Website Not Converting? Ten Checks to Run Before You Pay for a Redesign

If your website is not converting, the cause is more often something other than the design: broken tracking, enquiries that never arrive, the wrong visitors, or an unclear first screen. A redesign fixes none of those, so run the ten checks below before you pay for one. Each takes an afternoon at most, uses free tools, and says plainly when a redesign would not help. If you want a second pair of eyes on the results, our conversion rate optimization work starts from exactly this kind of diagnosis. We make no promise that fixing any of these will lift your results; the aim is to find the real problem before you spend money on the wrong one.

Decide what "converting" means before you test anything

Pick one action that counts: a submitted enquiry, a completed purchase, a booked call. Write it down. Without a single defined action, every metric looks like a problem or none does. Then note roughly how many people reach the page where that action happens. If only a handful of people visit your site, conversion rate is mostly noise, since one extra enquiry swings the percentage wildly. In that case the first job is bringing in relevant visitors, not tuning the page they land on, and checks 1 to 3 below matter more than any layout change.

The order that follows is a judgement, not a measured ranking. It reflects what we most often see go wrong on small-business sites, from broken plumbing at the top to harder commercial questions at the bottom. Work down the list and stop when you find a failure, fix it, and measure again.

Ten checks, most likely cause first

1. Your measurement is wrong

You cannot diagnose a conversion problem if the conversion is not being recorded. Google Analytics calls an important action a key event: Google defines it as an event measuring an action that is particularly important to the success of your business, and any collected event can be marked as one. Many sites never mark anything, or mark an event that fires on page load rather than on a real submission.

Test it free: open Google Analytics, enable debug mode through Tag Assistant, open DebugView, then complete your own enquiry or test purchase and watch whether the event appears with the right name. Google notes that events will not show if privacy controls or consent mode block them, so check that too.

A redesign will not help if the numbers you are using to judge the site were never reliable.

2. Enquiries or orders are failing silently

This is the check people skip because it feels too basic. Forms can break after a plugin update. Notification emails can land in a spam folder, or be sent from an address your mail provider rejects. Spam submissions can bury the real ones until nobody reads the inbox. Checkout can fail on one payment method and work on the others.

Test it free: submit your own form from a phone and a laptop, using an address that is not on your domain. Confirm the message arrives, in the inbox, within a sensible time. Place a test order through each payment option you offer, then check that confirmation emails reach you and the customer.

A redesign will not help if a form that fails today will fail again in a new layout. Checks 1 and 2 come first because they can make everything else misleading: a site with perfect copy and a dead form looks like a conversion failure and is really a delivery failure.

3. The wrong people are arriving

A page can only convert visitors who wanted what it offers. If your traffic comes from searches that are only loosely related to your service, from social posts that attract curiosity, or from a campaign aimed at the wrong audience, conversion will look poor however good the page is.

Test it free: in Google Search Console, the Performance report groups clicks and impressions by Queries, Pages, Countries and Devices. Read the queries that bring people to your key page, as a buyer would. Then compare traffic sources in Analytics. Ask whether a person typing those queries would plausibly want to buy from you.

A redesign will not help if the visitors are simply not buyers. The fix is in targeting, page topics or campaigns.

4. The first screen does not say what you do

If the top of the page offers a slogan instead of a plain statement of what you sell, who it is for and where you operate, people leave without reading further.

Test it free: show your page to someone who does not know the business for a few seconds, take it away and ask three questions: what does this company do, who is it for, what would you do next? If answers are wrong or blank, you have found the problem.

A redesign may help, but a rewrite of the headline and first paragraph comes first and is far cheaper to try.

5. The mobile page is awkward to use

Some of your visitors will arrive on a phone, and the Devices tab in Search Console shows how many. Text that is hard to read, buttons too close together, a sticky banner covering the form, or a menu that hides the call to action all push people away.

