What is CRO: conversion rate optimisation

CRO is the systematic process of increasing the percentage of users who complete a conversion. A conversion can be a purchase, a sign-up, a download or any action your business considers valuable. Conversion rate is calculated as conversions divided by sessions, multiplied by 100.

What differentiates CRO from "design opinions" is that it relies on data: analytics, heatmaps, session recordings, surveys and A/B tests. You do not change a button because "it looks better in green": you change it because data indicates that a specific change will increase conversions.

If your website converts at 1% and receives 10,000 monthly visits, you have 100 conversions. You can double traffic (expensive, hard) or improve conversion to 2% (same investment, double the result). CRO offers compounding returns: every conversion improvement applies to all future traffic.

Additionally, CRO improves the user experience for all visitors, not just those who convert. A smoother checkout, a clearer form or a better-communicated value proposition benefits the entire digital ecosystem.

CRO follows a continuous cycle: research → hypothesise → prioritise → test → analyse → iterate. The research phase identifies problems (where users drop off, what frustrates them). The hypothesis proposes a specific solution with an expected outcome. The test validates or refutes the hypothesis with data.

Prioritisation is key because resources are limited. The ICE framework (Impact, Confidence, Ease) scores each hypothesis from 1 to 10 on expected impact, confidence in the outcome and ease of implementation. Hypotheses with the highest scores are tested first.

You do not need a dedicated team or expensive tools to start with CRO. There are high-impact optimisations you can implement directly based on widely validated best practices.

These quick wins do not replace the full CRO process — which requires research and testing — but they address the most common issues affecting conversion on most websites.

The fundamental principle of CRO is that decisions are based on evidence, not personal preferences. The CEO may prefer design A, the designer may advocate for B, but only an A/B test with real user data determines which converts more.

This requires an organisational culture that values experimentation over hierarchical opinion. The companies that optimise conversions most effectively are those that let data settle internal design and messaging debates.

CRO requires a minimum traffic volume to work: without enough volume, A/B tests cannot reach statistical significance and results are unreliable. As a reference, you need at least 1,000 conversions per month to test changes with confidence within a reasonable timeframe.

If your traffic is lower, you can apply CRO best practices without A/B tests (optimisations based on qualitative research), work on improving acquisition first, or use tests with intermediate metrics (CTA clicks, scroll depth) that require less volume.