Deepak Ohri is CEO of the luxury Lebua Hotels & Resorts, headquartered in Bangkok. He recently told me that his properties in that city have the highest revenue per available room (RevPAR) of any local hotel.
That's quite an accomplishment. Bangkok has an abundance of luxe accommodations with great brand recognition: Mandarin Oriental, Peninsula, St. Regis, Shangri-La, Four Seasons, Banyan Tree.
Ohri told me that the secret to his enviable RevPAR is that he is methodical.
He contracts with a company that monitors the name Lebua on public chat, using software originally developed for the FBI. A team of 30 people then uses applied behavior patterning to map the transcripts and eliminate all those who are unlikely to visit Bangkok. Another analysis cuts out those unable to afford a luxury property.
If he started with 100,000 chat messages, he said, the first cut would reduce the number to 20,000, the second to 500.
Then it gets surgical. If, for example, three of those 500 live in Chicago, his team can, using clues in the chat, possibly find out who those three people are. And once he has names, he checks to see if he knows someone who knows them.
If it turns out that he does, he asks the go-between to arrange a meeting over coffee with the targeted potential guest and his staff, to talk about Lebua.
He said that, admittedly, the number with whom he or someone has just two degrees of separation is small, but that this method has yielded 1,100 contacts in the past eight years (he first used it for a Lebua restaurant in Bangkok). And once he connects with a target, he said, he stays in close touch. Those 1,100, he believes, are what give him the RevPAR edge.
By coincidence, later that day I sat down with Barney Harford, CEO of Orbitz. He, too, wanted to talk about microsegmentation, but applied in a very different manner.
He looks at microsegmentation as an extension of "sort." He characterized it as "de-averaging."
Orbitz, like other sophisticated websites, can tell a lot about who comes to visit. His staff knows, for example, whether you're using a Mac or PC. When he looked into the booking behavior of the two types of computer users, he found that a shopper on a Mac pays an average of $20 more per night. As a result, he can show a different (and slightly more upscale) set of hotels to Mac users.
Likewise, Orbitz has noticed that there are behavioral differences among people with different browsers, i.e., Internet Explorer vs. Firefox vs. Google Chrome.
As a result, he plans to build out multiple results pages, each tuned to different behavioral patterns. He can go deeper, presumably offering a different page for a PC user on Explorer than for a PC user deploying Google Chrome.
But it doesn't stop there. Orbitz is creating new data fields for properties, developing, for example, a "kid friendliness" scale. If you state you're traveling with children on the first screen, the hotels you'll see on the following page will be different from the ones seen by childless bookers. (And, going deeper still, different from those seen by parents booking on a Mac with Firefox.)
Harford said that, for now, it all happens in the background, but he plans eventually to sometimes say "We're recommending XYZ because ..."
Does this sound a bit Big Brotherish? At the risk of mixing my futuristic-paranoid literary metaphors, welcome to the Brave New World.
In travel, a case could be made that Big Brother is being employed to help the customer. As the general manager of any five-star hotel will tell you, people love it when you know and remember their preferences.
Former ASTA president Kathy Sudeikis told me that as a test, she once called the concierge at a Ritz-Carlton and told him she liked rocks in her pillowcase. Now, whenever she checks into a Ritz, there are rocks in her pillowcase.
It's possible to look at microsegmentation simply as a means to deduce who among the millions is likely to want rocks in their pillow without them having to tell you. The flip side: You may be miscast if you happen to be the lone Mac user who actually prefers to pay $20 less than the average PC owner for a hotel.
Email Arnie Weissmann at [email protected] and follow him on Twitter.