E-commerce Store Audit
Your store gets visitors. This tells you which ones you are losing to your own site.
A fixed-price, independent review of your online store. Twelve sections, a real basket taken through your checkout on a real phone, every finding located and evidenced, and a ranked list of what to fix first.
Around 70% of carts are abandoned. Most of that is not your fault.
Cart abandonment across online retail runs at roughly 70%. That figure is an average across fifty studies, compiled by the Baymard Institute, whose checkout research is the most cited work on the subject rather than the only measurement of it. The largest single reason people give is that they were only browsing and were never going to buy. No audit fixes that, and anyone telling you otherwise is selling you something.
Most of it no audit can touch. What is left is documented, specific and countable
- 40%Extra costs too high: shipping, tax or feesYour website
- 20%Delivery too slowOnly partly
- 19%Not trusting the site with card detailsOnly partly
- 18%Forced account creationYour website
- 17%A checkout that is too longYour website
Baymard’s survey of US online shoppers: one study, one sample, and not a measurement of your store. Respondents could give more than one reason, so the figures do not sum to 100 and are not slices of a whole. Your audit measures your own checkout.
Three of those five are properties of your website. Two are only partly. How fast you actually deliver is logistics, and whether people trust you with a card is substantially brand and payment provider, though your site can help with both by saying the right things in the right places. Knowing which of them is yours is the whole job.
Across IRP-hosted UK merchants, which skew towards small and medium retailers, June 2026 ran at a 2.03% conversion rate with an average order value of £127.06, and mobile accounted for 63.5% of sales. At that order value, a handful of recovered checkouts a month is a real number.
Cart and checkout
Seven steps to pay, against a 5.1 benchmark. Extra costs appear at step four.
- 01Extra costs first shown at the payment stepCheckout / step 4 of 5Fix first
- 02Guest checkout not offeredCart / account requiredFix first
- 03Product markup missing from server HTML148 products / section 11High
- 04Total price assembled during checkoutCheckout / section 12High
- 05Cancellation form not providedPolicy pages / section 12Medium
A document written to be used, not filed.
- 01A plain-language executive summaryThe three things that matter. Read nothing else and you will still act correctly.
- 02A scorecard across all twelve sectionsWith the method stated, so the numbers can be checked rather than believed.
- 03Your checkout, walked and documentedStep by step, on desktop and on a real phone, with the step count and field count recorded.
- 04Every finding locatedWhich page, which element, which measurement. Not a note that your checkout is long.
- 05A ranked priority listWhat to fix first, judged on impact against effort.
- 06What could not be verified, and whyA finding we cannot prove is an open question, not a conclusion, and it is recorded as one.
A mini audit in 2 business days.
Your homepage plus one page you nominate. No obligation and no call. We send the report and one short follow-up email, and if you do not reply you will not hear from us again unless you ask.
- 5 headline scores across performance, findability, accessibility, AI readiness, and trust and compliance signals.
- 5 findings, each named, located and checkable by you in 10 minutes on your own store.
- 1 thing already working well, because a useful review does not invent work.
- A clear boundary stating what the mini audit did not cover.
- 1 line on the full audit, so the next step is visible without becoming a sales sequence.
A free audit you have to take on trust is worth nothing. The findings are designed for you to verify yourself.
Request the mini auditTwelve sections. Eight for the website, four for the store.
Everything in the website audit, plus four sections that only apply when money changes hands.
- 01Performance
- 02Findability
- 03AI and agent readiness
- 04Accessibility
- 05Security and resilience
- 06Conversion and journey
- 07Compliance and disclosures
- 08Measurement
Product and collection pages
Product imagery and variant selection, price and delivery clarity before the cart, stock accuracy, description quality against the questions buyers actually ask, review display, filtering, sorting, and on-site search including what happens when it returns nothing.
Measured where the money is
Performance measured on collection and product pages specifically, because those are heavier than your homepage.
The questions buyers ask
Whether the description answers them, or leaves the visitor to guess and leave.
Cart and checkout
We take a real basket through your checkout to the payment step, on a real phone. Guest checkout, when the total cost first becomes visible, payment methods including the wallets that dominate mobile, card form quality, autofill, error handling and cart persistence.
Counted against a benchmark
Step count and form field count against Baymard’s 2024 benchmark for large US and EU stores, 5.1 steps and 11.3 form fields.
When it goes wrong
Out of stock at checkout, an expired session, an invalid code. The paths nobody tests.
Product data and shopping surfaces
Your product structured data against what Google requires for merchant listings, feed health, image resolution, and return and shipping data.
In the HTML your server returns
Which is what Merchant Center needs for automatic item updates, and which headless and app-heavy stores often get wrong while looking correct in a browser.
Ready for what is coming
Where the published agentic commerce standards are heading, so your product data is ready when they arrive.
Distance selling and consumer law
Whether your pre-contract information, cancellation rights, refund terms and cancellation form meet what the Consumer Contracts Regulations ask for. We check your store against what published requirements ask for. We are not lawyers and this is not legal advice.
The total price, and when
Whether it is shown when the rules say it should be, rather than assembled during checkout.
Claims that have to be real
Whether your countdown timers and low-stock messages reflect anything, and whether your review handling meets the standard now expected of the store.
Five of forty-three findings, and the one we would not call a finding.
A real store audit is published on this site, redacted. It rated twelve categories and produced forty-three findings. Five are set out in full with the evidence and the severity ratings. Among them:
- 01A market configuration that sent the founders’ own countrymen to a storefront built for somewhere else.
- 02A star rating going to Google that no customer had given.
