What a mobile website review should actually test

A website can look convincing in a presentation while leaving a visitor unsure what to do. The type is tidy, the photographs fit and the navigation collapses into a familiar icon. Yet a person trying to choose a service may still struggle to find what it includes, whether it suits them or what happens after an enquiry.

For a small business, this is a practical approval problem. The team needs a way to distinguish a page that looks finished from a journey that is ready to use. Our recommendation is to agree one customer task before approving the mobile experience, then observe someone attempting it without the designer narrating every step.

Start with the decision a visitor needs to make

Consider a hypothetical London workshop business. Its mobile homepage opens with an attractive film, followed by a founder story and a gallery. Workshop dates, location and booking details sit much further down the page.

None of those elements is inherently wrong. The question is whether their order supports the intended visit. A returning customer looking for a date has a different information need from someone discovering the business for the first time. A team should make that choice explicit instead of treating desktop order as the automatic mobile order.

Write the task in the visitor's language: “Find a workshop you could attend next month and work out what you need to do to reserve a place.” This connects the review to a real decision. It also makes missing information visible before the discussion becomes a debate about visual taste.

Reflow is one part of the review

W3C's explanation of WCAG 2.2 Reflow describes an accessibility requirement for content to remain available without unnecessary two-dimensional scrolling. For vertically scrolling content, the reference width is 320 CSS pixels, with exceptions for material whose meaning or use requires a two-dimensional layout.

That matters beyond phone screens: enlarging content on a desktop can also reduce the space available to a page. A narrow layout therefore deserves attention as part of accessibility, not merely as a smaller design presentation.

But this check answers a specific question about presentation. It does not establish that the visitor understands an offer or can complete a booking. Keep both questions in the review: can the content be read and operated, and does it give the person enough information to decide?

Give the participant a goal, then listen

The GOV.UK Service Manual's usability-testing guidance, published in 2017, recommends observing actual or likely users attempting relevant tasks. Its advice is to set a believable goal without revealing the route and to invite participants to explain their thinking.

For the workshop example, “tap the menu, choose events, then press book” would mainly test whether someone can follow instructions. A goal-based task lets the team discover whether its labels and sequence make sense to someone outside the project.

Notice the questions that arise. Does the person need the location before choosing a date? Do they interpret “enquire” as a reservation? Are they searching for information that is present but buried? Record the point of uncertainty and what happened next. A vague conclusion such as “mobile needs work” gives the next designer very little to act on.

Review the uncomfortable states

The most attractive screen is often the easiest one to approve. A useful review also asks what happens when a field is incomplete, no date is available or the visitor needs to return to an earlier choice.

These are suggested review scenarios, not findings from a client study. Their value is that each requires the team to define a response. If a booking cannot be completed, the page should explain the situation and the available next step. If an enquiry is submitted, the confirmation should accurately describe what the business will do next.

The same distinction matters in commerce. Our La Amazonas case study describes an online shop. A product page, basket and confirmation each serve a different stage of the purchase. That documented scope is a useful context for review; it is not evidence that we have measured a particular mobile conversion improvement.

Treat a quick check as a starting point

One observed attempt can reveal a problem worth investigating. It cannot establish how common that problem is across the audience, certify accessibility or forecast additional sales. Recruit people relevant to the service, include different access needs and devices, and choose the depth of research according to the consequences of failure.

There is also a balance to strike. Simplifying a mobile page by removing information can make an important decision harder. Before hiding a section, ask whether visitors still have a sensible way to reach what it contains.

For the next website review, bring three things: one customer task, a working mobile journey and a place to record observed difficulties. Resolve the most consequential uncertainty, then repeat the task. Approval becomes more useful when everyone can explain what the visitor is now able to do.

Editorial perspective, 16 September 2026. Sources checked on this date. The workshop scenario and review recommendations are illustrative; no client performance results are claimed.