• Bookmark and Share
  • PRINT

How Data Mining Impacts Your Travel

20121112-a-data-mining.jpg
© Tina Zellmer / Debut Art

Travel providers are secretly using your past to predict your future via data mining.

What if the next hotel you stay in (or airline or car rental company you use) knows what you want before you do? Believe this: They all want to, and right now, they are in a race to harness what the wonks call predictive analytics. The topic can lose us in the digital weeds but, ultimately, it is as simple as getting to know customers in ways that make a difference to this guest’s experience right now.

Benefits of Pedictive Analysis

Simon Talling-Smith, executive vice president for the Americas at British Airways, explains what this is all about: “Too many airlines treat even their frequent flyers as strangers. We have the data that will let us treat them as they want to be treated.”

Airlines—and certainly hotels—have massive amounts of guest data acquired via loyalty programs, Talling-Smith points out, and suddenly, travel providers have begun to analyze data in ways the industry hopes will increase personalization.

Big though the data may be, this topic quickly gets as small—and perhaps personal—as what you drink on your next flight. For instance: What if every time you fly, you start the trip with a scotch—perhaps a particular brand—and water on the side. Would it be that hard for a flight attendant to greet you with, “Ms. Jones, good to see you again. Would you like XYZ scotch and water?”

That information—and much more—is now in iPads that are issued to British Airways flight attendant team leaders, and other carriers are following suit.

A Game Changer

For centuries, every good hotel general manager has maintained a database, often on 3x5 cards, on his or her best repeat guests. Favorite flower? Preferred bottled water? Fluffy pillows or hard? Does this guest like to be recognized by name or does he prefer anonymity at the front desk? (Some celebs and the very wealthy always travel under a pseudonym.) That good GM knew all this, but he probably did not share it across the chain. The other, more telling issue: The data were gathered on only an elite group. Even an industrious GM knew little about 95 percent of his guests.

Today’s search for meaningful data may well be the travel industry game changer. “Arguably, data is the oil of the 21st century,” says Giles Nelson, chief technology officer at Bedford, Mass.–based Progress Software, where a focus is on developing tools for data analysis.

Truth is, however, these are still the early days, particularly in travel, says Steve Peterson, global travel and transportation leader at the IBM Institute for Business Value. Progress is coming fast, he adds optimistically. “The ability of travel providers to efficiently gather and analyze information about the needs and preferences of their customers combined with their ability to deliver useful insights to the customers and support staff is indeed revolutionary. These changes are fueling a predictive analytics–enabled transformation that has the potential to change the face of the travel industry in the years ahead.”

Adding New Channels of Information

At Destination Hotels, a management company based in Colorado, repeat guests are tracked not just in the company’s own database but also in publicly available information streams such as Twitter and LinkedIn. Now what if the hotel spots a tweet by an incoming guest: “Flight delayed. I’m starving. What I’d give for a pastrami sandwich and a cold beer!” And what if shortly after the guest arrives at the property, there’s a knock at the door and room service is there with the sandwich and the beer? “We can make that happen,” says Maureen Callahan, vice president of marketing at Destination.

The fine-tuning can get even more granular, says Nancy Kern, an assistant vice president at Destination. What if, for instance, a regular guest always books a spa session on every visit—or maybe it’s a round of golf—and what if the computer notices that the guest is returning soon and spa appointments are booking fast? “We can contact that guest while there still is time to get her or him booked into the spa and before we have a disappointed guest.” Will the contact be by phone or email or text message? That preference too is noted in the system so the guest hears from Destination as she prefers.

Turning Data to Insights

One key in bringing all this smarter analysis to the fore, elaborates Talling-Smith, is a change in how data are gathered and stored. Data had historically existed in silos with little sharing from one silo to another; the reservations team might not have shared its data with the in-flight teams, mainly because no easy way to do so existed. Now data are gathered in robust data warehouses, where, with a few clicks of relational tools, fascinating insights begin to tumble out of the computers.

Case in point: “When you don’t fly can be as interesting as when you do,” says Talling-Smith. He cited a hypothetical passenger who for a number of years has flown once a quarter, always in business class, round trip from New York to London. “He almost certainly serves on a board,” guesses Talling-Smith. But then he falls out of the reservations system. Is he off the board? Did he switch carriers? If the latter is true, British Airways wants to win that business back. “Should we offer an upgrade to first class?” asks Talling-Smith. “The data will help us decide.”

Data will let travel providers better predict what interests you, says Mark Simpson, president of Maxymiser, a company that specializes in using data to attempt to “optimize customer experiences.”

