Key takeways:
Start with discovery research. As the client tells you about the need, you’re going to start designing something in your head. When you see the current product, you’re going to start designing something in your head. When you have your first user interview, you’re going to start designing something in your head. Do all the contextual discovery research you can before you start the design.
Start with a small, end-to-end feature. This tool is big. It does so many things. It would be impossible to redesign it all at once, so we started small and chose two feature that had a distinct workflow within the larger system. This allowed us to prove the viability of our approach and design to the client, as well as gain insights into the larger workings of the system to make the redesign of the entire system easier.
Research early and often. I love when a project has the time and access to users to allow for several rounds of usability testing at key decision points in the design. We showed users low-fidelity wires, high-fidelity mockups, a fully interactive prototype, and learned important things at every step that changed the design for the better.
UEGroup was hired to redesign an existing software tool that tracks and enables the maintenance and logistics supply chain for a high-tech, next-generation military aircraft. Every part on this aircraft must be tracked, including engine hours, flight hours, takeoffs and landings, and using weapons. To accomplish this, all parts are tracked not only by part number but also with a manufacturer's code and serial number for that unique part. All parts have unique requirements for when they must be inspected and replaced, and the system lets the maintenance crew know when that should happen. All maintenance actions must be tracked in the system and the system must approve of all actions and parts before the aircraft can be released for flight. To add one final complication, this aircraft is sold to many nations which all have a unique terminology and requirements for what makes and aircraft ready to fly.
The existing software meets all of the above requirements, but is considered to be a hindrance rather than a helper. UEGroup requested to do a complete redesign rather than modify the current tool. The redesign began with extensive discovery research and visiting the people who use the current tool. Our discovery research tracked what the different maintenance and logistics specialists do on a daily basis, as well as periodic long-term requirements. We charted what happens in the physical world, in the software, and on paper. We tracked the flow of information and authority from start to finish and created diagrams to map both what happens today with the current software and what essentials would need to happen in a world agnostic of the current software.
Armed with a thorough understanding of what information, roles, and authorities are needed, we moved into the design phase. We created several initial possible designs that were intentionally very different. After discussing the pros and cons of the designs, we narrowed in a main design but kept a record of the other designs to use at later points when we faced specific interactions that needed special treatment. We conducted several rounds of usability testing with users of the current system. The participants represented as many demographics of the user base as we could find. The initial usability tests asked participants how they would complete tasks with our designs, what they expected to happen if they clicked certain icons, where they expected to find certain pieces of information, and how they liked the new design.
Several rounds of usability testing helped us refine our concept enough to invest in an interactive prototype that would allow us to perform a final round of in-depth usability testing where the participants had enormous freedom to use the software however they wanted. For this final round of usability testing, we wrote a testing plan that included tasks, measured time, and asked questions about what was expected and how the participants like the design. When we performed the actual test, participants found the prototype so intuitive and easy to use that many participants broke our test plan by jumping in and using the tool as they really would, not allowing us to time ask questions as they went. Participants not only used the tool, they enjoyed using it and were thrilled with the new direction of the tool. Their satisfaction is why this was the project I am most proud of in my career as a UX designer.