home » locations » uruguay » the path toward transformation in travel

The path toward transformation in travel


At the end of 2020, we introduced Sabre Travel AI ™–an industry-first Artificial Intelligence (AI)-driven technology platform created as part of a 10-year strategic innovation partnership with Google. Throughout 2021, we will explore the integration of AI and Machine Learning (ML) into several of our applications across travel retailing, distribution, and fulfillment.

We hope you’ll join us on the journey of unpacking Sabre Travel AI ™ technology as it transforms travel.

Acceleration into transformation

With the announcement of our Google Cloud partnership came a significant step forward in a journey toward digital transformation. As Sabre President and Chief Executive Officer, Sean Menke, stated, “For decades, Sabre has made travel easier for people on the go and within the industries that serve them…Google Cloud will help to accelerate our digital transformation and ability to create a new marketplace and critical products and systems focused on our customer needs for decades to come.”

The momentum for this transformation has been building for decades, as legacy software systems created a host of complexities and hardships. In airlines specifically, disparate databases were often tied tightly to individual functional areas such as revenue management, inventory control, shopping or departure control, creating a multitude of data silos over time. Today, some airline leaders cite accessing as many as 40 individual systems to complete complex business processes. This technical burden has prevented rapid advancement in many areas, including personalized retailing and real-time analytics–the focus of many other customer-centric industries over the last several years.

As we move Sabre applications to Google Cloud as part of our technology transformation, we can now unify these disparate data silos and integrate with state-of-the-art ML and compute infrastructure. In this blog installment, we will consider what Sabre Travel AI ™ means to retailing and the processes involved in personalized offer creation, and in subsequent installments, cover other aspects of how such technology might Sabre’s applications targeted toward distribution, fulfillment and operations.

First-, second- and third-generation technologies 

Originally, travel technology systems used for offer creation were built using hard-coded actions with pre-determined outcomes. The technologies of the 1980s and 1990s were efficient and built upon sophisticated understanding of operations research and yield management, but limited in their ability to react to competitors or changing market conditions in real time.

Generation 1: For example, an airline might have a static rule regarding the pricing of baggage, such that all travelers with a certain fare type are charged the same amount for their carry-on baggage. Even base fares were tied to single-letter booking classes, limiting the number of price points that could be specified for a given flight. The prices and options were determined by yield management algorithms and models managed by specialists–and the introduction of new products or retailing combinations was often manual and time consuming. The data available to such systems was also limited to an airline’s own flights and booking information, and customer segmentation that drove pricing and availability decisions was often simplistic.

As the internet boomed and data and analytics became more prevalent in our industry, platforms and software became capable of actioning flexible rules that allowed for dynamic human interventions and management. The introduction of unbundled ancillaries and the need to price them using flexible models became prevalent as Low-Cost Carriers (LCC) showed how profitable such product combinations could be. Models could sense and respond to changing market dynamics, while analysts could adjust rules to respond to such changing market conditions.

Generation 2: In this baggage example, it was now possible to take that static baggage fee and adjust it to a set of variable amounts based on factors such as loyalty tier or reaccommodation history. We could create separate systems that would use rules to compute how much prices might vary depending on factors such as distance traveled, time of day, etc. But these rules still only had access to an airline’s own data, and still required analysts to create and maintain them using time-consuming processes.

As an industry, we’re now on the cusp of the third generation–where it’s possible to implement machine-generated recommendations that learn from transactional data in real-time. Now, applications are market adaptive. We can take into account data about how other suppliers may be pricing their products and how consumers are reacting to such offers, while taking in a variety of other factors such as demand and intent data. Once coded and trained, these ML models learn to automatically segment users into meaningful subsets, creating truly personalized offers: the right price bundle, for the right customer at the right time.

Generation 3: With our baggage example, this enables an airline to offer a nearly unlimited amount of price points for their baggage fees by assessing vast amounts of customer and operational data. By automatically segmenting users and determining optimal prices that increase conversions and yield, airlines can now operate with higher degrees of automation and less manual intervention. We envision a world where truly unique offers–combinations of base prices and the right combination of ancillaries–are created automatically and offered to truly delight users. If a business traveler is not going to check-in baggage, perhaps they might be enticed by an offer that includes an ancillary they care about for the same price–such as early boarding or premium seating.

Overall, this evolution has allowed travel companies to move from intuition-driven to a more automated, data-driven and scientific decision-making model. Gone are the days of analysts making educated guesses about how to tune systems. Now, automated systems can continually evaluate the data to make optimal, personalized recommendations. These systems can also automatically segment users based on their willingness to pay and adjust prices and yields to not only maximize yields for airlines, but also satisfaction for users–since they’re getting the combination of product attributes they want.

COVID-19 as a further catalyst to change

It would be remiss to glaze over the change that continues to re-shape our travel landscape. COVID-19 has brought with it a need to better understand travelers, what they want and when they want it, at a rate simply unsuitable for legacy, manual processes. At the same time, many in the industry continue to look to technology for differentiation and to win business amidst slowly re-bounding demand. Many have found that digital and self-service technologies have helped handle the influx of this rapid change, which will likely take on new heights as the industry continues to recover.

Our partnership with Google has become even more critical for growing and facilitating personalized travel to bring value to modern travelers. There’s a great deal of emphasis placed on real-time customer service, retail offers and flexibility as we witness historical data patterns becoming less reliable. The accelerated pace of change in this new environment requires AI technologies that enable “an expansion of data sources for the real-time and predictive analysis of data for better decisions and the optimization of many workflows” (Phocuswire, 2018).

In our next installment, we’ll dive deeper into the application of AI and ML in travel.

To learn more about Sabre Travel AI ™, visit our microsite.


Share this post