This is going back to the days just prior to Universal Credits and long before the GDS guidelines were established, when the benefits system was so convoluted, so paper-based, so office reliant that it was in danger of collapse. The move online was both necessary and of major importance and had been going for sometime by the time I arrived and the processes involved were well established so any challenge to them was regarded with a mix of fear and annoyance.
The UX designers were not involved with the usability testing, strange as it seems now that wasn’t that unusual even five years ago, so I started by refining the existing wireframes into a series of lengthy hi-fi prototypes, heavily annotated to show where experience told me the system would fail. You didn’t need to go far down the heuristic route to see where the disconnect between rigid adherence to offline operating patterns and online interactive patterns were going to occur. Repeat user testing consistently showed a correlation with the results I had predicted and slowly changes to the system were made before the designs got anywhere near testing.
I began to work more closely with the Civil Service’s SMEs, modifying copy, challenging use cases and user stories and pushing back on real-life user journeys, explaining where and why user behavior would require a slightly different approach in a digital environment. In turn they were incredibly informative, approachable and passionate about their work and taught me a huge amount in return. The old cliché about jobsworth civil servants was proven wrong every day.
Although overly complex for what they needed to be the documents we ended up producing did facilitate the change of approach that was needed and did make the final UX better.
The accessibility of the site proved to be less than desirable, making basic but easily rectified mistakes, such as CTAs only being available halfway up the page and not easily navigable by screen readers. I ran reports and tests and the results changed the way the site was coded and laid out. This in turn was carried on through to the Directgov site.
My final act was to hand over the much improved and much better documented system to a new team of designers.
What would I change in retrospect?
I should have insisted on being involved with the usability tests, reliance on heuristics and personal experience wasn’t ideal. I did leave the project in a better place than when I joined it and I’ve since seen a lot of the engagement and interaction patterns I introduced then in other public sector systems now so I’m still glad to have been involved.