Skip to main content

Practice makes perfect: evolving our user research for GOV.UK

Posted by: , Posted on: - Categories: Case studies, Tips and techniques

The GOV.UK Finding Things team is committed to improving the navigation across GOV.UK.

We’ve already created a user-centered single taxonomy and improved our navigation design patterns.

In this post, I’ll explain how we’ve changed how we do our research cycles to better support the evolving needs of the project and team.

Starting with contextual research

We decided to run a pilot on a small part of GOV.UK and selected the theme of education. We narrowed our focus and selected early years. Our priority was to conduct discovery-style contextual research. This is research done in context of a specific location or activity, such as the user’s home or at work.

There are often a lot of unknowns at this stage. Ours were very broad: who are the users of education content on GOV.UK? How are we meeting their needs? How much of the content do they use?

Doing contextual research

To answer those questions, colleagues from the Department for Education helped us identify groups of users that early years content is aimed at. We interviewed 13 people from these groups including childminders, nursery managers, and pre-school managers.

During these interviews, we discovered another set of users: those who help users understand government guidance. This led to a second wave of research with 15 users, including childminding agencies, local authorities, and journalists.

The next step was to create a draft taxonomy, and use remote card sorting and tree testing to try it out with a larger number of users - 70 for the card sort and 65 for the tree testing.

Our research then moved on to people who work in schools. This gave us 24 additional perspectives from teachers, business managers, governors, and career advisers.

Original research process contextual research, card sorting, tree testing and more contextual research

This research provided valuable insights into how users think, feel and behave.

What we learned from the process

By the end of the second research round, we’d answered the unknowns we identified at the start of the research. We then moved on to think about how research could help us refine the taxonomy and our navigation patterns.

We examined the process and determined:

  • The time between research rounds didn’t let us readjust quickly - particularly when we discovered the new set of early years users during research
  • A month of contextual research with a final analysis session meant the team had to wait until the end for in-depth insights
  • The findings began to be repeated between user types. This was a good indication to us to move onto a different kind of research with different questions

A new approach

We ran a retrospective with the Finding Things team to look at how well our research questions, methods and findings had supported the team. We’d discovered a lot about the context of our users’ working lives, but we still needed to get to the details that would help us iterate the taxonomy.

We held a team workshop to gather outstanding questions about users’ behaviours, needs, language, and environment. Those questions were then prioritised by our information architect and designers.

The next challenge was to put together a research approach that would help us meet this goal and make improvements.

Streamlining interview research

Using all of this experience and input, we designed a research cycle to address the evolving needs of the project and team. Here is how it will work:

  1. The recruiting agency will have 2 weeks to recruit the first round of participants
  2. The first round of research will last a week. We will talk to 5 users of the same type. The recruiting agency will recruit for the second round concurrently
  3. The team will conduct a group analysis of the first round
  4. The team moves on to the second round of research and the agency will recruit for the third round of recruitment and so on until we have spoken to all remaining user types

New research process recruitment, interviews week one, group analysis week one to two, interviews one week and group analysis one to two weeks

We will also include tree testing during the interview sessions, so we can iterate our draft education taxonomy throughout the research cycle.

This will increase:

  • pace - while we’re doing our current round of research, our recruiting agencies will be finding people for the next round
  • flexibility - we’ll be able to adjust our recruitment criteria for future rounds and iterate our approach between rounds
  • collaboration - our team and other colleagues will be involved in the analysis of each round providing richer and more immediate insights

What’s next?

Now we’ve done the groundwork for this approach, it’s time to implement it. Follow our blog for updates.

Without the Department for Education, its agencies and the GOV.UK Content Team we wouldn’t have made so much progress. Thank you!

Keep in touch. Sign up to email updates from this blog. Follow Emma on Twitter.

Sharing and comments

Share this page