BART Usability Assessment

During my last semester at Cal, I took a graduate course called Needs and Usability Assessment in the School of Information — a whirlwind tour of UX research methods taught by a Director of UX at Salesforce. It was one of the most industry-oriented courses I took at Berkeley, a huge change from the academic neuroscience and psychology classes I was used to. A lot of my personal philosophy about UX still comes from that instructor's anecdotal advice.

The class covered the full toolkit — ethics and recruiting, usability studies, field studies and observation, ethnographic research, interviews, diary studies, focus groups, contextual inquiries, expert review, competitive analyses, surveys, heuristic evaluations, card sorting, and accessibility. For the final project, my team had to pick at least three of these methods and run an entire usability assessment.

We wanted a subject many people had experience with and access to, so we picked the Bay Area Rapid Transit (BART) system. All UC Berkeley students get a Clipper card with their student ID, which made finding participants — and experiencing the subject matter ourselves — far easier than past projects like contextual inquiries for Adobe Illustrator.

Public transit draws users from every age, culture, and economic background. That variance makes for complex data, but also for an interesting design challenge — it's why I'd one day like to do UX work in the public sector.

01 Narrowing the scope

We started with a broad list of questions spanning ticketing, the trains, and station quality — the entire BART experience. To get a feel for how riders physically interacted with stations, we ran an observational field study: my teammates and I stood near the train entrances at the Downtown Berkeley station for an hour at 5 p.m., writing observations using the AEIOU framework — time spent at ticketing machines, Clipper card use, luggage, traffic, and more.

Reviewing our notes, we realized the scope was far too big to tell one cohesive story. We narrowed to ticketing — a step riders spent a lot of time on and had real autonomy over, choosing how to pay and what type of ticket to buy. To sharpen further, we focused on riders who didn't use Clipper cards, since paper-ticket users interacted with the machines far more.

02 The studies

Cognitive walkthrough

We asked four subjects to walk us through their ticket-buying process, watching for what felt memorable, intuitive, or problematic. At each step we asked follow-up questions, and found that subjects only noticed the options they were interested in — ignoring other features the machines offered. We also began spotting patterns in payment methods and ride frequency.

Interviews

For our final study, we designed a 12-question protocol and interviewed six riders about their habits, why they bought paper tickets, their payment methods, and the issues they hit. Two interviewees were from outside the Bay Area and had never used BART — a chance to delve deeper into the behaviors we'd seen in the field study and walkthrough.

03 Key findings

The first trend was a clear relationship between frequency of use and payment method. Paper tickets can be paid for with cash or card and reused by recharging, much like a Clipper card — which let us sort users into a 2×2 grid. Because there's no way to refund a ticket, riders must spend any remaining balance on the next ride, or top it up if it's short. Many liked having both payment options, but found paying exact fares a hassle, since amounts were uneven and machines only give change in coins.

Cash

Card

Multi
use

Single
use

Grouping observations by our two target metrics — speed and ease of use — surfaced two sharp pain points:

• Speed: when paying by card, the machine defaults to adding $20. To change it, riders can only increment by $1 or $0.50 at a time — so paying a single $4 fare means pressing a button 16 times.
• Ease of use: the ticketing process is disconnected from train routes. Machines aren't always near a system map, and the fare chart lists stops alphabetically rather than geographically — especially confusing for out-of-town riders who plan trips around cities or sites, not specific stops.

04 Suggestions & improvements

To address what riders told us, we compiled a list of changes BART could implement:

• Let riders purchase by destination instead of finding and adding specific amounts
• Integrate route maps into the machine UI
• Make exact fares easier by using whole-, half-, and quarter-dollar amounts
• Change the $20 card default, and allow increments larger than $1 or manual entry
• Let machines give change in bills, not just coins