Conversion Rate Optimisation

The conversion rate is the share of visitors who take the desired action out of all visitors. Conversion rate optimisation (CRO) aims to raise that rate with evidence rather than guesswork: it investigates where users get stuck, forms a hypothesis about why, tests a change and measures the result.

CRO is not just about testing button colours or headlines. The clarity of the offer, the length of forms, page speed, trust signals and the mobile experience all fall within the same process. Because it makes better use of existing traffic before buying more, it is one of the strongest levers in performance marketing, and it cannot be measured correctly without a reliable data layer.

• Research: Finding problems through analytics data, heatmaps, session recordings, surveys and user interviews.
• Hypothesis: A testable statement in the form "this change will improve this metric, for this reason".
• Prioritisation: Choosing which test comes first based on potential impact, confidence and implementation cost.
• Testing: Trying the change in a controlled way with A/B or multivariate tests.
• Implementation and learning: Making the winning variation permanent and carrying the result into the next tests.

Consider a service company's enquiry form. Analytics data shows that a large share of users who open the form abandon it at the phone number field, and a short survey reveals that people are reluctant to receive sales calls.

The hypothesis is: making the phone field optional and explaining how contact will be made will increase the form completion rate. The test compares two variations over the same period with randomly split traffic, and the result is judged not only by form volume but also by the sales team's feedback on enquiry quality.

The basic formula is: number of conversions ÷ number of sessions or users. Which denominator is used should be stated clearly in reports. Alongside the main conversion, intermediate steps such as adding to basket or starting a form are tracked, so it is clear at which step users drop off.

For tests, sufficient sample size and statistical significance are set in advance, and tests are not stopped early. Guardrail metrics such as revenue per visitor, average order value and lead quality confirm that raising the conversion rate has not caused losses elsewhere.

A/B testing is a method for comparing two variations; CRO is the broader process that uses this method among others. A/B tests set up without research often try random ideas and fail to produce meaningful results. In CRO, a test is run to validate the solution to a problem that research has uncovered. On low-traffic sites, user research and direct improvements based on best practice can be more efficient than testing. CRO follows this logic in our performance marketing and mobile experience work.