For most people, your website isn’t a communications channel and it’s not a marketing asset. For most people, your website is your think tank.
It’s where a journalist goes at 5pm on deadline, where a policymaker checks whether your evidence supports their position, where a researcher looks for data and it’s where a donor decides whether to trust you.
If that experience is confusing, slow, or unclear, it doesn’t just affect your website. It affects your credibility. I’m Alice Vulliamy, UX strategist at Soapbox, and user experience is something I think about every day.
The supermarket test
When we run UX workshops for think tanks — most recently at the On Think Tanks School in Barcelona, which has become something of a yearly tradition for us — we often start with a simple analogy: the supermarket:
You walk in → You find the aisle → You locate the section → You pick the product.
That journey feels obvious because it’s well designed.
Think tank websites should work the same way:
You land on the homepage → You find a topic that interests you → You browse the related content → You pick the report you want to read.
If users can’t get from the entrance to the “product” without friction, they’ll leave — or worse, they’ll leave unsure whether you’re authoritative. And for research organisations, authority is everything.
Running this exercise in person brings the analogy to life. Being in the room together creates a different quality of discussion: ideas are tested more openly, assumptions are challenged more quickly and collective thinking sharpens in real time. It’s a reminder that UX is fundamentally about people — and people learn and solve problems best together.
Designing for four very different audiences
The real UX challenge for think tanks isn’t aesthetics. It’s diversity of audience.
Most research organisations are serving at least four core audiences at once:
- Policymakers;
- Researchers;
- Journalists;
- Donors
Each one arrives with a different job or task to do.
- Policymakers need concise, relevant insight and quick routes to expertise;
- Researchers need depth, data and citability;
- Journalists need speed, clarity and contact details;
- Donors need reassurance — who you are, what you stand for, and proof you can deliver.
Good UX doesn’t prioritise one at the expense of the others. Rather, it makes sure each group can complete their task easily.
At Soapbox, we start by turning those needs into simple, testable user stories, for example:
“As a journalist, I want to quickly find relevant research so that I can file before my 7pm deadline.”
If the website can’t support that story, it needs rethinking.
In our in-person workshops we explore this through rough paper prototyping. Working with simple materials allows teams to iterate quickly and focus on structure and journeys rather than visual polish. It’s a useful way to explore the problem space collaboratively before moving into digital design — and it often surfaces structural issues that might otherwise go unnoticed.
It also gives think tankers space to approach a different kind of challenge from their day-to-day policy work: thinking not just about what their ideas are, but how those ideas are encountered and used.
Structure first, design second
One of the biggest mistakes we see is organisations jumping straight to visual design – the colour palettes, fonts and hero images.
But UX lives in structure. Which is why before we design anything, we work through four stages:
Discover → Define → Design → Test
- Discover real user needs
- Define the underlying structure (your sitemap)
- Design prototypes around real journeys
- Test them with actual users.
Only then does visual design start to work properly — because it’s reinforcing clarity, not decorating confusion. And we’ve seen this make a transformative difference time and time again.
When we rebuilt the Institute for Fiscal Studies website, the goal wasn’t simply to modernise the look. It was to help diverse audiences — from journalists to academics — navigate complex economic research with confidence. Structured content, intuitive navigation and clear hierarchies were as important as visual design.
Research organisations publish at scale. UX must reflect that
Think tanks produce large volumes of high-quality content, from reports to briefings, data sets to articles and even events.
That scale creates a UX challenge: without strong hierarchy and metadata, everything becomes noise.
Three fundamentals matter more than ever:
- Machine-readable content: Publish in HTML as well as PDF. Meet accessibility standards. If screen readers can interpret your content, AI tools can too — and so can more of your audience.
- Clear hierarchy: Strong titles. Useful summaries. Descriptive subheadings. This helps humans scan and helps machines interpret.
- Structured metadata: Topics, authors, dates, content types — and semantic markup in the background. It’s the difference between “content on your site” and “discoverable research”.
This is ultimately about making your ideas usable.
UX in the age of AI
We’ve written recently about AI and research communications, so I’ll keep this short.
Two points matter most for UX right now:
First: AI will read your reports in full — even if humans don’t. That makes structure, summaries and metadata more important, not less.
Second: Think tank search has historically been poor. Modern AI-powered search — built on structured content and vector databases — can dramatically improve relevance. It’s one of the clearest quick wins for research organisations today.
But the fundamentals remain unchanged: clarity, hierarchy, accessibility. Ultimately think about like this: good UX for humans is good UX for machines.
What great UX actually delivers
When UX is done well, it means that your website does so much more for you than be a “shop window” for your work. It means your homepage communicates confidence, your navigation communicates competence and your publication pages communicate authority. It means that your entire digital presence reinforces trust, which is strategically critical to helping you achieve your organisational goals.
I think about it like this: your website is where your ideas meet the world. And if your mission is to shape public debate, influence policy, or strengthen research communities, that meeting point matters.
If you’re rethinking your website — or wondering whether your current one truly reflects the quality of your work — we’d love to talk.
And if you’d like to run a structured UX workshop with your own team, we’d be very happy to design and facilitate one with you. Drop us a line.
Ideas matter, but how people experience them matters just as much.