Your conversion rate dropped, or it never climbed, and you have a list of "fixes" a mile long. The problem is you don't actually know which one to make first — so any budget you spend is a guess. A CRO audit fixes that by turning guesswork into a diagnosis.

A CRO audit is a structured, repeatable review of your site that finds the exact points where visitors drop out of your conversion path — before you spend a dollar changing anything. Instead of a random pile of best-practice tweaks, it works in a fixed order: locate the leaking step, then diagnose why that step leaks using behavioral data, then confirm the cause before you touch the design. This guide walks you through that order, step by step, and shows where heatmaps, session recordings, and funnel data slot into each stage.

This is the walkthrough version, not the checklist version. If you want a quick tick-box list to run through later, keep our CRO checklist open in a second tab. What follows is the diagnostic process that tells you which of those boxes actually matter for your site — and in what order to check them.

Why run a CRO audit before you change anything

Most conversion "optimization" fails because it starts with a solution instead of a diagnosis. Someone reads that button color matters, changes a button, and the number doesn't move — because the button was never the leak.

A CRO audit reverses that. You find the leak first, then fix it, which means every hour of design and dev time is aimed at a problem you've confirmed is real. Three things make an audit worth the time:

  • It sequences your work. You stop debating ten ideas at once and follow the funnel to the one step that's bleeding the most.

  • It separates "where" from "why." Analytics tells you a page underperforms. Behavioral data tells you the actual reason. You need both, in that order.

  • It's repeatable. Run the same sequence next quarter and you can compare like for like instead of starting from scratch.

If you're new to the wider discipline, our conversion rate optimization guide covers the strategy this audit plugs into. The audit is the diagnostic engine; the guide is the map.

What you need before you start

You can't diagnose a leak you can't see. Before Step 1, get three things in place so the audit produces evidence instead of opinions.

  • A defined conversion path. Name the steps that lead to your primary goal — for an ecommerce site, that's usually product page → cart → checkout → payment → confirmation.

  • Behavioral tracking installed. You need heatmaps, session recordings, and a conversion funnel running on the pages in that path. On Lucky Orange, that's one snippet and you're collecting data the same day.

  • A baseline number. Record your current conversion rate per step so you can tell later whether a fix worked. If you can't measure it, you can't audit it.

With that in place, work the six steps below in order. Each step narrows the problem and hands the next step a clearer target.

The 6-step CRO audit framework

The order matters. You move from the widest view (the whole funnel) to the narrowest (one element on one page), and you only diagnose a page once the funnel proves it's leaking. Skipping ahead is how teams end up "optimizing" pages that were never the problem.

Step 1: Map the funnel to locate the leak

Start wide. Before you judge any single page, build a conversion funnel across your defined path and read the drop-off percentage at each step. This is the step that turns "our conversions are low" into "we lose 41% of visitors between shipping and payment."

The funnel doesn't tell you why — it tells you where, which is exactly what you need first. In practice: if 100 people reach checkout and 59 make it to payment, the shipping/payment transition is your highest-value leak, and it becomes the target for Steps 2 through 6. Audit the leakiest step first, not the page you personally dislike most.

Step 2: Audit the page structure at the leak

Now zoom into the page where the biggest drop happens and ask a single question: is this page communicating the right thing in the right order? Most structural leaks come from important content sitting below where anyone actually looks.

Pull a scroll heatmap for that page. If 70% of visitors never reach your trust badges, guarantee, or key benefit, the content isn't wrong — its placement is, and no one's seeing it. A click/attention heatmap adds the second layer: it shows what people are drawn to versus what you wanted them drawn to. For the full checklist of hierarchy, above-the-fold, and content-order checks, run our page structure audit on this page before moving on.

Step 3: Audit the primary CTA

With structure confirmed, isolate the single action you want the visitor to take on this page. A page can be beautifully organized and still leak because the call to action is buried, vague, or competing with three other buttons.

Use a click heatmap to see whether people are actually clicking the CTA — or clicking something else, or a non-clickable element they think is the CTA. A common finding: visitors repeatedly click a product image expecting it to add to cart, then leave when nothing happens. Work through our CTA audit to pressure-test the wording, placement, prominence, and count of calls to action on the leaking page.

