Ice-climbing, high-altitude mountaineering, whitewater kayaking, backcountry skiing, ultra-endurance running and mountain biking are just a few of the activities I enjoy during my free time. It would be easy for those who don’t know me to look at these hobbies and conclude I’m a thrill-seeker and a risk-taker, living on the edge in search of the next extreme experience.
But for me, being high up on a steep mountainside or kayaking through rapids with a singular focus on the task at hand, brings a sense of calm, clarity and connection. It’s a momentary escape from the fast-paced world we live in.
The reasons why people do the things they do or make the choices they make are often telling, and they can fall into certain categories. Commonalities emerge and, with a large enough sample size, it becomes possible to make accurate inferences about a group of people.
In the absence of individual insights, personalization can be challenging. However, by using any available contextual information and focusing on the ‘why’, an opportunity emerges. I believe this approach – sometimes referred to as non-personalization – could be used by airlines to build greater traveler-centricity as we evolve towards modern travel retailing.
Building traveler-centricity in the airline industry
One of the biggest challenges facing the airline industry right now centers around how to become more traveler-centric. In other words, putting the traveler at the heart of the airline decision-making process.
We’ve seen a rapid evolution of consumer expectations because of their experiences in other verticals: abundant choice, personalized recommendations and seamless fulfillment. Understandably, they expect a similar experience when it comes to travel, and solving the traveler-centricity puzzle is critical to success in modern retailing.
Personalization in some form is a key enabler of traveler-centricity, and when it comes to creating personalized travel offers, data is king. I mentioned in my recent blog that airlines collect more data than any other player in the travel ecosystem and they are no strangers to complex data analysis. Mathematical models relying on historical data have been used by airlines in many ways; price optimization, scheduling, operational efficiency and customer loyalty to name just a few. But in a modern retailing environment, airlines need to analyze and apply data in real-time – and therein lies the challenge.
The personalization challenge
Personalization in the traditional sense – designing something to meet someone’s individual requirements – presents a unique challenge for airlines because 70% of customers travel with an airline no more than once a year. This results in a high volume of anonymous travel shoppers and a very low volume of ‘shopping basket’ data that can be analyzed at an individual traveler level.
Even for travelers who do make multiple trips with an airline, the complexity of those trips can render the data less valuable for analysis purposes. For example, a traveler may book a week of skiing with a group of friends early in the year, a solo business trip to a city center for two nights and a two-week family vacation at a beach resort or theme park in the summer. Although the same traveler is searching for each trip, their requirements and preferences may vary significantly based on the trip purpose.
And let’s not forget, even where prior trip data is available, past behavior is not always an accurate predictor of future purchasing propensity. Consider the example of a once-in-a-lifetime vacation to Egypt to cruise along the Nile and see the Great Pyramid of Giza. That exact same trip is highly unlikely to be repeated when that traveler comes to book their next trip.
Could non-personalization be a solution?
The airline industry must consider a range of potential approaches to the broad concept of personalization and what it can deliver. For example, I believe airlines could see powerful results through non-personalization using trip-purpose segmentation, which seeks to better understand the ‘why’ behind traveler preferences.
Non-personalization means making trip recommendations that are not based on a traveler’s past behavior or personal data. Instead, it uses contextual information such as search parameters (which may include number and age of travelers, origin and destination point, dates of travel, etc.) from which to make inferences. In the case of an anonymous travel shopper, this may be the only information available to the airline.
Of course, where personal information is available, a hybrid solution is the smart choice – using personal data to enrich any contextual information for an even more targeted recommendation. I always cite the example of a traveler who had a negative experience on their previous trip. When that same traveler shops for their next flight, if the airline can identify the prior service failure and respond (by offering a free seat upgrade, for example) the traveler experience can be greatly enhanced.
Trip-purpose segmentation uses transaction data to group travelers into cohorts using trip-purpose (the ‘why’) as the defining characteristic of their purchasing behavior.
This concept is not new. From the earliest days of revenue management pricing, a Saturday overnight stay was used to define business trip-purpose vs leisure trip-purpose.
Today, thanks to advances in technology, the application of AI/ML to marketplace data can analyze combinations of shopping request attributes to identify commonalities across trips to create unique cohorts. For example, all itineraries that represent a first-ever trip to Paris creates a cohort for trip-purpose segmentation that can be used by airlines to create offers that better match traveler needs and are therefore more likely to convert into revenue.
The Sabre Labs team is at the cutting edge of developments in the arena of traveler-centricity through uncovering trip-purpose segments – you can read more about that work here.
Every day, I help Sabre’s airline partners explore new and innovative ways they can use data and technology to help them better anticipate traveler needs. To overcome the challenge of becoming more traveler-centric and reap the rewards as they accelerate towards modern travel retailing. It’s a complex challenge being faced across the industry, but with the pioneering work of the Sabre Labs team combined with Google Cloud Platform’s infrastructure, data analytics and machine learning, I’m confident Sabre’s airline partners will be ideally placed to succeed.
