One day last summer, MIT students with clipboards asked hundreds of people waiting for the Red Line at South Station what they thought of the T. When they returned the next day and asked the same questions, they found customers had become happier overnight.
The stations were no cleaner, the trains no faster. What changed? The MBTA had activated countdown clocks overnight, providing predictions in minutes of when the next trains should arrive.
Before the clocks, customers tended to be overly pessimistic about how long it would take, thinking they would need to wait longer than they did. After the countdown clocks appeared, satisfaction with the T’s performance jumped 15 percent.
Massachusetts Bay Transportation Authority leaders were thrilled, but the point of the surveys was not for the transit authority to pat itself on the back. Instead, it is part of a larger MBTA effort to harness data and gain insight into the customer experience.
“The more we can understand what is important to the customers at many levels is important,” said General Manager Beverly A. Scott. “We are really going to start being able to get our arms around customers’ perceptions, behaviors, what drives them — what makes for a competitive transit-travel package.”
By that, Scott means luring customers who have choices. In Atlanta, she presided over a transit system largely eschewed by those with cars. She was drawn to Boston by the mix of what she calls “lifestyle riders” and “lifeline riders.” Scott said she is committed to serving both, but attracting customers who have other options is key to expanding ridership, bolstering the T, and helping reduce congestion and greenhouse-gas emissions.
“You have to be competitive,” Scott said, wanting to entice riders who are open to the idea of transit but have been turned off by the uncertainty of waiting or the hassle of too many transfers.
Starting with South Station last summer, the T has activated countdown clocks at 30 subway stations, harnessing existing signs and audiovisual equipment originally installed for other purposes. Compared with the millions of dollars some agencies have spent on countdown clocks, the T did it on a relative shoestring, with an in-house team of tech-savvy employees. Costs (less than $1 million) were limited to new computer servers and software.
Some might snicker — the historic but frequently derided MBTA is something Bostonians love to hate — but the T is taking to heart the message from the Patrick administration to be more attentive to customer service, said Josh Robin, director of innovation.
Starting four years ago, the traditionally guarded MBTA shared data with software developers about bus and train schedules, real-time locations, and arrival predictions, moves that yielded dozens of applications for customers at little or no cost. Riders have given them high marks anecdotally, but thoroughly vetted customer satisfaction data are lacking, Robin said. So when he began the countdown sign project, Robin wanted a better way to measure his thesis “that short of new stations and new vehicles, which are billions of dollars’ worth of investment, I don’t think there are many things that would get you the same impact” on customer satisfaction.
Research elsewhere has shown that bus passengers without countdown information overestimate their wait times. And from package delivery to Internet downloads, society has grown to expect tracking tools and status bars — and to be distressed without them, Robin said.
But data about subway predictions were lacking, which is why MIT researchers were enthusiastic about gauging customer perceptions on the two days sandwiching the arrival of the countdown clocks at South Station — a window that MIT’s David Block-Schachter called “the gold standard” from a research perspective.
The challenge, Block-Schachter said, was to distill a survey to its essence to avoid delaying commuters trying to catch trains and to reach more than 1,300 people in two days. Researchers also needed to take into account the varying answers they might get from people depending on whether they were just arriving or had already waited several minutes.
So they asked riders to rate the MBTA overall — from poor to excellent — and they asked how many minutes people expected to wait in total, from the time they entered South Station to the time their anticipated train arrived. The researchers later checked to see when each rider’s train arrived, using data collected automatically by MBTA dispatchers.Continued...