UX design examples: look for the decision behind the design
Useful UX design examples show how a person completes a task, not just how an interface looks. A polished screenshot rarely explains what happens when information is missing, someone changes their mind or a request fails. Start by asking what the person needs to do and what uncertainty the design removes.
The examples below analyse documented GOV.UK Design System patterns. They are not Project Hero client projects or claims about measured commercial results. The business applications are hypothetical ways to explore the same principles. Treat each as a starting point for research, not a template that will automatically improve conversion.
Example one: review information before submission
The GOV.UK check answers pattern gives people a chance to review information before sending it. It displays answers with routes to change them. The design addresses a specific problem: someone may notice a mistake only when they see the whole submission together, rather than one question at a time.
A hypothetical training provider could explore this approach before a customer submits a course booking request. Show the chosen course, date and attendee details, with clear edit routes. Make the final action explicit so the customer knows whether they are requesting a place or paying for one. Test whether people understand the difference. A summary that conceals the commitment would miss the point of the pattern.
Example two: make errors actionable
The GOV.UK error summary component brings validation problems together and links them to the relevant fields. This combines an overview with a route to repair. It is more useful than an unexplained message saying that the form could not be submitted, especially when the problem is below the visible part of the page.
Imagine a business enquiry form asking for a project budget. If the format is invalid, explain the expected format and preserve the other answers. Do not blame the person for a server outage or describe that outage as invalid input. In testing, watch whether participants can find the problem and recover without help. The result to assess is successful correction, not whether the warning looks prominent.
Example three: adapt established patterns carefully
GOV.UK’s service manual recommends considering existing patterns, then adapting or creating them when user needs require it. This is a useful counterweight to redesigning every control merely to appear distinctive. Familiar interaction can let visual identity live in the surrounding presentation without changing how every task works.
For a hypothetical equipment-hire site, a familiar set of labelled filters may be a better starting point than an unexplained visual selector. But the actual needs still matter: visitors may compare sizes, availability or collection locations. Identify the information needed to choose, sketch alternatives and test the uncertain parts. Do not copy an entire government service interface simply because one of its patterns is well documented.
Test the whole journey, including the awkward states
GOV.UK describes moderated usability testing as observing participants attempting specific tasks. This helps reveal whether language, layout or interaction prevents completion. Use actual or likely users and tasks that match the decisions you need to make. Avoid showing a design and asking only whether people like it.
For the booking example, ask someone to find a suitable session, change the date and explain what will happen after submission. Include an unavailable date or a missing answer where relevant. Note hesitation, wrong turns and recovery alongside completion. Obtain appropriate consent for research and use safe sample information. A test does not need real payment details or confidential customer records to reveal a confusing journey.
Choose an example that matches your problem
Before adopting any design, write down the problem it is intended to solve. “People cannot tell whether their request was sent” is more actionable than “the page needs modern UX”. Decide what evidence would show improvement and what new friction you might introduce. An extra review page, for example, should earn its place by helping with a meaningful decision.
Keep a small decision record: the current problem, the proposed change, the pattern that informed it and the observations from testing. Revisit it after release with support feedback and suitable usage evidence. Good examples help you ask better questions. They cannot replace understanding your own audience, and they do not justify invented claims about higher sales or fewer errors.