What is conversion rate optimization (CRO)?

What CRO is, in plain terms

A conversion rate is the share of visits that end in the action you want. If 1,000 people visit your quote page and 30 send a request, the rate is 3%. Conversion rate optimization is the work of raising that share on purpose.

Google Ads Help defines conversion rate as the average number of conversions per ad interaction, shown as a percentage. Its example is 50 conversions from 1,000 interactions, which gives 5%. The same arithmetic applies to any page. The next guide, conversion rate math, walks through it in full.

CRO is not a design style or a bag of tricks. It is a loop: observe how people really use the site, form a theory about why they stall, change one thing, and check whether the result moved.

Why it matters to a business owner

Traffic costs money or time. You pay for it through ads, or you earn it through months of SEO and content. A site that converts poorly wastes a share of every pound or dollar spent getting people there.

Take a simple case. A service site gets 2,000 visits a month and 40 enquiries, a 2% rate. If the page changes lift that to 2.5%, you get 50 enquiries from the same traffic. The numbers here are arithmetic, not a forecast, and no one can promise that lift. They show why a small percentage change matters at scale.

CRO also makes other channels cheaper to run. When a landing page converts better, each ad click is worth more, so the same budget buys more enquiries. That is why we build ads and pages together.

How CRO works: the loop

Good CRO follows the same five steps every time. Skipping the first two is the most common reason programs fail.

• Measure: confirm that your tracking records the right actions, including micro-conversions such as starting a form, and set a baseline conversion rate per page and per traffic source.
• Research: find out why visitors do not convert, using analytics, heatmaps, session recordings, surveys and testing with real people, plus a quick heuristic evaluation. See CRO research methods.
• Prioritise: list the problems found, and fix the ones that affect the most visitors at the most important steps first.
• Change and test: make the change, and where traffic allows, run an A/B test so you learn whether it caused the difference.
• Learn: record the result, including the losses, and use it to choose the next change.

What actually gets changed

The changes that matter are usually unglamorous. Nielsen Norman Group first reported in 1997 that 79% of test users scanned each new page and only 16% read word by word. Pages written for scanning, with clear headings and plain language, performed better in its tests than promotional copy.

In practice, the common fixes fall into a short list: a clearer headline and offer, a call to action that says what happens next, fewer form fields, proof placed near the decision, faster mobile loading, and message match between the ad or link that brought the visitor and the page they land on, which keeps the information scent strong.

Baymard Institute, which studies checkout usability, aggregates 50 studies on cart abandonment. As of its September 2025 update, the average documented abandonment rate is 70.22%. Its research lists extra costs shown late, forced account creation and a long checkout among the reasons shoppers give. Many of those are fixable.

What is a good conversion rate?

There is no single good number. Unbounce, a landing page software vendor, reported in its 2024 Conversion Benchmark Report a median of 6.6% across industries, from over 41,000 landing pages. It also reported industry medians ranging from 3.8% to 12.3%.

Treat that as context, not a target. A vendor sample is not your market. Rates change with traffic source, price, offer, and what you count as a conversion. A page selling a 20,000-pound service converts far less often than a free guide, and that is normal.

The honest benchmark is your own history. Compare this quarter with the last one, by traffic source and device, and judge a change against that.

When CRO goes wrong

CRO fails in predictable ways. Tests are stopped before they reach statistical significance when a result looks good, tracking counts the wrong events, or a site with a few hundred visits a month runs tests it cannot finish. Each one produces confident but false conclusions.

Not every idea wins. Ron Kohavi and Stefan Thomke, writing in Harvard Business Review in 2017, described how one low-priority change to ad headlines on Bing raised revenue by about 12%. Their wider point was that teams routinely misjudge which ideas matter, which is why testing beats opinion.

Low-traffic sites have a different route. When there are too few visitors for a valid test, you rely on research and sound judgment, make bigger changes, and compare before and after with care. It is less certain, and you should say so.

How to start this month

Pick one page that matters, usually the one where paid or organic visitors land before they enquire or buy. Record its visits and conversions for the last 90 days, split by source.

Then read the research methods guide, watch 20 session recordings of that page, and write down every place a visitor hesitated. Compare the headline to what your top ad or search result promises, using message match as the test.

Change the clearest problem first. Record the date, the change and the result. That log is the beginning of a real CRO program.