Case Study: Navigation, Cognitive & Usability Research for a Public-Facing Digital Service

Overview

The client was a large public sector organisation building a new public-facing digital platform for searching, saving, and acting on listings.

To reduce the risk of shipping a hard-to-navigate service, research ran in two phases:

• Phase One (pre-build): we tested whether the proposed information architecture and navigation matched how people naturally understood the content. 
• Phase Two (pre-launch): we evaluated accessibility and usability with people with diverse disabilities and access needs.

Approach

We took a two-phased, mixed-methods research design approach, which included hybrid card sorting and tree testing. 

Hybrid card sorting is a research method where participants group and categorise content to help understand how they naturally expect information to be organised. Tree testing is a research method where participants complete tasks using a proposed navigation structure to test how easily they can find information.

The approach proposed was in two phases:

• Phase One: hybrid card sort (95 cards / 15 participants); tree test (11 tasks / 10 participants); cognitive walkthrough (5 participants)
• Phase Two: 5 moderated remote usability sessions with assistive-technology users, paired with expert WCAG 2.2 AA review

Every finding was rated on a 4-point severity scale - critical, high, medium, low - for consistent prioritisation. Based upon this prioritisation, we created a severity-rated, prioritised recommendation roadmap.

Challenges

• No evidence on whether proposed navigation matched user mental models
• Diverse cognitive, sensory & motor access needs required real testing
• Limited development time before launch demanded clear prioritisation

A challenge for some clients can arise from the use of automated accessibility testing tools only. In situations where the platform technically passes automated accessibility checks, research needs to be undertaken to see if it still create real friction for users.

Outcomes

Phase One: 33% card sort accuracy and 68% tree test success pinpointed clear information architecture, weak spots, especially one major content category needing a full rework.

Phase Two: 18 findings surfaced (1 critical, 2 high, 14 medium, 2 low), including an inaccessible screen-reader dropdown that fully blocked independent task completion.

Combined roadmap: fix the critical blocker; unify "my activity" and confirmation journeys; restructure weak information architecture categories; reduce visual & cognitive load; validate with a full WCAG 2.2 AA audit before launch.

Achievements

30 participants engaged across both research phases

18 usability & accessibility findings identified and prioritised

1 critical accessibility blocker caught before public launch

A clear, evidence-based IA restructuring plan

Why This Matters

Structure that makes sense internally does not always make sense to the people using it - and issues that pass automated checks often only surface when real people, with real access needs, try to use the real thing. Testing before build and after launch catches different problems; together they show whether a public service is genuinely usable by everyone it's meant to serve.

This case study has been generalised to protect confidentiality. Client identity and identifying details have been removed or altered.