In recent times, a large part of the airline industry’s success has been due to an ability that many industries have struggled with – optimising revenue using Artificial Intelligence (AI)
By Neeraj Gargi
Airline ticket prices are decided by algorithms that change fares depending upon several factors such as past bookings, remaining capacity, average demand per routes and the probability of selling more seats later, etc; all of which can be included in a strategy called “Airline Revenue Management”.
Sabre, the largest Global Distribution Systems provider for air bookings in North America, describes this as a system that is “used to determine the optimal price of selling a seat at any given point in time”.
Airline revenue optimisation, though, is becoming increasingly difficult citing several factors, not the least of which is the upward trend in candidates to the cheap carrier segment and the ever-increasing list of competitive in-flight services. Plus, with increasing competition in the industry and market volatility, the airline industry is looking for solutions that will offer ways to maximise profits and deliver better customer experience and customer service.
Changing airline dynamic pricing science
As CRM systems and e-commerce strategies go through the transformation to adapt to changing buying patterns, preferences, the inability for airlines to implement an effective dynamic pricing strategy is one of the biggest challenges that the industry faces in the digital era. This is where AI comes in. Welcome to a sophisticated dynamic pricing science powered by AI, cognitive algorithms and machine learning technology.
Today, AI is doing more than just providing a seat. AI has unlocked the previously untapped data to enable more informed decision making across the airline industry. To optimise airline revenue management (RM) – organisations need an approach that will provide fair pricing – which many experts believe AI can provide.
AI and cognitive technologies are used to make sense of data that can streamline and automate analytics, customer service and machinery maintenance and other internal processes and tasks. AI is useful in various elements of airline operations, which includes revenue management tasks. Here is how they implement data analytics in RM.
Flight routes in demand
While revenue management is all about the best ways of selling a service, carriers use AI to answer a very specific question – “Where to fly?” Data Scientist Konstantin Vandyshev of Transavia, a KLM company explains, “To define air routes, specialists have to analyse data and make decisions based on the insights. When researching a demand for a destination among different customer groups, they can rely on such data sources as search history and macroeconomic factors (for example, GDP).” It has industry-specific standards that the experts use to determine the customers’ willingness to pay (WTP).
Willingness to pay
Airlines collect and crunch data about customers to understand their preferences and buying behaviour well enough to suggest them offers they are most likely to buy. Therefore, revenue managers start by measuring willingness to pay – when the customer is likely to pay the maximum price to buy one unit of a product.
In terms of airline ticketing, customers in general are ready to pay more when there is less time before the departure of the respective flight. The willingness to pay in the airline industry thus, depends upon a factor called “day before departure”. So, in practice, the experts define the average WTP price, which would be something that at least 50 per cent of customers would be likely pay for a specific ticket on a particular day before departure.
Such willingness to pay is equivalent to price elasticity (the number of passengers that would buy a ticket if the price drops by a certain percentage) with some hypothesis between the market demand and supply.
This particular metric is connected to dynamic pricing, which is the practice of pricing a product on the basis of customer’s will to pay. To calculate the WTP correctly, data must be selected accurately. Revenue management in the airline industry combines elements such as similar markets, differentiate peak and non-peak seasons, business destinations, holidays and weekends to accomplish this.
Expected marginal seat revenue
The expected marginal seat revenue (EMSR) is an optimisation model that is calculated after the WTP is set. This metric can be explained as the expected value of the current seat and requires allocating a seat to a particular fare class.
Data scientists measure the EMSR by multiplying the sales profit by the probability of selling additional (marginal) seat allocated to the specific fare class.
Altexsoft writes, “The moment comes when sales probability of a higher-fare ticket is so low that the expected revenue in a lower fare class will be bigger. So, knowing these probabilities you can determine the fare-class allocation for each day before departure.”
In a best case scenario, the scientists will have to know the sell-up probabilities in the different fare classes and the days before departure in order to find the WTP and the EMSR accurately.
The sell-up probability will show if the customer will buy a higher priced ticket if their request is denied. Flight grouping as per flight dates and destinations are needed. The revenue management team will also perform a “click-stream” analysis which figures out the number of customers that saw a webpage displaying a specific price. Airlines also use historical sales data to find out willingness to pay and the EMSR.
Ancillary price optimisation
This is another method used to drive airline revenue via analytics based pricing. It allows the data scientists to learn about a customer’s inclination to purchase ancillaries (add-ons) like baggage. Like for instance, experts determine the markets and days on which travellers are more likely to spend more to check their bags.
Any modern commerce needs to work at a break-neck speed, this means that airlines need to respond to customers with a precision and consistency across all channels. Of course, vendors that respond first win most percentage of the deals. Static, outdated prices that were once designed to protect profit margins today represent lost revenue.
Dynamic pricing based on AI technology offers what any modern trade needs to sustain in today’s competitive airline industry- speediness, fiscally-sound deals that are consistent across all channels.
The core of a successful modern venture lies in their ability to price dynamically – based on the insights gained from data science, creating experiences that enhance customer loyalty. Thus, by leveraging AI and accepting dynamic pricing grounded on cognitive technologies as the foundation of how they sell, the airline industry will be able to synchronise their pricing strategies consistent across all channels in real-time – present to the right customer on right time at the right price.
Airlines who will keep their operations innovative, embracing a science-driven business will be able to stay in this competitive market and capture opportunities that otherwise would be lost without sacrificing margin or leaking revenue.
(The author is the Chief Technology Officer at Intelegain)