The hospitality industry wants to get to know you better. That means more than just your preference for a non-smoking room with foam pillows far from the elevators.
Big data is a game changer when it comes to pricing hotel rooms to reduce idle inventory, Martha Poulter, CIO of Starwood (hot) Hotels and Resorts, told Fortune this week.
"Today we can price our rooms based on a relevant combination of internal and external data sources and optimize accordingly," she noted. That means taking advantage of data about weather; outside events like conferences and concerts; public information about competitors' performance; and customers' past travel preferences. If you've shown a predilection for staying in cities near water and Starwood has a low occupancy rate at its Halifax or Boston Harbor properties, you might get a great promotion offering a discounted stay or package deal there.
Here's the thing about hotel rooms, there may be a lot of them—Starwood manages 358,000 across 1,260 properties—but they remain a relatively inelastic, limited resource. And it is a huge analytical problem to keep them booked and to make sure that needed maintenance happens during down time, not when a guest is in the room. If Starwood—or competitors Hilton (hlt) and Marriott (mar), for that matter—can boost occupancy even a few percentage points, that's a big deal.
By using Starwood's proprietary data analysis tool, Poulter said the company was able to boost the revenue and market share of one Halifax property by 20%. "That flows right to the bottom line," she said.
Another goal is to take as much friction out of the booking-checkin-checkout process as possible so that guests book through Starwood's own properties rather than Expedia, Hotels.com, or another third-party travel site. Perks like the Starwood Preferred Guest or SPG keyless entry phone app which lets guests at select W, Aloft, and Element properties download key information and room number to their device and use it to gain entry. (No check-in lines!) That service is not available to people who book through third parties.
When you think about the world of connected things, there are obvious benefits that come from the ability to perform preventive maintenance in remote, inhospitable environments such as mines or oil fields. If you get data showing that a gasket or filter on a big piece of equipment is starting to malfunction, you can get the right service person with the right part to the right place before anything goes haywire.
But the same sort of early warning system is also useful in more comfortable settings.
Sick of seeing room service trays piled up in the hall? What if there is a sensor in the tray and in the door so that housekeeping will know when you've put your tray outside without you having to call? And how nice would it be if staff knows a lightbulb is flickering and that the room is unoccupied so maintenance can get to work without disturbing anyone?
Data collection can also make it eaiser for guests to get acclimated to their new room. Some new properties are so concerned with maintaining a sleek, minimalist appearance that it can take a full 15 to 30 minutes for a guest to find and figure out the light switches, curtain controls and TV operation (I'm looking at you Aria.) A Microsoft executive cited data at a recent hospitality technology conference, that confirms my experience, noting that a typical hotel guest takes 12 minutes to figure out the thermostat, lights and TV controls in a hotel room. That's probably about 10 minutes too long. If hotels can get that sort of usage data maybe they'll be smarter about how they design these interfaces in the future.
Subscribe to Data Sheet, Fortune’s daily newsletter on the business of technology.