For the global travel industry, the subject of personalization has never been hotter. The number of articles, conference themes and startups pitches emphasizing personalization has reached the extreme hype phase of the cycle. In reality, the theme of personalization has been around for decades. For the travel industry, personalization has its own set of barriers especially for the infrequent leisure traveler. Let’s take a deeper dive into the subject and separate the hype from the reality.
I first became involved with travel industry personalization efforts when I advised Broadvision back in the late 1990s. Before the DOTCOM bubble burst, Broadvision was the premier platform for personalization. By 2000, Broadvision had contracts with major airlines (AA, TAM, Air Canada), Amadeus, EurailPass Forte Hotels, and IntraWest. Unfortunately, the Company’s fortunes were impacted hard by the DOTCOM bubble burst as their stock lost approximately 98% of its value by 2002. This did not mean the concept of personalization was dead, but instead indicated that the complex platform offered by Broadvision back in the 1990s was not able to meet the market’s needs.
Traditionally, personalization uses two basic techniques: explicit and implicit preferencing. Explicit preferencing asks the visitor to specify their preferences by selecting from a list of attributes such as “beach vacations”, while implicit personalization gleans a traveler’s preferences by their online behavior and past purchases.
So why is there so much hype about personalization today? Three driving forces(1) the customer experience with other e-commerce sites specially Amazon and Netflix, (2) the maturation of AI technologies (3) the growth of social media. Foremost on the AI front is the ability to analyze and understand vast amounts of unstructured data, including social media behavior. Incorporating unstructured data helps extend the knowledge of the customer, as well as providing better interpretation of supplier content (e.g. understanding the themes of customer reviews). This may include using a Google-like search bar to ask a more detailed vacation request (e.g. New York Hotel with lap pool and close to Grand Central Station). Behind the scenes Natural Language Processing (NLP) is used to understand text, traveler intent and match the request with hotels that not only have a pool, but ones where lap swimming is possible.
With the limited frequency of most leisure travelers, rather than offering true 1:1 personalization, many companies try to micro-target specific types of travelers within a group/segment (e.g. family, business, adventure travel, etc..). In general, the business travel segment is more attractive for personalization due to higher travel frequency and thus more data on the traveler’s preferences. In a corporate travel managed context, the challenge of personalization is that it may pit the traveler’s preferences against corporate policy. Dependent on what level of detail of customer behavior is captured, the trade off between personalization versus “the creepiness factor” is still a struggle as well. In addition, the fact that travelers may project different personas for a given trip or within the same trip, further complicates the effort toward 1:1 personalization.
I would propose that a more important use of customer data is to provide support and services based on the traveler’s context. Simply put, context is the immediate need of the traveler (e.g. what are they doing right now and what they will do next). Getting context right is extremely difficult. In order to provide the right context, the travel supplier or intermediary needs to track the trip closely and provide services on a real time basis. For example, something as simple as a text message identifying local coffee shops received when the traveler awakes is an example of a contextually relevant message. Personalization is also needed, especially in this case if customer data reveals that the traveler doesn’t even drink coffee!
Another example is offering “personalized” alternative itineraries when a trip gets disrupted. This may be an airline delay or a change in plans driven by natural disaster. To be effective, contextually based personalization needs to capture the intent of the trip. For example travelers who travel for business may lack the flexibility to dramatically change their itinerary and the impact of even a relatively short delay may cause issues for the traveler if they miss an important board or client meeting. For leisure travelers who planned a trip to Yosemite where reservations are often made a year in advance, contextually relevant personalized messages could provide alternatives destinations to accommodate the traveler due to the fires that has led to the Park’s closure.
So don’t get caught up in the hype around personalization and instead focus on a traveler’s immediate needs based on intimate customer knowledge. The more a supplier or intermediary can proactively communicate with a traveler throughout the travel cycle, the more effective the contextually relevant personalized communication can be and the more it provides a positive view of the supplier or intermediary.