United Airlines | 2018
Redesign the self-service “rebooking kiosks” located at United airport gates, thereby decreasing customer dependence on live airport agents and call/contact centers.
Worked with UX research to create direction for usability testing studies
Provided direction to UI designer responsible for building usability study prototypes
Defined UX strategy for each study based on user and stakeholder feedback
Co-managed relationship with stakeholder group
Pitched and received buy-in from airport strategy leaders
When flights are delayed or canceled, United customers can rebook their reservations themselves by using the mobile app, a web browser, or one of the airport kiosks located in the terminal. But even with so many options, in 2019, these self-service methods only accounted for 11% of total rebookings. Customers were relying heavily on agents and contact centers.
On days of irregular operations at the airport, this translated to long lines, long wait times, and a lot of frustration. And yet, a study by the company’s business team found that 50-70% of customers who waited in line at the service desk could have actually rebooked themselves. So, why didn’t they?
To learn more, my team and I visited O’Hare Airport on a busy summer day in 2019 when heavy thunderstorms disrupted air traffic.
This is what we discovered:
The rebooking request might seem too complex for a kiosk to process. I watched one customer interact with an agent for an hour, thereby holding up the line for hundreds of people. She was trying to get a group of 20 Brazilians to Omaha, Nebraska, but flights were canceled for the rest of the day and booked solid through the rest of the week.
Customers are confused by airline policies. They often asked if they could be rebooked on a different airline, if they’d get meal or hotel vouchers, if they could get a refund or a rental car, etc.
Stressed-out customers want to seek help and sympathy from a human. They think they can get better rebooking options if they explain their situation to a sympathetic agent.
Airport workers don’t often refer customers to the self-service kiosks. Gate agents, who do not process rebookings, suggest that travelers call the customer service phone line or visit the customer service center. They rarely mention self-service rebooking options.
Legacy kiosks are located far from some gates. Because rebooking kiosks are located at the customer service center, customers figure that if they have to walk there anyway, they might as well just get in line to talk to a person.
Needless to say, our airport observations uncovered so many big issues that discouraged customers from rebooking their flights with United’s self-service methods. To mitigate all of them, we’d need a large team and a lot of resources. The business team took our recommendation to heart and began planning a business case to reimagine the overall experience for customers needing to rebook. But my team and I hypothesized that, in the meantime, we could still improve the user experience of the rebooking application.
When we began the project in the fall of 2019, United was in the process of replatforming its app and website. We had no choice but to zone in on the self-service rebooking kiosks. We didn’t know much about them. Data from the business team was unclear, but we did learn that only 16% of customers who began a rebooking on the kiosk actually completed the transaction. The best-case scenario was that customers were opening the kiosk application only to discover that their automatic rebooking actually was the best option for getting to their destination. But we hypothesized that this kind of positive experience was rare.
Our first prototype involved some low-hanging-fruit updates to copy, but mostly it consisted of a baseline usability test of the existing kiosk rebooking application. We made no functional or flow changes, instead keeping the four rebooking options offered at the existing kiosk:
Testing it, we began to identify the biggest pain points:
Flights with available seats were presented alongside standby flight options.
While each flight type was clearly labeled as available or standby, many users didn’t understand what flying standby meant. Furthermore, in this design, there wasn’t enough screen real estate to adequately explain it. A reference button had been included on the page, but it was not given much importance amidst a crowded design.
Flights from nearby airports were given the same level of importance in the rebooking flow as flights from the original airport.
In reality, flying out of a different airport is a worst-case scenario—and so uncommon that I didn’t believe it needed to be featured in the self-service application at all. It raises too many complex questions for customers, which may ultimately have sent them to a customer service agent (e.g., “If I fly out of a different airport today, will United provide transportation?” and “I’ve already checked my bags, so how will I get them?”).
The four rebooking options were nested in the flow, leaving some users to wonder what tasks the kiosks could actually help them accomplish.
We hypothesized that communicating the capabilities of the kiosk from the get-go would provide more transparency for users.
By the time users arrived at the flight confirmation page, they still weren’t completely confident that they understood their new itineraries. That had to change in the next iteration.
Because of the confusion surrounding the above flow, our second prototype completely overhauled the navigation and decision tree. We thought about how live agents help customers: they give the customer all of her options up front and allow her to make decisions based on her individual needs. We tried to emulate this with our second design. We extracted the four rebooking options from the flow and put them on the front screen. This also separated standby options from the flights with available seats, which we hoped would lead to less confusion.
This design also failed:
Moving all of the rebooking options to the front overloaded the home screen with copy.
The hierarchy still did not prioritize rebookings on flights with available seats.
Participants expected to view these flight options immediately, but the button was buried at the bottom of the page under a significant amount of copy.
The prototype still encouraged customers to add a standby option to the itineraries they were automatically rebooked on.
We asked ourselves if this option really needed so much priority. Customers were already wary of flying standby. Additionally, if we pushed more people to add themselves to standby lists, wouldn’t we risk overcrowding our gate areas?
After rebooking, customers still had to go through the standard check-in flow.
Why not just automatically check them in?
Users wanted to select seats for their new flights and had trouble finding this option within the check-in flow.
Our first and second prototypes had been created specifically for the legacy kiosks located at airport customer service centers. However, when it came time for us to take our third stab at the redesign, another team at United had just begun a project to innovate the company’s airport gates. (Read about my involvement here.) They envisioned new kiosk devices to be located at each gate, but they weren’t sure what purpose they’d serve. We stepped in, designing our third prototype for the new device’s ratio and with an updated visual look.
In terms of strategy, we decided that since an agent wouldn’t be on hand to answer questions, the kiosk needs to help the user make decisions. In every rebooking case, no matter what, a customer needs to confirm themselves on a flight with available seats. Our new flow makes them do this first, then think about adding a standby option, which is now an advanced feature.
Our third prototype tested well. We needed to make small adjustments to copy, but overall, we were ready to socialize the design across the organization.
Overall, we knew we’d created a more streamlined and valuable tool for customers, and we were excited to implement it. It was scheduled to be released as part of a 2020 “Gate Innovation” proof of concept at O’Hare, with the new rebooking flow then translated to the app, the website, and ultimately, to the customer service centers of the future.
And then the 2020 coronavirus pandemic put our project on hold. The travel industry—and United itself—faced new, urgent challenges. I had to pivot and I had to do it fast. After having invested so much time and thought into United’s kiosks, I began asking myself, “How can we make them touchless?”
