Conversion-rate optimisation is the discipline of getting more of the traffic you already have to do the thing you want them to do. Most of it is folklore — "use red buttons", "remove the navigation", "add a countdown timer" — and most of the folklore does not survive contact with real testing. This guide is the field manual we use across UK SMB CRO engagements, with the eleven moves that consistently lift conversion and the experimentation framework that separates signal from noise.
Baseline benchmarks by sector
Knowing where you start matters more than knowing where you want to end. UK SMB conversion-rate benchmarks for 2026 traffic, measured on website-to-primary-conversion: trades (quote-form lead) 4-9%; restaurants (booking) 6-12%; professional services (consultation booking) 2-6%; e-commerce (purchase) 1.5-3.5%; SaaS (free-trial signup) 3-7%; B2B services (qualified lead form) 1-3%. Sites well below the bottom of the range almost always have foundational problems; sites at the top have done the work below systematically.
1. Above-the-fold clarity — pass the 5-second test
A visitor lands on the homepage. Five seconds. Can they answer three questions without scrolling? What does this business do, who is it for, what should I do next? Most homepages fail at least one of the three. The fix: a single H1 stating what you do specifically (not "we transform businesses" — "same-day website launches from £499"), a single subhead clarifying the audience and the offer, a single dominant CTA. Everything else below the fold.
2. Sub-2-second LCP on mobile
A 4-second LCP loses 30-40% of mobile visitors before the hero finishes painting. The conversion-rate cost is not subtle and the fix is well-documented (image budget, font subsetting, no render-blocking third-party scripts in the critical path, edge-CDN hosting). Tuning LCP from 3.5s to 1.8s on a mobile-heavy site typically lifts conversion by 15-25% without any other change. See the website speed optimisation guide for the full playbook.
3. Mobile thumb-friendly CTAs
CTAs sized for desktop mouse precision do not work on a thumb at arm’s length on a moving train. The mobile button minimum: 48 pixels tall, 80% of the available content width, contrast ratio passing WCAG AA against the surrounding surface. Sticky CTAs that follow the scroll on mobile (without covering content) lift conversion by 20-40% on most SMB sites because the conversion path is always one tap away regardless of where the user is on the page.
4. Three-or-fewer form fields on high-intent paths
Every additional form field reduces submission rate by roughly 8-12% on quote-form and lead-gen paths. The right count depends on intent: high-intent emergency-service paths (emergency plumber, locksmith) work best at 2 fields (postcode + phone); considered service paths (financial advice, architectural commission) tolerate 4-6 fields because the visitor expects to provide more context; e-commerce checkout works at 8-12 fields because the visitor has already decided. Match field count to intent rather than to internal convenience.
5. Trust signals next to CTAs
A CTA without a trust signal converts at roughly 60% of the rate of the same CTA with a trust signal beside it. The right signals depend on the audience but reliably include: aggregate rating with review count ("4.8 from 132 Google reviews"), accreditation badge (Gas Safe number, RIBA Chartered Practice, FCA FRN, NHS-registered), guarantee statement (money-back if not delivered, fixed price, no hidden fees), client logo strip for B2B. Place them adjacent to the CTA, not buried in the footer.
6. Social proof above the CTA
Testimonials placed below the fold get read by 8-15% of visitors; the same testimonials placed above the primary CTA get read by 60-75%. Lead with the strongest named testimonial. Specific outcomes beat generic praise — "doubled our booking volume in three weeks" outperforms "great service" by a wide margin. Photo of the named customer outperforms an anonymous quote by 40-60% on conversion-rate impact.
7. Transparent pricing where you can
Pricing transparency filters unsuitable prospects out before they enquire and accelerates conversion for suitable ones. The mistake most service businesses make is hiding pricing entirely under the assumption that "we need to qualify the lead first". Specific price ranges, banded pricing by package level, or clear from-prices all outperform "available on enquiry" by 25-50% on conversion rate. The visitors who are price-sensitive self-select out before wasting your time; the visitors who can afford you self-select in with intent.
8. Friction-free phone — tap-to-call
For trades, emergency services, hospitality, healthcare and any sector where calls are a meaningful share of conversions: the phone number is a tel: link, large, visible above the fold, tappable on mobile. PNG phone numbers fail this test. Phone numbers buried in the footer fail this test. The lift from getting this right on a phone-led conversion path is typically 30-60% of additional call volume against no change in traffic.
9. The right exit-intent intervention (most are wrong)
Exit-intent popups have a bad reputation because most of them are bad. A "wait! 10% off!" popup on a £25 product converts at 0.5-1.5% and annoys 98%. The right pattern: a relevant secondary offer triggered by exit on visitors who have shown intent (scrolled past 60% of the page, viewed three pages, dwelt longer than 90 seconds) rather than every visitor. The offer must be genuinely useful — a downloadable guide, a free consultation, a price-match commitment — not a generic discount. Done correctly, this lifts conversion 1-4 percentage points without the brand damage.
10. Urgency signals that are honest
Honest urgency works (genuine deadlines, real stock counts, scheduled launches). Dishonest urgency (fake countdown timers, fabricated stock-low warnings, "X people viewing this page right now" widgets that are not real) backfires once the visitor notices, and increasingly the visitor notices. Use urgency where it is genuine — "8 spaces left for August intake", "build window for next Friday closing in 6 hours" — and skip it where it would have to be fabricated.
11. Post-conversion confirmation that earns the next action
The confirmation page after a form submission or purchase is one of the highest-attention pages on the site (the visitor just committed) and one of the most wasted. Most confirmation pages say "thanks, we will be in touch" and end the conversation. The opportunity: name the next concrete step, set expectations on timing, offer the second valuable action (download the guide, follow the social channel, book a deeper conversation, share the offer with a colleague). Done well, the confirmation page lifts repeat conversion by 8-15% over six months.
The experimentation framework
CRO advice that is not tested is folklore. The framework: form a specific hypothesis ("removing the third form field will lift submission rate"), define the metric in advance (form submissions per unique session), define the significance threshold (95% confidence on a one-tailed test), run for enough volume to reach significance (typically 1,000+ conversions per variant on a 5-percentage-point lift), document the result. Most SMB sites do not have the traffic for true A/B testing on every change; for low-traffic sites the better framework is a structured roadmap of high-confidence changes shipped sequentially with before/after measurement over 28-day rolling windows.
What CRO is not
CRO is not "make the homepage prettier", though a coherent visual system helps. CRO is not "add more content", though clearly written content helps. CRO is not "rebuild the entire site", though sometimes the foundation is the problem. CRO is the systematic removal of friction and the insertion of trust at the moments the visitor needs them. Done well, it is invisible — the visitor does not notice the changes; they just convert at higher rates than they used to.