and various time series analysis methods and regression analysis are applied. The purpose of implementation of time series analysis is to reach the best forecasted demand values that are tested against historical realizations of 26 weeks of sales data. In order to generate models and compare results, R (programming language) is used as a tool. After obtaining the best fitted demand models for each market route, fare values are found by using a proposed fare production process. This process is developed to produce demand-based fare values. Forecasted revenue with forecasted demand and produced fare values are achieved for each market route, and all results are examined separately to analyze performance of demand models and fares.