麻豆原创

Airline dynamic pricing is a term that has been used in many different contexts, usually as some type of automated pricing tool. At 麻豆原创, we refer to dynamic pricing of airfares to describe an engine that can help airlines provide tailored offers to different customer segments, considering a carrier鈥檚 revenue management controls and data signals on the current competitive marketplace.

A close cousin to dynamic pricing is dynamic availability. With dynamic availability, rather than directly modifying the price amounts, an airline鈥檚 availability controls are updated (either by closing or opening booking classes). The airline鈥檚 ability to modify prices in real-time with dynamic pricing provides the maximum revenue performance, but today it is limited to airline direct channels and NDC offers. On the other hand, dynamic availability changes can be applied across all GDS selling channels, so while the revenue performance of each change is slightly less, the total applicable sales base is larger. Thus, both approaches drive value for airlines.

麻豆原创鈥檚 dynamic pricing and availability products have been shown to provide up to鈥�3% revenue gains. These gains are incremental to what an airline realizes from its revenue management system, and they arise from more customized offer decisions, continuous price points instead of fare classes (see illustration), and the ability to manage prices with finer granularity (i.e. at the roundtrip level rather than departure date only for traditional inventory control). This finer-grained, continuous pricing also benefits travelers because they gain access to new price points that were previously unsupported. For more information explore:

One of the biggest trends in airline marketing over the past 10 years has been the strong and steady growth of ancillary sales for air travel. According to a , airline ancillary revenue trends since 2013 show an average estimated annual growth rate of more than 15%. Ancillary sales growth in this period arises from both individual ancillary sales (a la carte) and the proliferation of airline branded fare bundles.

In the past, ancillary prices were a secondary consideration by airline marketing teams, and those prices were usually static across airline regions. But as sales (and revenues) have increasingly shifted to ancillaries, the opportunities for more dynamic pricing with AI/ML technology have grown. AI/ML models can help airlines recommend optimal prices for ancillary products or services using a mix of supervised learning (based on estimated price elasticities and factors like marketing segment, type of seat, position in cabin, etc.) and reinforcement learning methods (see 鈥楨xperimentation engines鈥� section below). 

In practice, our customers have seen incremental ancillary revenue gains of up to 10% from the use of AI/ML models for both dynamic price decreases and increases of ancillaries. For more information, see:

Our new experimentation engine platform, based on a type of AI/ML known as reinforcement learning, provides exciting opportunities for all travel sellers. Traditional AI/ML models build complex prediction models relying on historical data. Experimentation engines, however, do not rely on extensive historical data; they try out (i.e. experiment) multiple options and observe the outcomes. The models learn from those outcomes so the parameters that produce outputs are adjusted a bit, and the process can be run again (and again, and again).

One example use case is 麻豆原创 Hospitality鈥檚 SynXis庐 Retail Studio™ application for hotels, which has many similarities to airline retailing applications. A hotel can have many different ancillary products and choosing the right ones to present to each guest type is a complicated problem.  

The graphs below demonstrate how the experimentation engine model learns which product to offer two different customer segments, A and B. At first (see the left side of the graphs), the model offers all three products. With time, it settles on a clear solution: A prefers breakfast and B prefers high floor. The model learns as it gets more data points (from impressions and clicks), so the degree of 鈥檈xploration鈥� can be reduced as it gains confidence about the best products to display. 

Two of the most important questions in airline network planning when designing future schedules are understanding:

鈥淲here do people want to go?鈥� and 鈥淗ow many people want to travel between those origin-destination pairs?鈥�

If we know both answers, we know travel demand, and hence we can help optimize travel supply (i.e. determining the number of planes needed, where they should fly to and how they should be scheduled). That means we can build systems that estimate expected revenue and profitability, fine tune retailing and distribution strategies, plan resources, etc.

Accurately answering 鈥淗ow many people want to travel from city A to city B?鈥� is important for aircraft manufacturers, airlines, hotel operators, travel agencies, local businesses oriented to tourism and the investment community. There are about 250,000 city pairs in the world that have air service between them, and travel demand for them is changing every month; thus, answering that question with manual market analysis is virtually impossible.

This is where 麻豆原创鈥檚 new network planning tools can help. Our 麻豆原创 Labs team is working on an AI/ML model known as the 鈥淢arket Potential Forecaster鈥� which predicts how many people would travel from A to B per month for the next 12 months for all 250,000 markets. And it updates such predictions periodically. The underlying AI/ML model is trained on both historical trends and forward-looking air low fare search shopping data. Our colleague, Cuneyd Kaya, provided further model details in this recent 麻豆原创 article.

By using a combination of historical data, analyzing segment behaviors and forward-looking shopping data, the AI/ML-based market potential forecaster model has the potential to increase average prediction accuracy to up to 90%.

Some of the most common disruptions airline passengers face are flight cancellations and delays, but other kinds of disruptions exist. 麻豆原创鈥檚 AI/ML models for reaccommodation during irregular operations such as unplanned flight schedule changes, cancellations, etc. (called IROPS) can help in such scenarios by providing new itineraries accounting for global availability, passenger prioritization and airline practices.

These AI/ML models can also be used as a 鈥榳hat-if鈥� tool where airline analysts evaluate the effect of a particular disruption (say, a change of equipment to a smaller capacity plane). Based on the solution, they can decide if this is an acceptable action to perform. If the tool is used, and results are applied, passengers are automatically guaranteed a ticket in the new itinerary as the rebooking process handles this. 

Whenever the IROPS application is run, and passengers are informed of their new itineraries, those passengers need to decide if the new trip is something they are still interested in taking or they prefer rescheduling or completely canceling.  

A businesswoman with an important meeting Monday morning would cancel rather than take an alternative itinerary arriving Monday lunch time. Someone on a week-long vacation in Italy would likely accept to be rebooked to arrive half a day later.鈥疎xperienced analysts can recognize some of these patterns and offer only those reaccommodation options that are likely to satisfy travelers. To help airline analysts further, 麻豆原创 developed a machine learning model that helps evaluate if the replacement itinerary would be accepted or not.