邮轮公司如何定价:基于北美市场的实证分析

How to Set Prices for Cruise Cabins:An Empirical Study on the North America Market

近年来,邮轮产业的发展速度有目共睹,已经成为整个旅游和休闲业中发展最迅速的行业之一。与航空业和酒店业类似,邮轮业具有传统收益管理的行业特征:存量有限、易逝性产品、提前预订和需求不确定性等。收益管理中的动态定价理论与方法已经广泛应用于存量固定且易逝性产品特征的行业,从而帮助企业在有限的销售周期内最大化自身收益。文章通过分析目前北美邮轮市场的一般定价模式,提出了一种基于需求学习的动态价格调整策略。该方法通过一种两阶段框架为邮轮特定航线、未来航次的不同周期进行动态定价,从而在有限的销售时间内最大化未来的总收益。第一阶段,基于历史数据和当前数据确定未来航次不同周期的需求函数;第二阶段,以当前获得的需求函数为基础,确定未来航次不同周期的最优价格,并采用到下一周期的实际预订中。当预订完成且新一周期的需求数据被观测到后,需求函数按照第一阶段的方法再次更新。这一需求学习和价格调整过程循环发生,直到所有的航次都启航。最后,以北美邮轮公司的实际数据为例,进一步讨论了动态定价方法的实施过程。实证分析表明,该方法能在一定程度上提高邮轮公司的总收益。

In recent years,we have witnessed the rapid development in the cruise line industry,which is a dynamic and growing segment in the entire tourism and leisure market.As measured by total passengers,the industry has grown at an annual rate of over 7% between 1990 and 2011.In terms of capacity,the North American market alone increased from around 41,000 berths in 1981 to over 340,000 berths in 2010 at an average annual growth rate of 7.6%.At present,the cruise line industry is one of the most profitable categories in the entire leisure market.A significant number of countries and regions have become economically dependent on the industry. Like other reservation-based and capacity-constrained travel and hospitality industries such as airlines and hotels,cruise lines face the problem of selling a fixed capacity of perishable products/services to different customer segments over a finite booking horizon with the purpose of maximizing the total expected revenue.In cruise ships,there are many kinds of cabin types with fixed capacities and different kinds of fare classes that are sold in advance over a finite booking window.All common properties of revenue management（RM） problems are reported in the cruise line industry,i.e.,segmented market,fixed capacity,perishable inventory,finite selling horizons,advance sales,and stochastic demand.However,cruise lines have received comparatively little attention in the RM literature compared to airlines and hotels,perhaps due to the lack of access to proprietary data. As in airline and hotel RM applications,cruise line RM includes demand forecasting,capacity allocation,and dynamic pricing.The goal of dynamic pricing is to change demand conditions profitably by adjusting the product prices from the start to the end of the selling horizon.Capacity-constrained companies usually offer different prices in different periods to segment the market in order to allocate the fixed capacity to the right customer at the right time.Methods of dynamic pricing have been widely studied in airline and hotel RM,but the cruise line industry has received little research attention. In this paper,we consider the dynamic pricing problem of a cruise line company.We propose a method to dynamically adjust prices of cabins for non-departed cruises for a given itinerary during finite selling horizon.The objective is to maximize total revenue to come at a given reading point.Our method has a two-stage price adjustment mechanism that accounts for demand learning.In the first stage,the parameters of demand function in each reading period are adjusted by re-running the regression problem when new data are observed.In the second stage,a constrained nonlinear programming is solved to determine the optimal prices for remaining periods.The process is repeated and the prices for remaining periods are updated as new data are observed over time.We conduct numerical tests of our method using data from a major North American cruise company.