He gives an example. If you are based in Edinburgh and every January you book a weeklong holiday to the sunny Canary Islands, when you log on this winter it is simply annoying to be served ads for discounts on Berlin getaways. The system knows enough to know you do not want more gray skies. But, says Simpson, you will probably welcome seeing discounts on sunny getaways. “We look at harnessing data from previous engagements,” he explains. “We use that information to determine the next best content to create more engagement.”

Change of Plans

And then predictive analytics begins to up the travel ante. Imagine this: Every morning you take the Circle Line to work in central London (or maybe it is BART in San Francisco or the MTA in New York). Now further imagine—and this will be easy with the troubled Circle Line, which often has immense delays—that the Tube is going to be nonfunctional during the morning commute. Most regular commuters use cards that are smart enough to identify the user and to record historic usage. “What if you got an email that told you about the commuting disaster, offered an alternative, explained how to do it with minimal disruption and perhaps sweetened the offer with a discount on a cup of coffee or tea?” asks Martin Howell, director of worldwide marketing communications for Cubic Transportation Systems, which works with public transit companies in a number of big cities, including London.

Put that email in that passenger’s hands before he heads out for a bad morning wrestling with the Circle Line, and suddenly, a fraught day can become good, says Howell.

A next frontier: pushing that predictive power out to airline schedules. As major disruptions occur (due to anything from volcanic ash cloud to hurricanes), it’s still a matter of a computer putting pairs together and making recommendations.

“It’s complicated. It’s not just a flight. It’s a hotel, ground transportation, possibly other vendors,” says Minoo Patel, vice president of mobility and social media practice at NIIT Technologies, an India-headquartered IT company that is in the hunt for algorithms to market to corporate travel managers and individual frequent travelers. “Every traveler is different. Some want to spend the night where they are, leaving early the next morning. Others want out as soon as possible, even if it is a red-eye. And then we have to factor in corporate travel policies,” which may not permit that red-eye traveler to upgrade to business class. “The trick is getting enough detailed data on the individual, the company and the situation. When we have that we believe we will be able to predict the best recovery scenarios for this traveler.”

And that will create a happier traveler, even in the midst of a bad trip. And better travel, ultimately, is what this is all about.

Fast-Forward

So what does big data mean to you, the traveler? Not necessarily that much today. But in a short year or two, say the big data wonks, a lot of travel will begin to be shaped by predictive analytics.

  • Better targeted recommendations. Just as Amazon already uses predictive analytics to recommend books or music based upon your past purchases, watch for the same from travel providers soon, says Dr. Dave Dimas, a professor at the University of California at Irvine Extension who is involved in its certificate program in predictive analytics. Fly coach and always stay in boutique hotels? Why pester you with alternatives that won’t interest you? Dimas predicts that very soon, travel websites will display personalized recommendations that will be “very, very good.”

  • Total travel experiences. Watch for a raising of the recommendations bar as travel providers begin to use big data to offer “a more complete travel experience,” says Steven Ramirez, president of Beyond the Arc, a predictive analytics firm. When booking a hotel room, for example, it will become commonplace for the provider to offer airport transfers, perhaps a spa appointment, possibly a dinner reservation, Ramirez elaborates. And these suggestions will be precisely tailored to your particular choices as demonstrated by your past behavior. For the provider this is classic upselling and cross-selling. For the traveler, however, it will increasingly be viewed as a value-added convenience that simplifies the travel experience, says Ramirez.

  • Pinpointed restaurant recommendations. The recommendations keep getting smarter. For instance: You are going to Berlin, or maybe New York. What if a computer could point to where you might like to eat, based on where you like to eat in other cities and where your friends eat—with a particular focus on eateries known only by locals, not travel guidebooks? That is what Australian Sean Craig is cooking up with Eight Spots (eightspots.com), a site that intends to use predictive analytics to guide you to good eats and spare you bad and bland meals on the road. Input a handful of your favorite restaurants in your home city, and Eight Spots will stir the pot and come up with probable new faves in other towns, says Craig. Watch other purveyors do much the same as big data analytics becomes more common.

  • Easier hotel check-ins. Soon you will be able to check into your hotel before you set foot on the property, says Maryam Wehe, a senior vice president at big data firm Applied Predictive Technologies, which counts among its clients Holiday Inn, Choice Hotels and Hotel Indigo. Airlines have permitted this—indeed encouraged it—for some years, Wehe acknowledges, but watch, it’s coming to a hotel near you.

Robert McGarvey covers business, technology and travel for outlets ranging from The New York Times to the Harvard Business Review.


Market Place