Test it free: use your own phone and go through the whole journey to the conversion action. Then open Chrome DevTools Device Mode, which simulates mobile viewports and can throttle network and CPU. Chrome's own documentation calls it a first-order approximation and says you are not running your code on a real device, so treat it as a supplement to a real phone, never a replacement.

A redesign may help if the problems are structural.

6. The page is slow where real visitors are

Speed is judged by what real visitors experience, not by your own fast office connection. Web.dev defines three Core Web Vitals with "good" targets: Largest Contentful Paint within 2.5 seconds, Interaction to Next Paint of 200 milliseconds or less, and Cumulative Layout Shift of 0.1 or less, each judged at the 75th percentile of page loads.

Test it free: run the page through PageSpeed Insights and look at the field data section first. Google explains that field data reflects real users while lab data is a simulated load on one device, and that the two can differ. Use the lab results for debugging and the field results for judging. Our post on how speed affects SEO, CRO and revenue goes deeper on why this matters.

A redesign will not necessarily help: a new design built on the same heavy images and plugins will be just as slow.

7. The form or checkout asks for too much

Every field is a small cost for the visitor. Asking for a phone number, a company size, a budget and a long message before anyone has spoken to you stops casual enquirers. In checkout, forced account creation and unexpected extra steps do the same.

Test it free: count the fields and ask of each one, do we need this to reply? Check that browsers can autofill them. Web.dev explains that correct autocomplete values help browsers fill the right saved data into each field. Then complete the form yourself on a phone and note every moment of friction.

A redesign will not help if the new layout keeps the same long form.

8. Trust is missing at the moment of decision

Visitors need a reason to believe you before they give you details or money. What matters is what sits next to the button: who you are, where the company is registered, what you have done, how to reach a person, and what happens after they submit. Fabricated reviews and badges do harm here, so only show proof you can back up.

Test it free: read your page as a stranger would. Is there a real company name, a contact route, a clear process, policies for returns or refunds where relevant, and proof you can point to? Is the evidence close to the call to action or hidden three pages away?

A redesign may help in placing proof well, but it cannot supply proof you do not have.

9. The next step is unclear or competes with others

A page with five equal buttons has no primary action. "Learn more", "Contact", "Get a quote", "Download" and "Follow us" split attention, and visitors often choose none of them.

Test it free: on each key page, circle the one action you most want and then list every other clickable thing in the first screen. If the main action is not visually dominant, simplify. Check the wording too: "Submit" tells nobody what happens next.

A redesign may help, though a tidier layout of the existing page usually does the job first.

10. The offer itself is not competitive

This is the hard one, and it comes last because it is the least about the website. If your delivery terms, product range, service scope or reputation compare badly with the alternatives, no layout will compensate.

Test it free: put three competitor pages next to yours and compare what a buyer would weigh: what is offered, on what terms, with what proof and how easy it is to start. Read your recent customer emails and any objections you hear in calls. If people say the same thing repeatedly, it is feedback about the offer.

A redesign will not help. The fix is commercial.

Reading your results

When a redesign really is the answer

Everything above argues for cheaper fixes first. Here is where that position is wrong. If checks 4, 5, 7 and 9 all fail, and the cause is that the site is built on a structure that cannot be adapted, for example a dated template with no usable mobile layout, no way to edit key sections and plugins that conflict with each other, then patching is false economy. In that situation a website redesign is justified, and the results of these checks become the brief: they tell the designer what the new site must fix, instead of leaving the project to start from taste.

A redesign is also reasonable when the site no longer reflects what you sell, or when the business has changed direction. What you should avoid is paying for one as a first reaction to disappointing numbers.

If you want help reading the results

If you have run the checks and want help reading the results, send your findings, or just your site address and the action you want visitors to take, through the contact page and ask for a written quote. Website Prism Ltd is a registered England and Wales company, and the work is delivered remotely by a team in Pakistan. We do not promise outcomes, and we will tell you if a check points away from anything we would build.