A CRO audit that tells you what to fix first, and what to test.

A CRO audit starts by checking that the numbers are right. We compare orders in analytics with orders in the store, then build the funnel in a GA4 funnel exploration: sessions, product views, add to cart, checkout started and purchase, split by device, channel and landing page. If the e-commerce events don’t fire correctly, fixing them is the first item on the list.

The funnel shows where people leave; it doesn’t show why. For that we watch session recordings and heatmaps in Microsoft Clarity (free, and quick to install), read customer service emails and reviews, and look at on-site search terms. Rage clicks, dead clicks and quick exits from the same element point at the parts of a page that confuse people.

Then a person goes through the store on a real phone and a desktop, as a shopper would: landing from an ad, browsing a collection, choosing a size, adding to cart and checking out with each payment method. We check each step against the 10 usability heuristics from Nielsen Norman Group, such as visibility of system status and error prevention.

Checkout gets the closest look. Baymard Institute’s research on cart abandonment lists extra costs, forced account creation, delivery that’s too slow and a checkout that’s too long among the main reasons shoppers leave. We check when shipping costs appear, whether guest checkout is offered, how many fields are required and which payment options show up first on a phone.

Speed is part of conversion. We measure product and collection templates against Google’s Core Web Vitals thresholds (Largest Contentful Paint within 2.5 seconds, Interaction to Next Paint under 200 milliseconds, layout shift of 0.1 or less) and name the apps and scripts that slow them down or make buttons move.

Everything goes into one ranked list with two kinds of items. Clear fixes are things that are plainly broken or missing, such as a hidden returns policy or a broken size guide; those should simply be done. Test ideas are changes where the result isn’t obvious; those go into a test plan with a hypothesis, the metric, the pages and whether your traffic is enough to read a result.

This audit comes from the same work behind our conversion rate optimization service. Kashif develops Shopify themes and apps, so recommendations are written so that a developer can build them, with notes on what can be done in the theme, what needs an app, and what Shopify doesn’t allow.

Each test idea in a CRO audit comes with a rough sample check before it goes into the plan. We look at how many orders the page gets in 28 days and at the change you would need to see; a test runs for at least 14 days, two full weekly cycles, so weekday and weekend shoppers are both counted. Where traffic is too low for that, the idea moves to the fix list or waits until a sale period brings more visitors, rather than being called on a few days of noise.

Every item on the list carries its evidence: a screenshot or short recording, the funnel step it affects, the device where we saw it and the share of sessions that reach that template. Each item is also rated for effort (hours, days or a full sprint) and for how many visitors it touches. So a broken size guide on your best-selling collection sits above a cosmetic issue on a page with 50 visits a month, and your developer can start at the top without a second meeting to decide what matters. The list is shared as a sheet you keep, so your team can mark items done and track what is left.