Oli Grimsey-Jones
← All work Home Office · GOV.UK

Bringing user research into the EU Settlement Scheme

The first version of the EU Settlement Scheme online was rushed and shaped by policy more than by users. Applicants found it long and unclear, and identity verification caused real problems. In a user-centred design team, I led research sessions and rapid prototyping to show what a simpler service could look like.

AAWCAG accessibility target for the service
End-to-endprototypes built from GOV.UK React components
[add]e.g. rounds of research or participants
Role
Designer and prototyper in a UCD team
Client
Home Office
Worked with
User researchers, civil servants, policy
Tools
Figma, Miro, GOV.UK React prototype kit
Live
gov.uk/settled-status-eu-citizens-families ↗
Role & ownership

What I owned

I built the prototypes, ran research sessions with researchers, and argued for treating prototypes as disposable, not as codebases to maintain, so the team could change direction quickly.

Problem & constraints

What we were up against

Users

  • Too many screens and unclear information
  • Identity verification was difficult
  • A very broad range of users and devices

Business

  • Reduce friction online
  • Verify identity online to save manual processing costs
  • Meet WCAG AA

Constraints

  • Policy decisions already made
  • Must fit with other government services
  • Dummy data only: no live country lists or passport-scan feedback
Process

01Understanding users

We interviewed people who had used, or would use, the service, then built personas and empathy maps. Device type didn't follow age the way we expected, which changed our assumptions.

02Prototyping the happy path

We mapped the existing service as user flows in Figma for a full view, card-sorted the journey in Miro, and built a first prototype with the GOV.UK React toolkit and dummy data.

User flow of the existing serviceadd image
User flow of the existing service

03Testing with frustrated users

Early sessions were hard: participants were so unhappy with the existing service that anything we showed felt like more of the same. We started giving them five minutes to talk about their experience, recorded it and promised to pass it on. After that, sessions became much more natural.

04Iterating without attachment

After each round, researchers presented findings and I updated the prototype; for big changes we threw it away and started again. We didn't reuse patterns just because other services had them. Users wanted quiet reassurance they were signed in, and a full account dashboard mid-journey would have been jarring.

Prototype iterationsadd image
Prototype iterations
Outcome

Impact

Design decisions grounded in research with real applicants, and a team that prototyped faster. [add any known outcome]

Takeaways

What I'd take to the next project

  • Let people be heard firstFive minutes to vent turned hostile sessions into useful ones.
  • Prototypes are disposableThrowing them away kept the team honest about what users needed.
  • Don't copy patterns by defaultWhat works in one service can be jarring in another.