Most websites do not have a traffic problem. They have a problem converting the traffic they already get, and the cause is usually sitting in the design.
Why design decides conversion rate more than copy does
The common assumption is that conversion is a writing problem: sharpen the headline, tighten the offer, add urgency. Copy matters, but it only gets read if the design lets it be read. A visitor arriving on a UK small-business website makes a rapid, largely unconscious assessment before absorbing a single sentence: does this load, does it look legitimate, can I see what this company does, and can I find the next step. Fail any of those and the copy never gets a hearing. This is why two sites selling identical services at identical prices can convert at wildly different rates, and why a redesign that only changes the words tends to produce nothing. Design is not decoration sitting on top of the message. It is the delivery mechanism for the message, and when it fails, it fails silently. Nobody emails to say your form was too long. They just leave.
Six design decisions that actually move conversion rate
These are the levers that show up repeatedly when we audit underperforming UK sites, roughly in order of how often they turn out to be the binding constraint.
Page speed and Core Web Vitals
Speed is the most reliable conversion lever because it affects every visitor before anything else can. Google's Core Web Vitals give concrete targets: Largest Contentful Paint under 2.5 seconds, Interaction to Next Paint under 200 milliseconds, and Cumulative Layout Shift below 0.1. Missing these on mobile is the single most common defect we find.
Mobile layout, not just mobile compatibility
Most UK service businesses now take the majority of their traffic on phones, yet many sites are merely mobile compatible rather than mobile designed. Tap targets sat too close together, key information pushed below three scrolls of hero imagery, and phone numbers that are not tappable all cost enquiries.
Visual hierarchy and scannability
People scan before they read. If every element competes for attention with the same weight, nothing gets attention. Clear heading structure, short paragraphs, real whitespace and a single obvious focal point per screen let a visitor find their answer without effort.
Form length and friction
Every additional field measurably reduces completions. Ask only for what you genuinely need to make the next contact, use sensible input types so mobile keyboards behave, validate inline rather than on submit, and never make a user re-enter data they have already given you.
Trust signals in the right place
Reviews, accreditations, client logos, real photographs of real people and a visible UK address all reduce the perceived risk of enquiring. Placement matters as much as presence: a trust badge in the footer does nothing, whereas the same badge beside the submit button works on the exact hesitation it is meant to answer.
Calls to action that name the action
Vague buttons underperform specific ones. "Get a free quote" or "Book a 15-minute call" tell the visitor what happens next, whereas "Submit" or "Learn more" ask them to guess. One primary action per page, repeated at natural decision points, beats five competing options.
What a realistic conversion rate looks like
Benchmarks get quoted carelessly, so treat any single industry-wide figure with suspicion. What matters more is measuring the right thing consistently. A few practical points:
- Define the conversion honestly. An enquiry form submission, a phone call and a newsletter signup are not the same event and should not sit in the same number.
- Segment by device. Mobile and desktop conversion rates frequently differ by a factor of two or more. A blended figure hides which one is broken.
- Segment by traffic source. Branded search converts far better than cold display traffic. A falling blended rate often just means the traffic mix changed.
- Compare against yourself. Your own trend, measured the same way month to month, is more useful than any published benchmark.
The number to chase is not an industry average. It is a rate that is improving against your own baseline while the quality of enquiries holds up.
How to find your own conversion leaks
You can do most of this diagnostic yourself in an afternoon:
- Run your key pages through a speed test on mobile. Test the actual landing pages people arrive on, not just the homepage. If you are on WordPress, our guide to WordPress speed optimisation covers the usual culprits.
- Load your own site on your phone, on mobile data. Not on office wifi. Try to complete an enquiry as a stranger would, and note every point where you hesitate.
- Check where people drop out. In your analytics, find the step in the journey where the biggest proportion of visitors leaves. That is where to spend the budget.
- Count your form fields. Then remove every one you do not need to make first contact. This is the cheapest conversion win available to most businesses.
- Ask someone outside the business to use the site. Five minutes of watching a real person struggle beats a month of internal debate. Say nothing while they do it.
