Storefront audit · No. 002
Storefront audit — a powersports retailer on Shopify
Your most expensive product tells every buyer they are saving minus three thousand dollars.
Ungated. Nobody asked for this and nobody paid for it. Want one for your store? Request a free audit →
Unsolicited and independent. We are not affiliated with the store, and we had no access to its analytics, ad account, or admin. Every finding below comes from the public storefront. The store is not named here, and we name no store. Product names, collection names and the store's own name are redacted inside quoted strings; everything outside the brackets is copied verbatim from the live pages at the time of review. Nothing here estimates traffic, conversion rate or revenue, because we have no way to know any of it.
Summary
The price block on your product pages renders a negative discount. On the flagship four-seat cab enclosure it reads
$3,185.00 USD Save $-3,185.00 USD, and it does the same on every priced product we opened. The collection tiles show the price cleanly, so the defect appears only on the page where the decision is made.Underneath that, the add-on records your product-options app creates are browsable and purchasable as products in their own right. A shopper can land on a page titled
MODEL YEAR, with no description, and add twenty-five dollars of nothing to the cart. That page carries eight customer reviews, which means people have already been routed through it.
- $-3,185.00 The saving advertised on the flagship enclosure
- 3 of 3 Priced product pages checked that render a negative saving
- $25.00 Price of an option record a shopper can buy on its own
- 3 Different lead times published for the same product
F1 · The price block advertises a negative discount Critical
On the highest-priced fixed-price product in the catalogue, a four-seat cab enclosure, the price block reads:
$3,185.00 USD Save $-3,185.00 USD
It is not confined to that product. The same pattern renders on a rear windshield at $679.00 USD Save $-679.00 USD and on a front windshield at $849.00 USD Save $-849.00 USD. Three of the three priced product pages we opened carry it, which makes it a property of the product template rather than a data error on one item.
The collection tiles for those same products show the price cleanly, with no savings badge at all. So the only place a shopper meets the negative number is the page where they decide — after they have clicked, and while they are reading the price.
A compare-at price that is set equal to the selling price produces exactly this. The badge is doing arithmetic on a discount that does not exist and printing the result.
Seen on the public storefront at the time of review.
Effort to fix: 1 hour.
F2 · Option records are sold as products Critical
Your product-options app writes its option sets into the catalogue as real products. They have machine-generated URLs of the form /products/option-set-…-checkbox-4, and they are reachable, indexable and buyable.
One of them is titled MODEL YEAR. It is priced $25.00 USD, it has no product description of any kind, it offers two model-year choices, and it renders an Add to Cart button. A shopper who arrives on it can put twenty-five dollars of nothing into their cart and check out.
That page also displays 4.1 / 5.0 based on 8 reviews. Reviews do not attach themselves to a page nobody reaches. Real customers have been routed through this record, which means it is not a theoretical exposure.
Seen on the public storefront at the time of review.
Effort to fix: 2–3 hours.
F3 · Those option records sit in the main browse collections High
They are not hidden in a corner of the catalogue. The all-products collection and the quick-order collection both list them inline among real merchandise:
ADD-ON TINT - From $75.00 USDCHOOSE TINT VALUE - Free - $25.00 USDMODEL YEAR - Free - $25.00 USD
Read the last one again. A single tile calls the same item both free and twenty-five dollars. A shopper browsing your full catalogue meets that contradiction next to products costing three thousand.
The cost here is not only the stray sale. It is that browsing the catalogue teaches a first-time visitor that the store is not tended, and that is the most expensive thing a storefront can teach.
Seen on the public storefront at the time of review.
Effort to fix: 1 hour, once F2 is done.
F4 · The same product is given three different lead times High
You make these goods to order, so the lead time is part of the price. It is published in three places and the three do not agree.
On a windshield product page, the top of the page states Windshields: 7+ days. Further down that same page, the FAQ accordion says:
It's typically, around 4-6 days and shipping is around 3-5 days depending on your location.
The dedicated lead-times page, which your site-wide announcement bar drives traffic to, says something different again: ~7+ business days to manufacture. 5-10 business days for shipping transit.
On the enclosures — the two- to three-thousand-dollar end of the catalogue — the spread is wider. The lead-times page says ~7-14+ business days to manufacture. 5-10 business days for shipping transit. The FAQ page says Typically 2 weeks. with shipping around 3-5 days. Business days and calendar days are not distinguished, and the transit windows differ by a factor of two.
A buyer spending three thousand dollars on something that does not exist yet reads the delivery estimate more carefully than any other sentence on the page. Three answers is worse than a slow answer.
Seen on the public storefront at the time of review.
Effort to fix: 2 hours. Pick one set of numbers, state whether they are business or calendar days, and render them from one source on all three surfaces.
F5 · Your strongest proof is filed where nobody buying can see it High
The most expensive fixed-price product in the catalogue, at $3,445.00 USD, displays Be the first to write a review.
It is not that you have no reviews. A near-identical enclosure one step down the price list shows 4.6 / 5.0 based on 7 reviews, and your dedicated reviews page aggregates Based on 677 reviews.
So the store has six hundred and seventy-seven pieces of evidence, and the single page where the largest decision is made has none of it. Reviews from the products a buyer is cross-shopping — the same enclosure for a different vehicle, the same part in a different size — belong on that page too, clearly labelled as such.
Seen on the public storefront at the time of review.
Effort to fix: 2 hours.
F6 · Duplicate add-on records split the same item in two Medium
Within one add-on collection, MODEL YEAR appears twice, at two different URLs and two different prices. Alongside it, ADD-ON TINT at From $75.00 USD sits next to ADD TINT at $75.00 USD.
