What a CRO audit should include: your GA4 funnel, session recordings, a tracking check and fixes ranked in dollars, and how to judge one before you pay.
A CRO audit should tell you which fixes are worth the most money on your store, and prove it with your own data. That means your GA4 funnel, recordings of your own customers, a check that your analytics counts every order, and a ranked list that puts a dollar figure and a confidence level on each fix. Contentsquare's 2026 benchmark found friction such as rage clicks, slow pages and errors in 35.2% of 99 billion sessions, and its own advice is that each business has to study behaviour on its own site to find its biggest pain points. A generic checklist finds the problems every store has. Your customers' sessions show the ones that cost you sales.
Australian audits range from a AU$699 Shopify review to a AU$14,500 audit with competitor benchmarking and a workshop, so two quotes for "a CRO audit" can describe different jobs. This guide sets out what a good one contains, what you should hold in your hands at the end, and how to tell the two apart before you pay.
What should a CRO audit include?
A CRO audit should include seven parts: a tracking check, the purchase funnel, session recordings, heatmaps and scroll, forms and checkout, real-user speed, and a heuristic review. The first six use your own store's data. The heuristic review compares your pages with published good practice.
| Part | What it checks | Data it uses | What a thin audit does instead |
|---|---|---|---|
| Tracking check | Whether analytics counts every order, once | GA4 purchases against Shopify orders for the same dates | Trusts the dashboard |
| Purchase funnel | Where shoppers drop out: product, cart, checkout, payment | GA4 funnel exploration, split by device and channel | Quotes the store-wide conversion rate |
| Session recordings | What shoppers do at the step where they leave | Microsoft Clarity or a paid recording tool, 200 or more sessions | Watches a handful, or none |
| Heatmaps and scroll | What people click, miss and never reach | Clarity heatmaps on the top templates by revenue | Reviews screenshots |
| Forms and checkout | Fields, errors, surprise costs, payment options | Recordings of the cart and checkout, form errors, Shopify checkout settings | Applies a checkout checklist |
| Real-user speed | Load and response times your customers get | Chrome UX Report field data, Shopify's web performance report | Runs one lab test |
| Heuristic review | The page against known good practice | Published guidelines (Baymard, NN/g) and the reviewer's experience | This is the whole audit |
Tracking check
The tracking check comes first because every other number in the audit depends on it. Compare the purchases GA4 records with the orders Shopify records over the same 30 days. If GA4 is missing a slice of orders, every funnel rate built on it is wrong, and so is any before-and-after measurement of a fix. Our guide to GA4 on Shopify covers why the two disagree and which gaps to fix.
Purchase funnel
The funnel shows where the money leaves. GA4's funnel exploration takes up to 10 steps and breaks each one down by a dimension such as device category, so a good audit reads product view, add to cart, begin checkout and purchase separately for phone and desktop, and for paid, organic and email traffic. A store-wide conversion rate hides the gap. Contentsquare's 2026 benchmark puts desktop conversion at 3.4% against 2.0% on mobile, while mobile carries about 70% of traffic, so the phone funnel is usually where the largest leak sits.
Session recordings and heatmaps
Recordings explain why people drop out at the step the funnel points to. Microsoft Clarity is free, records without sampling and keeps data for 30 days, and it flags dead clicks (taps that get no response), rage clicks and JavaScript errors. An auditor who reads 200 recordings of people who reached the cart and left will find problems no checklist names, such as a delivery-date picker that fails on one phone model or a discount field that sends people off to search for a code.
Forms and checkout
Checkout gets its own pass because it sits closest to the money. Baymard's benchmark counts 23.48 form elements in the average US checkout against an ideal of 12 to 14, and "extra costs too high" is the top reason for abandonment, named by 40% of US shoppers who abandoned. On Shopify, what the audit can recommend depends on your plan: apps that change the information, shipping and payment steps of checkout are available only on Shopify Plus. An audit that recommends checkout changes your plan cannot make has skipped this check; our Shopify checkout guide lists what each plan allows.
