Reinventing a Business Model to Restore Travelocity’s Financial Strength

The Problem

Pioneering online travel booking service Travelocity began to sustain serious market share erosion as competition from new market entrants Orbitz and Expedia muscled into the market. In 2002, six years after Travelocity’s 1996 launch, the company sustained a 32% decline in annual bookings. Travelocity’s original business model was based upon earning commissions from airlines from flights booked on the Travelocity system. However, airlines had begun to curtail paying commissions to travel agents of any kind — websites or human. As a result, Travelocity’s future was very much in doubt. It became clear that Travelocity would need to become a travel retail service, collecting fees from individual consumers.

The Analytics Solution

Travelocity Revenue Management teamed up with Sabre Research Group in a top-to-bottom review of drivers of online retail profitability. Thanks to its size and the nature of its business, Travelocity is able to collect detailed information on more than a half a million online shopping sessions each day. The trick was to efficiently sift through that massive data to analyze what customers ask for, the alternatives they receive, and the actions they take. Using multinomial discrete customer choice models, Travelocity and Sabre analyzed that data and estimated customer preference for brand, quality, and screen placement. The models enabled Travelocity to make better planning and retailing decisions and to create a process to determine which content is the most compelling to consumers, thereby maximizing consumer interaction. In addition, the team developed an analytics-based enterprise network model to gauge the impact that new or modified supplier agreements would have on site content and revenue. It supports the negotiation of agreements with airlines. That model is a large-scale, mixed-integer program that generates the parameters that control the display of air shopping responses, plus financial forecasts and sales targets. Also, the team developed an analytics-based process to quickly measure the performance of keyword searches and recommended adjustments to the bidding strategy for each paid search. Yet another element of the turnaround strategy for Travelocity involved the creation of customer choice models, nonlinear optimization, and simulation processes to set prices for so-called merchant products, including hotels and travel packages.

The Value

With its more sophisticated business model and tactical decision-making processes, Travelocity witnessed a doubling of its revenue over a three-year period and the restoration of its market share. Analytics-based decision support modeling in network planning, pricing, advertising, display, and customer behavior all contributed to this successful turnaround.