CRO audit: how to run it step by step

A CRO audit is an analysis of a website or online store that shows where and why users abandon a purchase or an enquiry. It is the first stage of conversion rate optimisation (CRO). It combines quantitative data from analytics with qualitative data, e.g. from session recordings and surveys. The result is a list of issues ordered by their impact on sales, plus hypotheses for A/B tests. An audit does not increase traffic, but it shows how to make better use of the traffic the site already has.

What does a CRO audit cover?

A CRO audit covers the whole user journey: from landing on the site to a purchase or a submitted form. It checks data, user behaviour, content, technology and trust.

A CRO audit differs from an SEO audit. An SEO audit looks for obstacles to winning search traffic. A CRO audit looks for obstacles that stop that traffic from turning into customers.

How do you run a CRO audit step by step?

A CRO audit runs in six steps: from the business goal, through the data, to a list of hypotheses for testing.

• Set the goal and the main conversion. In a store this is usually a purchase; in a service business, a form submission or a phone call. Also write down micro-conversions, e.g. add to cart or newsletter sign-up.
• Check that tracking is correct. Compare the number of orders in GA4 with the number of orders in the store system. Without reliable data, every later conclusion will be wrong. Missing events can be added in Google Tag Manager.
• Analyse the funnel. Find the step where most people drop off. Split the data by device, traffic source and new versus returning users.
• Find out why users drop off. Quantitative data shows where the problem is, and qualitative data shows why. Session recordings, heatmaps, short on-site surveys and user tests all help.
• Do an expert review. Go through the purchase path on a phone and on a computer as a customer would. Check the clarity of the offer, shipping costs, forms, error messages and trust elements.
• Write hypotheses and set priorities. Describe each issue as a hypothesis: what to change, for whom and what effect you expect. Then order the list by impact and effort.

A good audit report does not end with a list of errors. It shows which changes to implement right away and which to check first in an A/B test.

Which tools should you use and how do you set priorities?

Four groups of tools are enough for a CRO audit: analytics, behaviour research, technical tests and surveys. Combining their results matters more than the number of tools.

Page speed is measured by PageSpeed Insights and Google Lighthouse.

Priorities are most often set with the PIE method, developed by Chris Goward of the agency Conversion. Each issue is scored in three categories:

• Potential: how much improvement you can expect on a given page,
• Importance: how valuable the traffic reaching that page is,
• Ease: how hard the test is to prepare, technically and organisationally.

A popular alternative is the ICE method: impact, confidence and ease. Whatever the method, you first test changes on pages with high traffic and a clear problem. The effect is best converted into revenue or ROI to justify further tests.

Which issues does a CRO audit find most often?

Stores lose the most customers in the cart and at checkout. According to Baymard Institute, the average cart abandonment rate is 70.22%, based on 50 studies from 2006–2025.

Baymard estimates that a better checkout design can raise a large store's conversion rate by 35.26% on average. Typical issues found in an audit are:

• shipping costs shown only at the end of checkout,
• mandatory account creation before purchase,
• too many form fields and unclear error messages,
• slow page loading on phones,
• no reviews, company details or clear return policy,
• a landing page that does not match the ad's promise,
• misconfigured conversion tracking that distorts the conclusions.

Trust is built, among other things, by customer reviews and shopper photos, i.e. user-generated content (UGC). In the audit it is worth checking whether they are visible near the buy button.

The free UX audit checks the product page, cart and checkout on mobile, and the free GA4 audit checks event tracking in your store. What a rebuild costs is covered in how much a website costs, and we handle design and build as part of our websites service.

What sources is this article based on?

Data on cart abandonment, reasons for abandonment and the possible conversion uplift comes from Baymard Institute's “Cart Abandonment Rate” overview. The 70.22% average is based on 50 studies from 2006–2025, and the reasons for abandonment come from a survey of US shoppers.

The PIE method and its author are described by Conversion, the agency where Chris Goward developed it. Definitions come from the zest.agency glossary.