Real-user speed
Speed belongs in the audit when the field data shows a problem. Shopify's April 2026 study of its own stores found that for every 100 milliseconds slower a store loads, conversion "tends to be about 3.5% lower", which Shopify describes as "a clear correlation between performance and conversion, though not direct causation". Use real-user data from the Chrome UX Report or Shopify's web performance report. A single lab score from one test run says little about what your customers get.
Heuristic review
The heuristic review compares your pages with published good practice, and it belongs in every audit as one part of seven. It catches the known problems quickly: a hidden shipping cost, a forced account, a size guide nobody can find.
Why isn't a best-practice checklist enough?
A checklist finds the problems that most stores share and misses the ones your customers hit. In Nielsen's heuristic evaluation research, single evaluators found only 35% of the usability problems in an interface, averaged over six projects, which is why NN/g recommends three to five evaluators working independently. NN/g's current guide adds that heuristic evaluations "cannot replace user research" because "user-experience design is highly contextual".
Baymard's guidelines are the best checklist available. They come from usability testing and from benchmarking 344 leading US and EU ecommerce sites against 650+ guidelines. They tell you what tends to go wrong. They cannot tell you which of the improvements they flag matter at your traffic, your average order and your customers' devices.
Our own store shows the difference. The audit we ran on Paperlake, a store we run, read Clarity data for the 30 days to 18 November 2026:
| Page | What the recordings showed | Verdict |
|---|---|---|
| Book builder | 18% of 953 sessions clicked something that did nothing | Fix first |
| Meta landing page | Visitors from Meta scrolled 21% of the page on average, across 347 sessions | Fix |
| US landing page | A script error fired on 24% of 70 sessions | Fix |
| Order confirmation | 46% of 65 sessions clicked dead text on the thank-you page | Leave |
No checklist would have named the book builder, because it is unique to that store, and it was the most valuable finding. The last row is as useful. Almost half of buyers clicked dead text on the confirmation page, and the audit said to leave it because those people had already paid. An audit that ranks by money tells you what to ignore.
What should you receive at the end of a CRO audit?
You should receive a ranked list of fixes with a dollar figure, the evidence and an effort estimate for each one, plus a tracking verdict, a baseline number and a plan for who builds what. A PDF of observations without a value per fix leaves you to do the ranking yourself.
Fixes priced in dollars
Each fix should carry the revenue at stake, worked out from your own traffic and average order. The arithmetic is simple enough to check: sessions affected a month, times the conversion you could recover, times your average order, times 12.
An illustrative store (every figure below is an assumption): AU$3M a year online, a AU$120 average order, about 2,080 orders a month.
| Fix | Evidence | Revenue at stake a year | Confidence | Effort |
|---|---|---|---|---|
| Script error on the mobile cart | 800 sessions a month hit it; they convert at 10% against 25% for cart sessions without it | Up to AU$172,800 | High: the error stops the next step | Hours |
| Shipping cost shown only at checkout | 4,500 sessions a month reach checkout and 45% complete; recordings show people leaving on the shipping step | AU$194,400 if completion rises 3 points | Medium: the size of the lift is an estimate | A day |
| Product image zoom gets dead clicks | 12% of product-page sessions tap the main image and nothing happens | Not priced: no clear link to purchase yet | Low | A day |
| Dead clicks on the order confirmation page | Shoppers tap order details after paying | AU$0 in conversion | High | Leave |
The first row works out as 800 × (25% − 10%) = 120 orders a month, × AU$120 = AU$14,400 a month, × 12 = AU$172,800 a year. That is an upper bound, because not every affected session would have bought. The second row is 4,500 × 3 points = 135 orders, × AU$120 × 12 = AU$194,400. The first row ships first even though it is smaller, because the confidence is higher and the work takes hours.
Evidence for every fix
Every fix should point to its evidence: a recording, a funnel step, a heatmap or a form error count. You need it to brief a developer, and to argue with the auditor when you disagree.
