Federal Government
Anonymised · NDA
Improving how citizens find their way to applying for a federal benefit
The problem
The information was the barrier
Our team owned the information experience — the pages citizens read to learn how to apply, not the form itself. Over years of contributions from many teams, the guidance had become fragmented and hard to follow. Three core problems drove people away before they ever started an application.
Three problems identified
Overwhelming information
The volume and reading level of the guidance exceeded what many users could process — especially first-time applicants and those with cognitive disabilities.
Looping behaviour
Analytics showed people bouncing back and forth through the same pages without making progress — a clear signal that the path forward was not visible.
Unclear prerequisites
It wasn't clear what documents were needed before starting — people began the process, stalled, and left to search for information they should have had upfront.
Research & discovery
Mapping where people got lost
Before proposing solutions, we built evidence through three sources. The same problems kept surfacing across all of them — confirming these were systemic information issues, not isolated page problems.
Research methods
Journey mapping
Mapping the end-to-end experience exposed where people stalled — unsure which benefit applied, whether they qualified, or what they needed before starting.
Prior research review
Reviewing past research from other teams on the same programmes surfaced the same recurring problems — systemic information issues, not isolated to one page.
Citizen journey map — the information experience before redesign
Usability testing
55% → 97% — what the sessions revealed
We ran multiple rounds of moderated testing before and after the redesign across three cohorts: two distinct applicant cohorts and participants with cognitive disabilities. Testing all three mattered — each surfaced friction the others missed, and the cognitive-disability cohort exposed comprehension barriers that standard testing alone would have overlooked.
What the sessions found
Information overload
Volume and reading level were the dominant barrier — people couldn't quickly find what applied to them, so they read everything and still left unsure.
Back-and-forth navigation
People bounced between pages hunting for a required form or eligibility detail — because it wasn't where they expected it, the next step was rarely clear.
Unclear prerequisites
People didn't know which documents they needed before beginning. Many started, hit a wall, and left to find them — and not all came back.
Cognitive accessibility
Participants with cognitive disabilities faced heightened barriers — dense reading levels, no clear path, and the lack of a visible starting action all compounded.
The redesign
A clearer, more linear path through the information
Working within the established government design system, the redesign focused on three principles. The before/after diagram and representative wireframes below show the structural shift — from scattered, volume-heavy pages to a focused, step-by-step information flow.
Information structure — before and after the redesign
Anonymized under NDA
The information architecture diagram has been removed to comply with NDA requirements. Full documentation — including the before/after structure — is available on request.
Representative wireframes — program overview, eligibility screening, and pathway selection
Anonymized under NDA
Representative wireframes have been removed to comply with NDA requirements. The full wireframe set — covering program overview, eligibility screening, and pathway selection — is available on request.
Prerequisites upfront
Required documents and eligibility criteria were surfaced at the start of each benefit section, so people could prepare before beginning rather than discover blockers mid-journey.
Outcomes
Designing for clarity, not just compliance
What changed — and what we learned
The 55% to 97% task completion improvement was measured in moderated testing. In quarterly programme reviews attended by around 300 stakeholders, findings were presented screen-by-screen, ranked by severity, paired with a recommendation, and backed by session recordings. Prototypes replaced written descriptions as the primary communication tool — stakeholders could see the problem and the solution in the same sitting. Given more time, I would track completion through the live site in production, run a larger dedicated cohort for cognitive-disability testing, and revisit the language at the start of the benefit hub where the highest drop-off occurred.
