Guide
Conversion rate optimisation guide
Last updated:
What does conversion rate optimisation actually mean?
Conversion rate optimisation is the practice of improving the percentage of visitors who complete a valuable action — an enquiry, a booking, an add-to-cart, a signup — through evidence, not guesswork. It is a diagnosis-then-test discipline: find where real visitors are dropping off, form a specific hypothesis about why, then test the change rather than assuming a redesign will fix it.
CRO is not the same job as a redesign. A redesign changes structure and templates; CRO works within a living page, usually with much smaller, targeted changes — a clearer headline, a shorter form, a stronger proof point above the fold — measured against a defined outcome.
How do you find out where visitors are actually dropping off?
Start with the numbers you already have: analytics funnels, scroll depth, form abandonment and page-level bounce rates tell you where people leave, not why. Session recordings and heatmaps fill in some of the why — confusing navigation, a form field nobody wants to fill in, a call to action people scroll straight past.
The most reliable signal is usually the simplest one: read your own enquiry forms and sales call notes for the objections and questions that come up repeatedly. If prospects consistently ask a question your page never answers, that is a conversion leak analytics alone will not show you.
What should you test first?
Test the change most likely to move the outcome that actually matters — usually the primary call to action, the clarity of the offer above the fold, or friction in the enquiry/checkout path — before cosmetic changes like colour or imagery. Big, structural hypotheses (does removing a form field increase completions? does a clearer headline reduce bounce?) tend to produce clearer results than small visual tweaks.
Prioritise pages with real traffic and a defined goal. Testing a page that gets a handful of visitors a month wastes time on statistical noise instead of a real signal.
How much traffic do you need before testing makes sense?
This is the honest limit most CRO pitches skip: formal A/B testing needs enough visitors and enough conversions per variant to reach statistical confidence, which usually means thousands of monthly sessions to a specific page, not dozens. Below that volume, structured qualitative review — heatmaps, session recordings, direct user feedback, expert heuristic review — produces more reliable improvements than a "test" that will never reach significance.
Lower-traffic sites still benefit from CRO thinking; the method just shifts from statistical testing to evidence-based iteration: change one specific thing, watch the qualitative and directional numbers for a defined period, and keep or revert based on the evidence you actually have.
What's the difference between CRO and a redesign?
CRO diagnoses and improves within a living page or funnel. A redesign changes the underlying structure and templates. Many businesses jump straight to a redesign when the real problem is one confusing step in an otherwise sound page — diagnosing first is usually cheaper and faster than rebuilding first and hoping it helps.