A tracking verdict and a baseline
The audit should state whether your analytics can be trusted and agree the one number future work will be measured on, usually revenue per visitor, measured over your last 30 days. Without a baseline written down before anything changes, nobody can say later whether the fixes worked. Optimizely found that about 12% of experiments win on their primary metric across 127,000+ experiments run on its platform from 2018 to 2023. Most ideas do not move the number, and only a measurement against the baseline shows which did.
Who builds what
The audit should say who will build each fix and roughly what it costs. Most Australian audits are a diagnosis: of the providers in our CRO cost guide, Studio Elephant and Fontis quote the build separately, and Studio Elephant prices follow-on Shopify work at AU$140 an hour. Ask before you buy whether the auditor will build the list, and at what price.
How long does a CRO audit take, and what does it cost?
A CRO audit takes about 2 to 6 weeks and costs AU$699 to AU$14,500 on published Australian price pages. Turnarounds on those pages run from under 10 business days (Studio Elephant, Cart Forensics' focused audit) to 2 to 4 weeks (iOnline) and 15 to 30 business days for Fontis's three tiers.
Add time before the audit if you have no recording tool. Clarity keeps 30 days of data, so install it two to four weeks before the auditor starts, long enough to capture a few hundred sessions on your main templates and a normal trading week. A store with 20,000 sessions a month gets there faster than one with 5,000.
Most of the price difference is research depth and scope. A AU$700 audit covers a few page types on phone and desktop. A AU$7,500 audit covers the whole journey with your own analytics, recordings, more templates and sometimes competitors. The full price tables, with what each provider's page says you get, are in our guide to what CRO costs in Australia.
How do you judge a CRO audit before you buy it?
Judge an audit by its inputs and its outputs: what data the auditor will read, and what you will hold at the end. Ask these before you pay:
- Which of our data will you read: GA4, Shopify orders, recordings, form errors, field speed data?
- How many recordings will you watch, and of which sessions?
- Will you check that GA4 counts every Shopify order?
- Will each fix carry a dollar figure worked out from our traffic and average order?
- Can we see a redacted example of a finished audit?
- Will we agree a baseline number, and over which dates?
- Who builds the fixes, and what does that cost?
- Who does the work: the person on the sales call, or someone else?
The sample is the most useful of these. If every fix in it could apply to any store, yours will read the same.
What are the red flags in a CRO audit?
The red flags all point to an audit that reviewed pages without reading the store's data.
- No analytics or recording access requested. An auditor who never asks for GA4 or Clarity has not read your customers.
- A score out of 100 with no dollar figures. A score measures distance from a checklist, and you need to know what each issue costs.
- More than 50 recommendations with no ranking. You will build the easy ones and skip the valuable ones.
- Checkout changes your Shopify plan cannot make, recommended without a note about the plan.
- Lifts promised before anyone has seen your data.
- Speed judged on a lab score alone.
- No baseline. If the number is not written down before work starts, nobody can say afterwards whether it moved.
Can you audit your own store in an hour?
You can do a useful first pass in an hour with free tools, and below about AU$1.5M a year online it is usually the right first step. Compare GA4 purchases with Shopify orders for the last 30 days. Build a GA4 funnel exploration from product view to purchase, split by device. Watch 20 Clarity recordings of phone sessions that reached the cart and left. Then open your store on your own phone and check what a new shopper learns about shipping cost and delivery time before checkout.
What takes days is the rest: reading 200 or more recordings, pricing each problem from your own numbers, and deciding which fixes are worth the build. Installing Clarity takes minutes. Turning what it records into a ranked, built list is the work an audit charges for.
What does Studio Slate's audit include?
Our paid audit costs AU$2,500 to AU$4,000 plus GST, depending on the size of the store, and the full fee is credited to the first month of the retainer if you start within 90 days. The three founders read your sessions, recordings, heatmaps, funnels and form data, price each leak from your own traffic and average order, rank every fix, and agree with you in writing the one number we will be measured on, usually revenue per visitor against your last 30 days.
