As airlines across the globe move to embrace offers and
orders, one of the key areas of value creation lies in creating optimal offers
that tailor flight and ancillary products to customer
preference. To do so, many are turning to sophisticated revenue management
platforms with expertise in AI to offer customers more personalized experiences
and generating extra percentage points of income as a result.
A number of forward-thinking carriers are exploring how such
systems can be used to dynamically price ancillaries, such as seat selection,
bags and Wi-Fi.
PROS principal consultant and formerly the company’s vice president of product
management for its airline products John McBride said, “Airlines have
very rich data for the right to fly, meaning just to get on the plane. We do an
amazing job of modeling price elasticity for the right to fly, but when it
comes to ancillary data, that’s more difficult.”
PROS started its work on dynamic pricing of
ancillaries with Portuguese carrier TAP in 2017. Some of the early work
discovered that the way that you present ancillaries, particularly on mobile
devices, matters.
“What you bubble up to the top matters based on the
context. Have they already purchased this type of fare? Are they the type of
passenger who is staying one day and so they probably don't want bags but may
want Wi-Fi instead? We have proven that the display order or ranking matters.
But what about the price? Of course, the price matters of an ancillary but the
data is not very rich.”
How do you figure out price elasticity when airlines
rarely change the price of an ancillary?
“That was probably one of my favorite problems to
solve,” McBride said. “We had to create algorithms that are capable of
figuring out price elasticity, and we call them explore-exploit models.”
It is the concept of whether to stick with something
you already know or try something new in the hope that it might be better –
like eating out at a favorite restaurant or venturing somewhere new that might
turn out to have better food.
The algorithm behind PROS Dynamic Ancillary Pricing (DAP) platform works to
dynamically set the price of ancillary products based on trip features (like
day of week), as well as based on market conditions and passenger segment
interests, preferences and willingness to purchase. No personal data is
used.
AirBaltic, which operates more than 130 routes from
Riga, Tallinn, Vilnius and Tampere to Europe, the Middle East, North Africa,
and the Caucasus region, has integrated DAP into its 2e Systems internet booking engine
to drive extra revenues from customers.
The objective of the implementation was:
- to achieve optimal prices for ancillaries and deliver better
conversion
- to introduce workflow automation to free airline employees
from the repetitive tasks and help avoid any sporadic human errors
- to gain a competitive edge to stay ahead of the market by
leveraging the platform’s AI capabilities
Justin Jander, senior director of project management at PROS,
explains how this works with seat pricing for two different groups of
passengers on the same flight.
Choosing a flight from Riga to Frankfurt, Jander considers
the seats available for a single passenger and then for the same flight but for
two passengers travelling together. Both are offered a range of different seat
types that might have extra legroom or be closer to the front of the plane.
Jander explains that the pricing of the seats differs
depending on the request made.
“You can see these two seats are nineteen euros, but when it
was just one passenger making the booking, the same seat was fourteen euros. So
what DAP recognized is that as a passenger, there's a higher likelihood to
request seats together when you have an extra passenger in the request. It
understands the booking request, details of what they want, and the different
price sensitivities or elasticities of the passenger making that decision in
real-time deciding how to price those two things differently. So, it really
opens the door up for more revenue opportunity.”
Iuliia Granja Velasco, e-commerce project manager at
airBaltic said, “The first stage was to identify the segmentation parameters
that AI would use, for, proper forecasting prices and, achieving better offer
for the passengers.”
“At the same time, we were working on technical
implementation for our booking engine to pass the data about passengers' seat
reservation behavior to a processing system, and from where AI was supposed to
pick it up and, process it.
During the second phase, we already went live with the DAP,
but that was still a learning and adjustment stage, when AI was, picking more
and more data every day, learning our customers' behavior patterns, and,
predicting the price more accurately with each passing day and, of course,
adjusting on the go.”
“The third stage was all about A/B testing. We were running
tests, splitting our traffic, and comparing prices that were generated by AI
with those that were generated by a rule-based system.”
“At the end of the day, the outcome was very, very positive.
We gained six percent revenue per percent increase, and that was the major
outcome, of course. And on top of that, we managed to cut our maintenance
efforts in managing rule-based pricing, which was sometimes time-consuming.”
Eva Plakane, senior vice president of revenue management of airBaltic says that the airline
is keen to integrate AI capabilities into its workflows to enhance efficiency
and optimize its operations and that DAP exemplifies this commitment.
She said, “DAP has delivered tangible and quantifiable
results, driving measurable growth and empowering airBaltic to thrive in a dynamic and
competitive market landscape.”
The creation of AI-driven optimized offers as branded fares
or bundles represents a significant opportunity for carriers. In the offer and
order setup that many airlines are moving to, they can evolve to dynamic
bundles of products and services that are constructed in real-time to match
customer demand, as well as the airline revenue goals. Upsell offers,
regardless of if bundled or not, can also be offered real-time based on
AI-driven recommendations for what is most relevant to the traveler.
It's not about squeezing the customer for every last cent
but about them being willing to pay more for a personalized fare that better
meets their needs. It’s win-win for both airline and customer.