Three priority tiers of audit findings, and nothing left in the backlog
A US leather goods retailer
The Brief
The store had conversion gaps along the full length of its funnel, and a prior audit had already named them clearly. The problems were not a mystery. The document existed.
That is usually where the story ends. An audit is worth exactly what gets built from it, and most of them are implemented halfway, then quietly parked as a known-issue list nobody reopens.
The findings were specific. Product pages had no sticky Add to Cart, so a shopper who had scrolled a long page to decide had to scroll all the way back up to act. Discount value was not shown near the price. The cart drawer offered no payment logos or trust signals at the exact moment a customer decides whether to hand over a card. There was no first-time-visitor capture at all. And a set of usability faults sat underneath it: an empty cart that dead-ended, a review rating that linked to the wrong place, irrelevant dropdown options, a cramped search box.
All of it had to be fixed on a live store that was trading throughout.
The friction, point by point
What the audit found, and what replaced it.
| Funnel stage | Result | Status |
|---|---|---|
| Product page | 2,042 | Sticky Add-to-Cart with variant and quantity |
| Price | 1,936 | “Save $X” messaging above the fold |
| Cart drawer | Full store | Payment logos and warranty, returns and reviews strip |
| First-time visitors | None | Discount popup, does not reappear once dismissed |
| Empty cart | None | Continue Shopping call-to-action |
| Review rating | Zero | Links to the reviews |
The Approach
Every change was built in a copy theme first, never on the live store. On a trading Shopify site that is the whole discipline: a theme edit that looks fine in the editor can break a variant selector on mobile, and you find out through a customer who could not check out.
Each fix was QA-tested on its own rather than as a batch, so when something behaved unexpectedly it was obvious which change caused it.
Then we produced a before-and-after report, so the retailer could see each change side by side and sign off before anything went to the live store. Nothing published on our judgment alone.
What we Did
Added a sticky Add-to-Cart bar with variant and quantity selectors, so the buy action follows the shopper down the page instead of waiting at the top for them to come back.
Moved discount messaging above the fold. “Save $X” placed next to the price, where the saving is part of the decision rather than a surprise discovered later.
Put trust signals in the cart drawer. Payment logos plus a strip reading “Hassle-Free Warranty, Easy Returns, 3,000+ Five Star Reviews.” The drawer is the last screen before checkout and it was the emptiest one on the site.
Introduced a sticky site-wide urgency banner linking through to the sale page, giving every page a route to the current offer.
Built a first-time newsletter popup with a discount, set not to reappear once dismissed. Capture on the first visit, and no nagging on the second.
Fixed slow-loading review and testimonial sections on the homepage and made the social proof metrics visually prominent. Proof that arrives after the shopper has scrolled past it is not doing any work.
Made the empty cart recoverable and fixed a misrouted link. A “Continue Shopping” call-to-action replaced the empty-cart dead end, and the review rating was corrected so it navigates to the reviews rather than to related products.
Cleared the smaller usability debt. Back-to-top button, clickable footer logo, irrelevant options removed from the collection dropdown, a widened search auto-suggest dropdown, and active-state highlighting on the menu. Individually minor, collectively the difference between a store that feels maintained and one that does not.
Results in Detail
Every audit item across all three tiers implemented, QA-signed, and published live. No items deferred to a known-issue backlog.
This is the result, and it is rarer than it should be. Audits routinely land with the critical tier shipped, the high-priority tier half done, and the medium tier abandoned on the grounds that it is only cosmetic. Here P3 shipped too, which is why the small stuff — the footer logo, the dropdown, the search box — is on this page at all.
A documented before-and-after report delivered, validating each fix on the live site.
Before
After
The report is what makes the work checkable rather than asserted. Each change is shown in its prior and current state, so the retailer can verify what was done without taking anyone’s word for it.
The purchase path tightened at multiple points, from a persistent Add-to-Cart and visible discounts through to cart-drawer trust signals and a recoverable empty-cart flow.
Each fix removes friction at a different stage, which is the point of working from a funnel-wide audit rather than optimizing one page. A sticky buy button does not help if the cart drawer then loses the sale.
Tools & technologies
What the build runs on
verified