- Work through a structured review. Our website audit checklist covers the technical and structural items that are easy to miss.
Accessibility is a conversion issue, not just a compliance one
Accessible design gets treated as a legal box to tick, which undersells it. Low-contrast text, tiny fonts and forms that cannot be navigated by keyboard exclude people who were ready to buy. The W3C accessibility guidelines set a minimum contrast ratio of 4.5:1 for normal body text, and a surprising number of otherwise smart-looking UK sites fail it because a designer chose light grey on white. The same fixes that make a site usable for someone with a visual impairment make it easier for everyone reading on a phone in daylight, which is most of your traffic.
Where design stops and conversion rate optimisation begins
Design fixes the obvious defects. Conversion rate optimisation is the ongoing, evidence-led process of testing changes and keeping the ones that work. The two get conflated, and the distinction matters commercially: design work is a project with an end, whereas optimisation is a programme that compounds. For most SMBs the correct sequence is to fix the structural problems first, because there is no point A/B testing button colours on a page that takes nine seconds to load. Once the fundamentals are sound, structured testing is where the remaining gains live, and our conversion rate optimisation playbook sets out how to run that properly without a large testing budget.
Landing pages deserve separate treatment, because paid traffic behaves differently from organic and needs a tighter page with fewer exits. We cover that in our guide to landing page design.
Redesign mistakes that quietly cost conversions
Redesigns are the most common way a business accidentally reduces its own conversion rate. The recurring causes:
- No baseline. If you did not record conversion rate by device and source before the rebuild, you cannot tell whether the new site is better or worse. Record it first.
- Aesthetics decided in a boardroom. Internal preference is not evidence. The people choosing the design are rarely the people who have to use it under pressure on a phone.
- Heavier pages. Full-width video headers, oversized imagery and a stack of new scripts routinely make the new site slower than the old one, which quietly undoes every other improvement.
- Lost content and broken URLs. Dropping pages that ranked, or changing URLs without redirects, removes the traffic that was converting in the first place.
- Changing everything at once. When twenty things change together and the rate falls, you cannot tell which one caused it. Responsive behaviour in particular is worth verifying carefully, as we set out in our piece on mobile friendly website design.
If you would rather not run this yourself, this is exactly the work our web design and conversion rate optimisation services are built around.
Frequently asked questions
How does web design affect conversion rate?
Design controls whether a visitor can load the page, understand what you offer and complete the action, all before your copy gets read. Speed, mobile layout, visual hierarchy, form length, trust signals and call-to-action clarity are the six levers that account for most of the difference between sites with similar traffic and very different results.
What is a good conversion rate for a UK business website?
There is no single credible figure, because it depends heavily on what you count as a conversion, the sector, and whether traffic is branded or cold. A lead-generation service site converting enquiry forms behaves nothing like an ecommerce checkout. Measure your own rate consistently, segmented by device and source, and judge progress against that baseline.
Is UX design the same as conversion rate optimisation?
No. UX design is the discipline of making a site usable and coherent. Conversion rate optimisation is the ongoing process of testing specific changes and keeping what measurably works. Good UX usually raises conversion, but it is not driven by the same evidence loop, and treating them as interchangeable leads to testing trivia while structural problems go unfixed.
Does page speed really change conversion rate that much?
It is the most consistent lever we see, mainly because it affects every visitor and compounds with everything else. A slow page loses people before any other design decision gets the chance to matter. Google's Core Web Vitals thresholds give a concrete target to work against rather than a vague instruction to be faster.
How long before design changes show up in conversion data?
You need enough conversions to distinguish a real change from noise, which for most SMB sites means several weeks rather than several days. Judging a change after a fortnight of low traffic is how businesses talk themselves into reversing improvements. Fix obvious defects immediately, but hold your judgement on the numbers.
Can a redesign make conversion rate worse?
Frequently, yes. The usual causes are a slower site, lost pages that were ranking, URLs changed without redirects, and design decisions made on internal taste rather than evidence. Record your baseline before you start, keep the pages that were performing, and change things in stages you can actually attribute.
