A business is receiving fewer enquiries from its website. The homepage looks dated, a competitor has launched a more modern site, and a redesign feels like the obvious response.
But fewer enquiries can have very different causes. Search traffic may have changed. A contact form may no longer submit reliably on mobile devices. Visitors may not understand the offer. Or the reporting may be wrong while the actual number of enquiries is unchanged. Those situations call for different decisions—and only some justify redesigning the website.
The question worth answering first is not “Does the website look old?” It is “What is stopping the website from producing the result the business needs?”
First, establish what has actually changed
A decline in website performance should be defined in business terms. For a service company, that might mean fewer qualified enquiries; for an online shop, fewer completed purchases; for a booking service, fewer confirmed reservations. A click on a contact button is not the same as an enquiry received, and an enquiry is not automatically a suitable sales opportunity.
The distinction matters when using analytics. In Google Analytics 4 (opens in a new tab), a key event measures an action selected as important to the business. It does not independently confirm that a form was delivered, a booking was accepted, or a lead was qualified. Compare recorded events with the systems that receive the result: the form backend, sales inbox, CRM, order records, or booking platform.
Then check whether the comparison is fair. Traffic source, seasonality, advertising spend, device mix, and changes in how the business defines a qualified lead can all alter the apparent trend. Low volumes can also make short-term percentages misleading. Before recommending a large project, we need to know whether the underlying business result has genuinely changed.
One symptom can support several explanations
Imagine the same initial complaint in five different situations: “The website is bringing us fewer enquiries.” The table illustrates why a single symptom is not yet a diagnosis. Each pattern suggests a direction for investigation; none proves its cause on its own.
| Observed pattern | Plausible explanation | Next useful check |
|---|---|---|
| Analytics events fall; received enquiries do not | Tracking or consent measurement may have changed | Compare backend records with event data; inspect tagging |
| Received enquiries and relevant search visits both fall | Acquisition, demand, or search visibility may have changed | Compare comparable periods, queries, sources, and lead rates |
| Received enquiries fall; visits steady; mobile form fails | An interaction or submission defect may be blocking users | Reproduce failure and verify server acceptance |
| Visits steady; forms work; users misunderstand the service | Offer clarity or page usability may be limiting action | Observe representative tasks and test clearer content |
| Multiple journeys are confusing; structure resists fixes | A structural redesign may be justified | Map critical journeys and compare repair versus rebuild costs |
The first two patterns could leave the page design largely irrelevant. The third points to a specific technical failure. The fourth may call for better content or a narrower UX change. Only the final pattern begins to make a structural redesign a strong candidate—and even then the cost of fixing the current experience should be compared with rebuilding it.
Work through the evidence before choosing the solution
Consider a hypothetical service business whose analytics show a drop in contact-form submissions. The initial temptation is to improve the homepage’s appearance and make the enquiry button more prominent.
The first check is whether the business actually received fewer enquiries. If the received and verified enquiries are stable, but the analytics event count has fallen, the likely investigation is the measurement setup: tagging, consent behavior, event firing, or a changed form implementation. A redesign would not be supported by that evidence.
Suppose instead that confirmed enquiries have fallen. The next step is to look at relevant visits and completed journeys by device and landing page. In this illustrative case, visits to the important service pages are broadly comparable with the earlier period, but the decline is concentrated on mobile.
A direct task check then reproduces a validation error that prevents the mobile form from completing. That is much stronger evidence than a general conversion-rate decline: the failure is observable, tied to the affected journey, and can be retested after a fix. The sensible first intervention is to correct the form and verify that test submissions are both received and measured. There is still no case for redesigning the homepage.
What if enquiries remain low after the defect is repaired? Then the original explanation was incomplete. Recheck the traffic mix, service-page content, user tasks, and other points where visitors may leave. The purpose of the diagnosis is to select and challenge an explanation, not to defend the first one indefinitely.
Look for evidence that could change your mind
Good website analysis should be willing to disprove its own hypothesis. If we suspect visitors cannot find the contact action, observe whether representative users can locate it. If they can, the problem may lie in trust, offer clarity, qualification, or what happens after the click. If we suspect slow performance, examine real-user measurements and reproduce the affected interaction rather than treating one laboratory score as a verdict.
Google Search Console (opens in a new tab) can help investigate search visibility through queries, impressions, clicks, and indexing information. It cannot explain whether an enquiry was qualified. Core Web Vitals (opens in a new tab) help describe loading, responsiveness, and visual stability; they do not determine whether the service proposition makes sense. Different evidence sources answer different questions, so one metric should not be asked to explain the entire customer journey.
Sometimes the outcome is uncertainty rather than a definitive cause. When data quality is weak or traffic volume is small, a focused task review, instrumented test, or limited change may produce more useful evidence than a premature promise about what a redesign will achieve.
Choose between repair, improvement, and redesign
A targeted repair is appropriate when a reproducible fault interrupts an otherwise sound journey: a submission error, broken navigation, inaccessible control, or configuration problem. A focused improvement is appropriate when the structure works but users struggle with one offer, page, or step. Those interventions are easier to attribute and often easier to reverse.
A broader redesign becomes credible when the problems are connected and structural. Examples include important services that no longer fit the site’s information architecture, recurring confusion across multiple journeys, layouts that cannot reasonably support accessible or mobile use, or a platform whose constraints make every meaningful improvement costly and inconsistent.
The key is the comparison, not a universal preference for smaller work. If several isolated fixes would leave the central problem intact, a redesign may be the more responsible investment. Conversely, a new visual design is hard to justify when the evidence points to a faulty integration or an acquisition problem.
Turn the findings into a decision someone can verify
A website review should produce more than a list of recommendations. For the hypothetical mobile-form problem above, the decision can be recorded in a way that makes both the reasoning and the next check clear:
| Decision element | Illustrative record |
|---|---|
| Problem | Confirmed enquiries down, especially on mobile |
| Evidence | Relevant visits broadly comparable; mobile form error reproduced |
| Immediate action | Repair and test validation and submission; no homepage redesign yet |
| Success check | Submissions received and recorded; monitor qualified leads by device over a comparable period |
| Reconsider if | The fix works technically but confirmed enquiries remain low, or new evidence points to another cause |
This record is useful even if the final recommendation is a full redesign. It distinguishes the business outcome from a visual preference, shows what is known and what is only suspected, and sets a condition for revisiting the decision. A client can then challenge the evidence or agree on the scope without having to accept an unexplained recommendation.
Preserve the parts of the website that already work
If a redesign is justified, the work still needs a baseline. Identify which pages bring relevant visitors, which contact paths are working, what content supports purchasing decisions, and how important actions are measured. Otherwise, the new website may be more attractive while quietly losing useful capabilities.
Search migration is one example. If URLs change, Google’s site-move guidance (opens in a new tab) recommends mapping old URLs to their new destinations, using appropriate redirects, and monitoring the transition. If URLs do not change, the same migration plan is unnecessary, although technical, accessibility, content, and conversion regressions still need attention. Our technical SEO guide covers related search foundations in more detail.
When the immediate issue is a contained defect or reliability problem, maintenance and technical support may be more appropriate than a rebuild. The recommendation should follow the diagnosed need, not the size of the possible project.
The right redesign starts with the right diagnosis
A website can be visually dated and still complete its main tasks well. It can also look modern while losing enquiries through a broken form, weak content, irrelevant traffic, or misleading measurement.
Before paying to replace the interface, establish the business outcome, verify the evidence, test competing explanations, and choose the change that addresses the most credible cause. Sometimes that will be a repair. Sometimes it will be a series of improvements. And sometimes it will be a full redesign. What matters is being able to explain why.
