Dynamic Pricing in the Ferry Industry

选择“最佳”票价是渡轮行业的主要挑战之一，特别是对于同时提供个人和货运服务的渡轮运营商。在厘定票价时，营办商须(1)预测各类乘客及其车辆对渡轮服务的需求；(2)确定不同类型车辆应分配多少空间；(3)估计车辆可以多容易地挤进可用的甲板空间。预测需求，为不同的客户找到最有利可图的空间分配，以及机票定价都属于收益管理（RM）的总称，它最初是为航空公司开发的，但适用于广泛的行业。什么是新的在这个项目的第一部分是包装的包含。通过将渡轮上车辆的最佳包装纳入RM，我们将找到定价和分配解决方案，以提高渡轮服务的效率，并确保定价适当地反映将车辆包装到有限的甲板空间中的成本。传统上，RM专注于使每次旅行的收入最大化，但有必要考虑更大的前景，并考虑价格对运营商长期盈利能力的影响。轮渡主要用于不经常旅行的游客和经常使用的客户，例如货运、通勤和定期长途汽车服务。通过优化长期收入，例如一年以上，以及考虑个别航次，我们将能够考虑到老客户的总贡献。例如，全年运营的货运公司或定期使用渡轮的通勤者不应该在夏季高峰期间被市场淘汰，而应该提供一个反映他们对公司的长期价值的价格。令人惊讶的是，这方面的工作几乎没有开展，这与几乎所有的运输供应商都有关，对于避免在高峰时段为老客户定价过高至关重要。通过与P&O渡轮公司和Red Funnel公司合作，我们将使用真实的数据来为模型提供信息，这两家公司分别在英国大陆和欧洲大陆以及怀特岛之间经营渡轮。这些真实的数据是从各种内部或外部系统/来源（如订票系统、公司网站、一线操作系统、营销活动记录、市场竞争力报告等）收集的。本项目将首先研究如何将这些数据联系在一起，以便它们可以用于构建和测试我们提出的定量模型。因此，该项目的成果将成为利用“大数据”潜力的一个很好的例子。在收集和准备数据后，我们将开发模型来估计票价，使渡轮运营商的收入最大化。提高收入将通过两种方式实现：(1)由于更有效的打包算法，增加了可以打包到渡轮上的车辆数量；(2)根据不同航次的预测需求进行价格优化。改善渡轮上的车辆组合及其包装方式将提高渡轮服务的效率，对环境产生积极影响。这项工作在若干工业部门具有广泛的影响，特别是在运费的最佳定价方面，在确定运费时需要考虑到包装。开发优化长期收益的方法，可以改善任何行业的定价，在这些行业中，既有定期流量，也有偶尔流量。

Choosing the "best" ticket prices is one of the key challenges in the ferry industry, especially for ferry operators providing a service to both individuals and freight. When setting the ticket price, the operators need to (1) forecast the demand for ferry services by various types of passengers and their vehicles; (2) decide how much space should be allocated to different vehicle types; and (3) estimate how easily vehicles can be packed into the available deck space. Forecasting demand, finding the most profitable allocation of space to different customers, and pricing of tickets fall under the umbrella term of Revenue Management (RM), which was originally developed for airlines, but is applicable across a wide-range of industries. What is new in this first part of the project is the inclusion of packing. By incorporating optimal packing of vehicles on the ferry into RM, we will find pricing and allocation solutions that increase the efficiency of ferry services and ensure the pricing properly reflects the cost of packing a vehicle into the limited deck space.Traditionally, RM has focused on maximising the revenue on each individual journey, but there is a need to look at the bigger picture and consider the effects of prices on the long-term profitability of the operator. Ferries are used by tourists who travel relatively infrequently and by regular customers, e.g. freight, commuters and regular coach services. By optimizing revenue in the longer term, e.g. over one year, as well as considering individual sailings, we will be able to take account of the total contribution of regular customers. For example, a freight company that operates year round or commuters who use the ferry regularly should not be priced out of the market during the peak summer season, but should be offered a price that reflects their long-term value to the company. Surprisingly little work has been carried out in this area, which is relevant to nearly all transport providers and is vital to avoid over-pricing tickets for regular customers at peak times. Through working with P&O Ferries and Red Funnel, who operate ferries between mainland Britain and the Continent, and the Isle of Wight, respectively, we will use real data to inform the models. These real data are collected from various internal or external systems/sources (e.g. ticket booking systems, company's websites, frontline operating systems, marketing campaign records, market competitiveness reports, etc.). This project will first look at how to link these data together so that they could be used in building and testing our proposed quantitative models. The result of this project will therefore become a good example of utilising the potential of "Big Data". After collecting and preparing the data, our models will be developed to estimate the ticket prices which maximise revenue for the ferry operator. Improving revenues will be achieved in two ways: (1) increasing the number of vehicles that can be packed onto the ferry thanks to more effective packing algorithms; (2) optimizing prices based on forecast demand for different sailings. Improving the mix of vehicles on the ferry and the way they are packed will increase the efficiency of ferry services, having a positive environmental impact. The work has wide-ranging implications in a number of industry sectors, particularly in optimal pricing for freight, where packing needs to be taken into account when setting delivery charges. Developing methods for optimizing revenue in the long-term could improve the pricing in any industries in which there is a mix of regular and occasional traffic.