Why Work with a Research Partner for Your Website Redesign?

Because the people who commission a website are rarely the people who use it. A research partner brings those users into the room at the three points where redesigns go wrong: the brief, the structure, and the weeks before launch.

Here's what that looks like at each stage.

Before design: check the brief

The partner finds out who actually uses the site, what they came to do, and where they give up. That means reading the analytics, then talking to real users to understand why the numbers look the way they do.

Analytics tell you where people leave. Only people can tell you why.

During design: test the structure first

Most redesigns start with page designs. The bigger risk sits underneath, in how content is grouped and labelled. A good partner tests that structure with users before anyone designs a page. Then they test key pages and tasks as the design takes shape, and feed what they learn straight back to the designers.

Before launch: prove it works

The last round of testing checks the real thing, on real devices, with the tasks that matter most to the business. It's the cheapest point to catch a problem that would otherwise go live.

What it looked like at the University of Melbourne

The university's research websites were hard to navigate. Students, academics and industry partners couldn't find what they needed.

Our research and design team worked inside the university for more than nine months. We mapped current and future journeys, built personas and a clear value proposition, and ran one-to-one testing with 16 users. That shaped three new platforms: Graduate Research, Find an Expert, and Research Commercialisation.

The result was more student enquiries and applications, more connections through Find an Expert, and stronger industry interest in research partnerships.

How to evaluate a research partner for a redesign

• Ask to see a redesign where research changed the brief. If it never has, they're confirming, not researching.
• Ask who will run the sessions. It should be senior people, not a junior team reading a script.
• Ask what you'll get at each stage. A tested structure and clear design decisions, not a long report.

Pitfalls to avoid

• Relying only on analytics, without talking to users.
• Researching once at the start and never testing again.
• Testing only with people who already love the site.

For why skipping this step costs more than doing it, read why skipping research in a redesign costs more.

Related: Behavioural research and strategy and the University of Melbourne case study.