Design Task - Check if you can get Legal Aid

The problem

The 'Check if you can get legal aid' service does something important - it helps people in often difficult, stressful circumstances understand whether they can access legal support. But important services still have gaps, and this one had several. Nobody had formally audited it against usability or accessibility standards. Some issues were immediately visible to a trained eye. Others only surfaced through tooling. The task was to assess the service honestly, identify what was working, what wasn't, and propose a clear path to improvement.

What I did about the problem

I started with Jakob's Ten Usability Heuristics as a structured lens for the assessment - not as a checklist to rush through, but as a framework to ensure I was looking at the service systematically rather than just reacting to what caught my eye. I identified genuine strengths: the 'Before you start' content sets clear expectations early, the plain language throughout is well-judged, and the 'hide this page' feature for vulnerable users navigating domestic abuse situations is a thoughtful, human-centred design decision - one being actively developed in collaboration with DWP and Digital Scotland across the wider GDS community.

Then I looked at where the service fell short. Using WAVE, the web accessibility evaluation tool, I identified contrast errors in the navigation menu - low contrast between text and background that would affect users with low vision. I found that the session timeout warning was buried in inset text at the bottom of the page - easy to miss, and potentially frustrating for users who lose their progress. I flagged accessibility concerns with the modal dialog pattern, drawing on my own project experience of a front-end developer discovering modal issues during an audit - in that case the solution was moving the content to a new page rather than risking excluding users. And I identified that 16 single-column cards for legal problem categories created unnecessary scrolling, when a grid layout - already evidenced by HMRC's tax pattern work - would make the content more scannable.

I also conducted a competitor analysis across legal aid services in Ontario, South Africa, New South Wales, and California - not to copy what they did, but to ask the right questions: how easy is it to check eligibility? How many steps does it take? Is the criteria clear? What can we learn?

For next steps I proposed using a Lean UX canvas with the team, stakeholders, and users to test my assumptions - because all of this was desk research and heuristic evaluation without any direct user insight. I would then prioritise using a hypothesis prioritisation canvas to decide what to test, build, or discard, before moving into ideation with 'how might we' questions and low-fidelity sketching, through to usability testing.

What came out the other side

The assessment produced a prioritised set of evidence-based recommendations - from quick accessibility wins like contrast fixes and heading structure, to more considered changes like the modal dialog approach and card layout. Importantly, I was clear throughout that these were informed assumptions, not validated findings - and that the Lean UX canvas and usability testing were the essential next steps to confirm what was actually worth solving.

What I learned

Heuristic evaluation is most valuable when you're honest about its limits. You can spot a lot by looking carefully and knowing your frameworks - but the moment you put real users in front of a prototype, you find out what actually matters to them. The assessment is the start of the process, not the end.