Two live records for one thing split whatever reviews and history each accumulates, and give a shopper a choice between two options that are not actually different. Whichever one they pick, half the evidence is on the other.
Seen on the public storefront at the time of review.
Effort to fix: 1 hour.
F7 · The main navigation points at filtered URLs, and the filter lands in the page title Medium
Several primary navigation items link to tag-filtered collection URLs rather than to the clean collection. The filter is then concatenated into the title tag. The glass-replacement item resolves to a page whose title reads, with the store's own name redacted:
Glass Replacement for [category] GLASS REPLACEMENT REPLACEMENT GLASS | [store name]
That is past a hundred characters, with the same phrase written three ways. Search results show roughly the first sixty, so the visible part is the useful half and the rest reads as keyword stuffing to the ranking system.
The store's highest-value category does the same thing, leaking a bare lowercase tag: the title ends vault | [store name]. The clean collection URL, which is not the one in the navigation, produces the correct title.
Seen on the public storefront at the time of review.
Effort to fix: 1 hour.
F8 · Housekeeping that is visible to shoppers and to search engines Medium
Four smaller things, grouped because none of them takes more than a few minutes.
- Two pages titled
TESTINGandTESTING 3are live, and both are listed in the store's own public page sitemap — actively submitted to search engines. One of them contains real product content. - A seasonal gift collection is live and indexed out of season, carrying an active offer:
USE DISCOUNT CODE PRESENT10 AT CHECKOUT FOR [ten per cent] OFF ANYTHING IN THE [seasonal] COLLECTION. We did not attempt checkout, so whether the code still redeems is not something this audit can tell you — which is itself the problem, because a shopper will find out by trying. - The footer misspells your main product category on every page of the site, as
WINDHSIELDS. - A navigation-linked packages collection holds two entries, of which the only actual package shows
Sold Out; the remaining item is a single bare wheel. Another nav category holds exactly one product.
Seen on the public storefront at the time of review.
Effort to fix: 1 hour for all four.
What is already working
W1 · You publish lead times before the sale, not after Strength
Your site-wide announcement bar reads ATTENTION!! CLICK HERE FOR INFO REGARDING MANUFACTURING AND LEAD TIMES and links to a page that breaks the estimate down by product type, covers custom work, and states plainly that These times are approximate and may vary based on current demand and production schedules.
For a made-to-order manufacturer that is the right instinct, and most of your competitors bury it. F4 is not an argument against this page. It is an argument for making the other two surfaces agree with it.
W2 · You publish your negative reviews, unedited Strength
Your reviews page carries a two-star review in full, with its original spelling intact, describing an order that arrived incomplete and a request that was not accommodated. It has not been removed, softened or buried.
That is uncommon and it is worth more than the aggregate score above it. A page of uniformly perfect reviews is read as curated; one bad review left standing is what makes the other six hundred credible.
W3 · The policy infrastructure is real Strength
Product pages carry explicit fitment data. The warranty is a properly drafted document covering defects in materials and workmanship that occur during normal, intended use, with honest and specific exclusions rather than a blanket disclaimer. The returns policy states its window and discloses the restocking fee up front instead of at the point of refund.
Fitment, a real warranty and a stated returns window is more policy infrastructure than most stores this size maintain, and it is the part that is hardest to add later.
Priority order
| ID | Fix | Effort |
|---|---|---|
| F1 | Clear the compare-at price wherever it equals the selling price, so the savings badge stops rendering a negative number | 1 hour |
| F2 | Take the option-set records out of the catalogue: no product page, no URL, no add-to-cart | 2–3 hours |
| F3 | Remove the option records from the all-products and quick-order collections | 1 hour |
| F4 | Pick one set of lead times, say whether they are business or calendar days, render them from one source everywhere | 2 hours |
| F5 | Show reviews on the highest-priced pages, including clearly-labelled reviews of the same part for other vehicles | 2 hours |
| F8 | Delete the test pages, retire the out-of-season promotion, fix the footer spelling, hide the thin categories | 1 hour |
| F6 | Merge the duplicate add-on records | 1 hour |
| F7 | Point the navigation at clean collection URLs and stop the tag from entering the title tag | 1 hour |
Method
The checklist behind this audit comes from walking twelve high-volume storefronts end to end and writing down what each one does at every step: how the funnel is structured and how the store bills, not how polished it looks. A store gets read against that structure.
Findings here are drawn from the public storefront only — the homepage, the navigation, the collections, the highest-priced product pages, the policy and lead-time pages, and the store's own published sitemaps. Quoted strings are copied verbatim from the live pages at the time of review, with the store's name, its product names and its collection names redacted inside brackets. We make no claim about the store's advertising. No analytics, admin or ad-account access was requested or used, so nothing here estimates traffic, conversion rate or revenue.
Two things we could not check and therefore do not assert: whether the seasonal discount code still redeems, and whether variant selection behaves correctly on the enclosure pages. Neither appears above as a finding.
What to do with this
F1 and F2 are the whole job. Between them they are an afternoon, they need no redesign, and until they are done your most expensive product advertises a negative discount and your catalogue sells an item that is not a thing.
This audit is free and public. Hand it to whoever already works on the store — there is nothing proprietary in it, no gate, no form, and no call required to read all of it. If you would rather we did the work, send the store URL to hello@altamiracommerce.com and we will reply with what we would change first and what it would cost. The same reading applies to a storefront in any trade.