Mara logo

Usability Testing

Watching real users attempt real tasks in a product or prototype to find where they struggle, without steering them.

Process & Methods

The method is simple and easy to ruin. You give a participant a task framed by their goal — 'send £50 to a friend' — not by your interface — 'use the transfer screen' — and then you keep quiet. The moment you explain something, you have replaced the data you came for with your own knowledge. Five to eight participants per audience surfaces most of the significant problems.

It answers a different question from A/B testing. Testing tells you which version wins; usability testing tells you why people fail, which is what you need in order to have a better version to test. It also works before anything is built, on a prototype, which is when findings are cheapest to act on.

In practice

A fintech signup tested with six users. All six paused at a field labelled 'Legal entity name', unsure whether to enter their own name. Before: the team assumed the drop-off was form length. After relabelling to 'Company name (as registered)' with an example, completion rose 22%. The catch — the field was legally required to stay, so only the label could change.

Learn more

You may ask

Frequently Asked Questions

How many users do we need to test with?

Five to eight per distinct audience uncovers the large majority of serious issues. Testing more of the same audience mostly re-finds the same problems; if you have two very different user types, test both groups separately.

Can we test with colleagues?

Only for a smoke test of the script. Colleagues know the product and the vocabulary, so they cannot reproduce the confusion you are looking for. Any real finding needs people outside the company.

Related terms

All terms

Let's talk about your product.

Happy to look at what you're building and say where design would move the needle.

Contact Us