Segmentation of Theatre Audiences: A Latent Class Approach for Combined Data

Abstract

Theatrical productions are supposed to be perishable good, since the tickets for a particular play cannot be inventoried and sold after a time of play. In the revenue management of a perishable good price discrimination is widely used. Since the theatre audience is heterogeneous in terms of visit purpose, ability to perceive quality, willingness-to-pay, the strategy of price discrimination is developed in the context of theatre segments. In this paper, we segment consumers of Perm Opera and Ballet Theatre and propose marketing instruments to increase theatre revenue. Since development of detailed price discrimination strategy requires data on consumer’s purchase history, her behavioral and socio-demographic characteristics, we collect and combine two data sources: data on ticket purchases and data obtained from discrete choice experiment. We modify latent class logit model for joint revealed and stated preferences data where data on consumer characteristics is only partially observed and employ the model to segment the audience. We identify four segments of the theater's audience. The study reveals theatregoers segments with different willingness-to-pay for performance and seat location characteristics. Segmenting allows to develop detailed recommendations on the pricing strategy for various theater audiences.