Yet travel marketers are dealing with a very particular set of behaviors that make it hard to use data the way many other marketers do. Travel can be expensive and purchases are often paid for by someone else (work), or packed with emotion (leisure.)
Because of these differences, travel marketers must step up their game.
More Detail Required
Most standard ad targeting begins with demographic data about customers. But beyond a certain income level, demographics are a poor indicator of a consumer’s travel behavior.
Take two married women in the same upper-middle-class Zip code. Clothing retailers can be relatively certain that both will buy a few sweaters in the fall. They can focus on a few details like color and size and have a relatively successful “targeted” ad campaign.
Travel companies need much more specific data. One of the women above might travel extensively for work, earn loyalty rewards, and take frequent short domestic holidays with her family. The other may rarely travel, taking one big trip a year somewhere abroad. Airline marketers would need to discern business from leisure travel, international from domestic preferences, loyalty status, frequency and timing of messaging, and more.
The Past Doesn’t Always Predict The Future
A recent Wall Street Journalreport illustrated how more companies are starting to use customer lifetime value scores to prioritize customer service, discounts, and other interactions with customers.
For travel brands, as we’ve seen, demographics aren’t enough to calculate lifetime value, but neither is past spending behavior. Instead, travel brands must not only use many more unique personal inputs to calculate a more accurate potential future value of an individual.
Using a potential value model takes new data points into consideration. For example, someone may travel extensively — with another airline. The current value of that person might seem to be low, but her future value is enormous. Many college graduates starting their first jobs will begin traveling much more extensively, and will be signing up for rewards programs for the first time. These “potential value” inputs should often override past spending behavior.
Take the example of a hotel mega chain. If it were to prioritize a current high spender over a current low spender using past data only, it would miss the opportunity to delight a newly minted consultant — who just got assigned a long-long-distance job in a city she needs to travel to four nights every week for the next year.
Every Insight Is Relative
Hans Rosling, the author of “Factfulness,” likes to explain that data is relative, and that anyone looking at data needs to consider the context. For example, Thailand gets more than 35 million tourists per year. That’s a lot! Or is it? In fact, it is a lot. But France, the top destination, gets nearly 90 million, while the average country receives far fewer.
Marketers should put their own data in context. Perhaps a hotel or airline’s most valuable customers actually travel more to another destination, or spend more with a different airline. Perhaps that other resort or airline has a higher average spend overall. These relative data points are required to put marketing plans in the right context.
Innovative travel companies are turning to machine learning to contextualize data even more, and improve over time. The largest “pre-transaction” travel website is even tagging all of its community content to create metadata that will help it create more personalized search results and marketing communication.
Travelers are sophisticated and diverse, and travel marketing must be, too.