It is built for stores turning over about AU$1.5M a year or more online with at least 20,000 visits a month. Below that, a AU$699 to AU$1,500 audit from the table in our cost guide, or the one-hour pass above, is the better first spend. If you are weighing an audit against ongoing work, our guide to a CRO retainer or a one-off audit sets out which fits which store.
Sources
- NN/g, Jakob Nielsen: The Theory Behind Heuristic Evaluations (1994) - single evaluators found 35% of usability problems, averaged over six projects; three to five evaluators recommended
- NN/g: How to Conduct a Heuristic Evaluation (2023) - heuristic evaluations cannot replace user research
- Baymard Institute: cart abandonment rate statistics (updated 28 October 2026) - 23.48 form elements against an ideal 12 to 14; 40% extra costs; 39 potential checkout improvements
- Baymard Institute: cart and checkout usability research - 344 benchmarked sites, 650+ guidelines
- Contentsquare: frustration in the 2026 Digital Experience Benchmark - 35.2% of sessions affected by friction, 99 billion sessions on 6,500 sites; and the engagement chapter: desktop 3.4% against mobile 2.0% conversion, mobile about 70% of traffic
- Microsoft Clarity FAQ and semantic metrics - free, no traffic limits, no sampling, data kept 30 days; dead clicks, rage clicks and JavaScript errors
- Google Analytics Help: funnel exploration - up to 10 steps, breakdown by dimension
- Shopify Help Center: customising your checkout with apps - checkout step apps limited to Shopify Plus
- Shopify: Store Speed and Conversion, What the Data Shows (27 April 2026) - 100 milliseconds slower, about 3.5% lower conversion; correlation
- Optimizely: Evolution of Experimentation (2023 ebook, 127,000+ experiments, 2018 to 2023) - about 12% of experiments win on the primary metric
- Provider turnarounds and prices, each read on the provider's own page on 20 November 2026 and named without links: Studio Elephant, Cart Forensics, iOnline, Fontis
- Paperlake audit figures: Microsoft Clarity data for the 30 days to 18 November 2026, a store we run
Questions
What should a CRO audit include?
A CRO audit should include a check that your analytics counts every order, the purchase funnel step by step by device and channel, a read of session recordings and heatmaps on the pages that lose the most money, a review of forms and checkout, real-user speed data, and a heuristic review against known good practice. It should end in a list of fixes ranked by what each is worth to your store, with the evidence for each and a baseline number agreed in writing.
How much does a CRO audit cost in Australia?
Published Australian CRO audits cost AU$699 to AU$14,500, and most sit between AU$1,500 and AU$7,500, on the price pages of 12 providers read on 20 November 2026. The low end reviews a few page types against a checklist; the higher prices add your own analytics, recordings, more templates and competitor benchmarking. Building the fixes is usually quoted separately.
How long does a CRO audit take?
Published turnarounds run from under 10 business days for a focused Shopify audit to 15 to 30 business days for a larger one. If Microsoft Clarity or another recording tool is not installed yet, add two to four weeks before the audit starts, because the auditor needs recordings of your own customers to read and Clarity keeps them for 30 days.
Is a free CRO audit worth it?
A free audit is worth taking if you treat it as a sample of how the provider thinks. Most free audits are a short review of one page against good practice, because nobody can read your analytics and recordings properly for nothing. Judge it on whether it names specific problems on your store with evidence, or lists generic advice that would fit any store.
Can I do a CRO audit myself?
You can run a useful first pass in about an hour: compare GA4 purchases with Shopify orders, build a GA4 funnel exploration split by device, watch 20 Microsoft Clarity recordings of people who reached the cart and left, and check what a phone shopper sees about shipping costs before checkout. A full audit takes days because reading 200 or more recordings and pricing each fix is the slow part.