Step 4: Audit the mobile experience

A page can pass Steps 2 and 3 on desktop and still hemorrhage conversions on mobile, where most traffic now lives. Mobile is where layouts break, tap targets shrink, and forms turn hostile — and desktop testing hides all of it.

Filter your session recordings to mobile visitors only and watch five of them move through the leaking step. You'll see the real failures fast: a sticky header covering the "Buy" button, a form field that won't accept autofill, a zoomed-in layout that requires pinching. Then run our mobile UX audit to check tap targets, viewport, and mobile form behavior systematically instead of anecdotally.

Step 5: Audit page speed

Speed is the leak that fires before any of your content gets a chance. If the page is slow to load or slow to respond to taps, visitors abandon before Steps 2 through 4 even matter — which is why it sits here, after you've confirmed the content and interaction are sound.

Watch recordings for early exits: visitors who land, wait, and bounce within the first few seconds are usually a speed signal, not a content one. Pair that with your funnel timing — a step with a high immediate drop-off and long load time is a speed problem wearing a content costume. Our page speed audit covers the specific load, render, and interaction metrics to check and what "good" looks like for each.

Step 6: Confirm the cause, then prioritize

You now have suspects from the funnel and each narrow audit. Don't ship fixes yet — confirm. This is the step that stops you from "fixing" a symptom and wondering why the number didn't move.

Do two things:

  • Watch recordings of visitors who dropped at the leaking step. Aggregate data suggested the cause; individual sessions prove it. If ten in a row hesitate at the same shipping-cost reveal, that's your confirmed leak.

  • Ask the visitor directly. A well-timed exit survey ("What stopped you from checking out today?") captures intent that behavior alone can't infer — the "surprise shipping cost" you'd otherwise only guess at.

Then rank your confirmed leaks by impact: fix the one draining the most conversions on the leakiest funnel step first. That single ordering is what separates an audit from a to-do list.

What this looks like in practice

Say your checkout converts at 2%. Step 1's funnel shows a 41% drop at the shipping step. Step 2's scroll heatmap shows shipping cost appears below the fold, so it's a surprise. Step 3 confirms the CTA is fine. Step 4's mobile recordings show the surprise is worse on mobile, where the total jumps on a separate screen. You skip Step 5 because load times are healthy, and Step 6's exit survey has three people typing "shipping too expensive." Your fix isn't a new button color — it's showing shipping cost earlier. That's a diagnosis, not a guess, and you got there in one sitting.

Frequently asked questions

What is a CRO audit?

A CRO audit is a structured review of your website that pinpoints where and why visitors drop out of your conversion path. It combines analytics (to find which step leaks) with behavioral data like heatmaps and session recordings (to explain why), so you can fix confirmed problems instead of guessing.

How long does a CRO audit take?

A focused audit of one conversion path can be done in a day or two once your tracking is collecting data. The framework is deliberately sequential, so you spend your time on the single leakiest step rather than reviewing every page on the site at once.

What tools do I need to run a CRO audit?

At minimum you need conversion funnels, heatmaps, and session recordings on the pages in your conversion path, plus a way to survey visitors. Tools like Lucky Orange bundle all of these so you can move from "which step leaks" to "why it leaks" without switching platforms.

How is a CRO audit different from a CRO checklist?

A checklist is a quick-reference list of best practices to verify. A CRO audit is the diagnostic process that tells you which of those items actually matter for your site and in what order to address them, using your own visitor data as the evidence.

How often should I run a CRO audit?

Run a full audit quarterly, and a lighter one any time a key metric drops or you ship a major change to a page in your conversion path. Because the framework is repeatable, each audit is comparable to the last, so you can see whether your fixes moved the number.

Where do heatmaps and session recordings fit in a CRO audit?

Heatmaps come in at the page and CTA stages to show where attention and clicks land. Session recordings come in for mobile diagnosis and final confirmation, letting you watch real visitors fail at the exact step your funnel flagged. Funnel data comes first, to locate the leak before you diagnose it.

Your next step

Don't try to audit your whole site at once — that's how audits stall. Build a funnel for your single most important conversion path, find the step with the biggest drop, and run Steps 2 through 6 on that one step. Start a free trial of Lucky Orange and watch your first five recordings of the visitors who leaked — you'll usually spot the cause before you finish your coffee.

Lucky Orange Blog Author Icon

Lucky Orange