- 03Contract documents pointing at a dead domain.
- 04Terms nominating a governing law under which the store’s own refund policy did not work.
- 05A paid app running against product data that did not contain the fields it needed to function.
- OpenAnd the item we entered as an open question rather than a finding, because we could not prove it without placing a test order: a currency converter sitting over an unchanged market configuration. That is published too, with the two-minute test the owner could run themselves.
Each mark is one finding. The outlined mark is the one we could not prove.
Audit clients are anonymised by default. We publish findings, not names.
No scanner has ever taken a basket to a payment screen.
We use the same measurement tools any competent auditor uses. That is not the differentiator and we will not pretend it is. What you are paying for is the judgement about which measurements matter for your store, plus the work no scanner can do at all.
An automated scan
- Flags issues without deciding which ones matter here.
- Cannot put a basket through a checkout.
- Cannot read a refund policy and compare it to the terms.
- Sends every store in your sector a near-identical export.
This audit
- A real basket, taken to the payment step on a real phone.
- Keyboard-only navigation of every primary journey.
- Your policy pages and your terms, read against each other.
- Your product data checked in the HTML your server returns.
We build stores. So we have an interest in what your report says.
Pellmark builds and manages websites and online stores and sells monthly care plans. Start a care plan within 30 days of your report and we credit the audit fee against it. That is a commercial interest in what your report concludes.
Three fences hold it, and all three are binding on every audit we run.
The report does not recommend Pellmark’s own services. It names what is wrong and what to do about it.
There are no links to our services inside the report, and no conclusion that you need a rebuild.
Every report carries a paragraph disclosing this relationship. It is in the document itself, not only on this page.
We also check our own audits and say when they were wrong. The published store audit went through three versions before it was fit to send, and the second corrected four findings from the first. If you want us to carry out any of the work, you raise it, and it is a separate conversation.
One clear fee. Fixed scope.
Twelve sections, up to eight pages, plus the checkout
Covers all twelve sections, up to eight pages plus a complete checkout run-through, the full written report and the 45-minute debrief call. No per-page pricing and no scope negotiation.
GBP prices shown are the total payable.
Fixed-price e-commerce audits from UK agencies commonly start between £1,500 and £2,400 plus VAT, and specialist consultancies start higher.
Start a Studio Care or Premium Care plan within 30 days of receiving your report and the audit fee is credited in full against the plan. On Essential Care the credit is capped at three months of the plan fee, because Essential includes neither the SEO work nor the AI-readiness foundation that a good part of your report will point at.
The 30-day window exists because the report is a snapshot of your store on the day it was checked. Care plan clients receive an audit as part of their plan regardless.
Terms and practical details
Invoicing and the delivery clock
Invoiced in full before work starts. The five working days begin on the first working day after both the invoice is paid and we have your URL and page list. Your confirmation email states the working day the report will arrive.
On the checkout run-through
We take your checkout to the payment step. We do not complete a purchase unless you ask us to, and if a full test order would be useful we agree it with you first and you decide how it is paid for.
Cancelling and refunds
Full refund before work begins. Once under way, we refund the portion not yet worked. A completed audit is not refundable on its findings: a report concluding your store is in good order is a delivered result.
If a finding is wrong
Tell us and we will check it. Where we agree, we correct the finding and reissue the report at no charge.
The Boundaries
What this audit is not.
Not legal advice.
We check your store against what published UK requirements ask for and show you where it does not match. We are not lawyers.
Not a full accessibility conformance audit.
It is an expert review against WCAG 2.2 AA with manual keyboard and screen-reader checks. If the findings are extensive, we will say so and tell you what a full conformance audit involves.
Not a conversion testing programme.
It is a diagnosis, not an experiment. We identify where the checkout leaks and what the documented causes are. We do not promise a percentage. Anyone quoting you an uplift figure before looking at your store is quoting a study average, not a prediction about your business.
Not a sales pitch.
If the audit concludes your store is fundamentally sound, it will say so, and you will have a documented answer instead of a suspicion.
Not selling online yet?
The Website Audit is sections one to eight of this page, without the four store sections. Up to five pages and a 30-minute debrief. Same method, same report structure, same fences.
- Up to five pages
- Eight sections
- 30-minute debrief
- 5 working days
FAQ
Questions people ask before booking.
Any. Shopify, WooCommerce, Squarespace, Wix, BigCommerce, Magento or custom. The sections are the same; only the fixes differ.
Read-only access to analytics and Merchant Center where you have them makes the audit substantially better. We can run without it at the same fee, and the report states plainly what could not be verified.
We run the checkout to the payment step. We do not complete a purchase unless you ask us to, and if a full test order would be useful we agree it with you first.
No. An audit is included in your plan. Ask us and we will schedule it.
No. We check your store against what published UK requirements ask for and show you where it does not match. Acting on it is a conversation for you and your solicitor, and the report is written so you can hand it to one.
Yes, and that is the point of it. The call exists so the person who has to do the work can ask questions. Move it as often as you need to, at no charge. If we have to move it, the next slot will not be more than two working days later.
Audit clients are anonymised by default. We publish findings, not names. Nobody wants their defects public, and discretion is part of what you are paying for. Our website builds are named because a finished site is something a client is pleased with. An audit is the opposite kind of document.
Reply to the report, or email hello@pellmark.com with “audit” in the subject line, and you will hear back within one working day. The audit services are covered by our terms of use.
Find out what your checkout is costing you.
Start with the free mini audit: five findings on your own store, each one you can check yourself in ten minutes. No call, no obligation.
Request