We’re creating a more connected travel industry, underpinned by sustainability and long-term investor relations.
References to The Godfather aside, an optimized offer for airlines is one that targeted passengers can’t refuse. It’s a proposal to a customer to sell a seat and/or ancillaries at a particular price. This price is optimized to maximize revenue for the airline, while also being acceptable to the customer. In a simple scenario, the price could be fixed and or even published.
It is much more effective and powerful if this price is dynamic and targeted. That means being tailored exactly for the customer who requested it and when they requested it. The optimized offer is an effective method of conveying revenue management solutions directly at a level of detail and accuracy previously unobtainable.
To truly support optimized offers, we need to re-imagine the traditional revenue management and availability processes. For example, business travelers may receive offers containing changeable fares with lounge access, while leisure customers receive basic fares and options with checked baggage. Prices are assigned to each offer taking into account aspects such as customer loyalty, market conditions, analyst adjustments, and price optimization.
Besides the real-time systems, forecasting and optimization components are necessary to drive the offer construction and dynamic pricing systems. Next generation revenue management systems will incorporate advanced customer choice preferences and discrete choice models. This technique generates choice probabilities for each alternative flight or bundle that a customer might consider. Choice models require large sets of current, historical and competitive data, thus yielding results that are a natural fit with dynamic offers and pricing.
When we look at the task of optimizing both prices and offer mixes in a network environment, the number of potential combinations that the customer could buy is staggering. Customers may choose pure air products, or fares bundled with a variety of options. Of course the challenge increases further with optimization at the network level rather than leg based, crossing multiple airports and equipment types. The task quickly exceeds the capabilities of direct optimization techniques.
Fortunately, there have been a number of new advances in Artificial Intelligence and Machine Learning in recent years; and companies such as Google, IBM, and Facebook have effectively implemented these technologies to better predict weather conditions, optimize ad relevance, understand human speech, and power intelligent assistant systems. One common requirement necessary for successful Artificial Intelligence and Machine Learning systems is relevant shopping data, which is key to develop customer choice models that can help form expected purchasing patterns. We see a strong role for big data and Artificial Intelligence and Machine Learning in optimizing dynamically built and priced offer.
Want to learn more about this subject? Check out our report, A spotlight on Total Offer Optimization: Fast forward to customer centric revenue management , for more.