ATM Improvements | KeyBank
Summary
KeyBank’s ATM team saw its Customer Satisfaction (CSAT) scores drop from 68% to 32% after the launch of new ATM software on existing hardware. The team agreed to try my experience design team’s user-centered approach. My team interviewed 25 people to understand the ATM user and created composite personas and scenarios around how each persona used the ATM on a typical day. The team then had a virtual design workshop to ideate solutions that would align with the user’s needs. They identified two problem areas: a confusing bill denomination screen and old hardware with non-functioning buttons. We built and tested a user interface that they compared against the vendor's interface and tested the time on task for the bill denomination. We also installed mirrored plates over the non-functioning buttons to create a sense of safety for customers. The fascia plates tested successfully and were installed across all of the KeyBank ATMs, saving KeyBank over $2.8 million.
“Design is not an expensive new overhead but something that companies are already engaged in and could almost certainly do more cost-effectively.”
The Challenge
With the launch of new ATM software on existing hardware, KeyBank saw its Customer Satisfaction (CSAT) scores around ATM usage drop from 68% to 32% in just a few months. The ATM team had opted to use its tight budget to upgrade the software (the interface and screens) before upgrading the hardware (the ATM terminal itself.) In their view, they were incrementally improving the customer experience. In reality, they’d upended it.
My experience design team connected with the channel manager for ATM to see if we could help. The ATM team worked with company that manufactures and installs self-service transaction systems to procure both the hardware and software for the ATM, so at first they were concerned that our contribution would be limited. After all, we were a product design team. We couldn’t re-code the vendor software, nor could we build better physical machines. So how could an experience design team have an impact?
Nevertheless, in the face of their swiftly dropping CSAT scores, and the limited budget they had to fix their problems, the team agreed to try our user-centered approach.
Research
The first order of business was to understand the ATM user. As the ATM channel managers told us, “[ …the user was ] everyone. There’s no specific demographic profile, or customer type, that uses ATMs. Anyone can use one. Even people who are not our customers.” While the ATM team stated this expecting that this made our research harder, it actually made it much easier. Because in design research, it is not about who is using a product. It is about how they’re using it. What they need it to do, what they expect it to do, and what they value about it. And it meant that we could enlist the perspectives of anyone who uses ATMs. Which is a lot of people.
My team interviewed 25 people about their experience using ATMs and three themes emerged: safety, speed and convenience. We created composite personas to humanize the experience, and scenarios around how each persona used the ATM on a typical day. And we had a virtual design workshop with the ATM team, branch professionals, marketing, technology and leadership.
Ideation
In the virtual workshop, it was evident that many of the cross-functional team members already could identify the problems that had plunged the CSAT scores. And they even had ideas for how to solve them. But focusing on the personas and scenarios helped them to think more deeply about the experience of the customer and ensure their ideas aligned with the user’s needs.
For example, one problem area was the bill denomination screen. The current interface was confusing, requiring users to designate the number of bills they wanted of each type, but defaulting to the lowest bill available. That meant it was very easy to accidentally withdraw $300 in $5 increments. As one team had a scenario of a person driving through with their child to get $20, and another team had a scenario for a person who was getting out $1400 to buy a piece of furniture, it enabled the team to think through the interface design from both perspectives and not just err on the side of generic controls for each.
The other problem area identified had to do with juxtaposition of new software with old hardware. The new software was a touchscreen interface. The old hardware had buttons on each side. In the workshop, branch professionals said they had been covering the buttons with tape in order to alert customers that they weren’t functional. But in the workshop as we talked about the pillars of safety, speed and convenience, one of my team members proposed that we use mirrored plates to cover the old buttons, which would reinforce a feeling of safety for the customers and prevent customers from seeing the old, non-functioning buttons.
Prototyping & Testing
ATM hardware (before) with buttons on the side
As mentioned above, we couldn’t re-code the vendor software. But we could very easily and cheaply build and test a user interface that we could compare against theirs, and point them in the direction we wanted to go. So that is what we did. We simulated the experience of using ATM screens on a tablet and usability tested this experience with 10 users to ensure that it was straightforward. But since we were also interested in the core value of speed, we tested time on task for bill denomination to ensure that our interface was faster.
In addition, a star designer on my team created a dozen pairs of inexpensive mirrored fascia plates. We sent them to pre-selected branches with installation kits that included: instructions on how to mount them, the mirrored plates, and a weather-proof adhesive. We asked the branch team to provide us with feedback on the plates. We wanted to ensure that the solution worked before rolling it across the entire “fleet” of ATMs.
Impact
The fascia plates tested successfully and were installed across all of the KeyBank ATMs. In disguising the buttons on the old hardware, we prolonged the company’s original investment and saved KeyBank over $2.8 million (the cost of the fascia plates vs the cost to upgrade the entire fleet to new machines.)
ATM hardware (after) with mirrored plates covering the buttons
Our prototype for the bill denomination screens was given to the vendor partner and they implemented that design not just for KeyBank ATMs, but for other clients as well. They didn’t charge us for the design of the software interface.
In August of 2022, the KeyBank ATM project won a Bank Customer Experience Summit Award, a panel that recognizes innovation in the banking sphere.
Additional design and research by Ian Liao, Samantha Gustafson and Aaron Simmons
In accordance with my non-disclosure agreement with KeyBank, I have omitted confidential information from this case study. All information in this case study is expressed by me, and does not necessarily reflect the views of KeyBank or any of its employees.