Card Sorting

Card sorting is a user research technique where you write content items, features, or topics onto individual cards — physical or digital — and ask participants to group them in ways that make sense to them. The result is a window into how your audience naturally categorises information, which is often quite different from how your organisation structures it internally.

There are three main variants. In an open card sort, participants create their own group names, revealing their natural categories. In a closed card sort, you provide pre-defined categories and participants place cards into them, which tests whether your proposed structure works. A hybrid approach combines both — some fixed categories with the freedom to create new ones.

The beauty of card sorting is its simplicity. You don’t need a prototype, a finished design, or even a clear information architecture — in fact, the whole point is to let users shape it before you commit. Sessions typically take 15–30 minutes, and with digital tools, you can run them remotely at scale.

Analysing results involves looking for agreement patterns — which cards consistently end up together, and which category labels participants naturally reach for. Dendrograms and similarity matrices help visualise these clusters, but even a simple spreadsheet analysis can reveal powerful insights.

Why it matters

Information architecture built on organisational thinking rather than user thinking is one of the most common reasons products feel confusing. Card sorting bridges this gap directly. It takes the guesswork out of navigation structure and replaces it with evidence of how your users’ mental models actually work.

It’s also remarkably cost-effective. A card sort with 15–20 participants can give you high-confidence data about content structure, and the whole process — from setup to analysis — can happen in a single week.

In practice

• A university redesigned its website navigation using card sorting with prospective students. They discovered that students didn’t think in terms of “faculties” — they thought in terms of careers and interests. The resulting information architecture organised courses by outcome rather than department.

• A healthcare app ran a closed card sort to test whether patients could find symptoms, medications, and appointment features under the navigation labels the team had proposed. When 40% of participants placed “Prescription refills” under “Appointments” instead of “Medications,” the team restructured accordingly.

• An intranet team used a hybrid card sort with employees from different departments. The results revealed that the HR, IT, and Finance teams each had completely different mental models for the same content, leading to a multi-path navigation approach.