Find where your catalog loses the buyer — in 60–90 seconds
The scan opens your catalog, a product page and the cart the way a buyer does, and scores what it finds: one CRO score from 0 to 100, the issues that cost the most and a ranked plan to fix them. It is read-only and it runs on any public Shopify storefront today — including yours. What is still to come is the B2B scoring: price-list pages, quote forms, MOQ and pack-size tables, and the reorder path.
- Read-only — never changes your storefront
- CRO score 0–100 · Target 70+
- Shareable public report
10 points below Target 70+
Six weighted parts, scored the way a buyer reads the catalog.
from public signals
60–90 s
From URL to score
Four passes over the live storefront — connect, read, score, quantify — then Copilot writes the brief
0–100
One CRO score, target 70+
“Healthy” starts at 70; below 45 the verdict reads “Critical friction”
6
Weighted parts
Homepage clarity /15 · product page structure /25 · cart and checkout /20 · trust /15 · offer and AOV /15 · technical /10
≈
Revenue at risk, always marked
A directional estimate from public storefront signals — never presented as measured
Every figure above is the scan as it ships today, scored with DTC weighting. The B2B areas described on this page — price-list pages, quote forms, MOQ tables, the reorder path — are roadmap, and the screens are illustrative.
One score. Six weighted parts. A target of 70+.
Every scan writes a CRO score from 0 to 100 and the verdict that goes with it — Healthy, Needs attention or Critical friction — plus the highest-impact root cause and the action recommended next. Scores are trended across your last six scans against a dashed Target 70 line, so you can see whether the storefront is moving toward the benchmark or away from it.
- A weighted breakdown you can read part by part, with the weakest part ranked first and linked to the module that can improve it
- Each part is scored the way a visitor sees it — what is above the fold, what the cart does, which trust marks are where the decision is made
- On a B2B catalog the same six parts still apply; the roadmap adds the areas a buyer cares about that a shopper does not
CRO score
Why conversion is underperforming and which root cause matters first.
Audit snapshot
Latest completed scan · beautyskin.com
64
of 100
Several conversion areas are underperforming.
6 points below Target 70+ · 6 / 6 parts backed by evidence
Score history · last 6 scans
Score breakdown · weakest first
Each part links to its moduleHighest-impact root cause
Weak above-the-fold product page structure
PDP structure scores 15 of 25 — price, variant, add-to-cart and the rating summary sit below the fold on mobile.
CRO score
64 / 100
Target 70+
Evidence coverage
6 / 6 parts
Homepage · PDP · cart · trust · offer · technical
Revenue at risk
≈ $17,200 / mo
Directional estimate from public signals
Expected recovery
≈ $12,850 / mo
Sum of itemized leak estimates
Six places a trade catalog loses the order
These are the checks the B2B twin would add to the six shipped parts. None of them is scored today — the scan reads a catalog with DTC weighting, which is useful but not complete.
The price list behind a wall
A form with no preview, no tier example and no idea what happens after submitting. The most common single reason a qualified buyer leaves without asking.
No quote path on the page
A product a buyer cannot request a price for from the page they are reading — the request lives in a contact form three clicks away, or in a rep’s email address.
MOQ and pack size in a PDF
Minimums, case contents and pallet quantities buried in a downloadable spec sheet instead of stated where the quantity is chosen.
Lead time unstated
No delivery window on the page, so the buyer has to ask before they can plan — and asking is a delay that competes with a supplier who answered on the page.
No reorder path
A returning buyer forced through discovery to re-find what they bought last quarter, instead of a reorder that starts from their own history.
Compliance proof missing
Certificates, test reports and warranty terms absent from the page where a compliance officer decides whether to escalate the supplier at all.
Where the revenue leaks — with the fix one click away
Revenue recovery turns the scan into a funnel: At risk → Recoverable → In execution → Recovered. Under it sits every leak with its source, priority, a monthly estimate or “Not estimated”, the size of the change, and an “Open fix” link that lands in the module able to close it. That machinery ships today; the B2B twin changes which leaks it can name, not how they are worked.
- The biggest leak is called out first; every other row is ranked by its estimate
- Impact by module shows each module’s configured estimate — labelled “typical, not measured” — and whether it is Active or Ready
- “Recovered” reads “Not measured yet” until measured attribution exists; it is never modelled from the estimates above it
Revenue recovery
Where your store is leaking revenue, backed by evidence — and the fastest path to recover it.
Recovery funnel
From estimated exposure to measured recovery — real data only
Estimated stages carry an ≈. “Recovered” stays “Not measured yet” until measured attribution exists — it is never modelled.
Where revenue leaks (4)
Ranked by estimate · source shown per rowNo sticky add-to-cart bar on mobile product pages
Copilotcritical priorityMobile add-to-cart rate 8.9% vs 13.4% on desktop; the buy box scrolls away on every product page.
Free-delivery threshold not shown inside the cart drawer
Copilothigh priorityShoppers leave the drawer without seeing how far they are from free delivery.
No active cart upsell or bundle logic
CRO scorehigh priorityOffer / AOV system scores 8 of 15 — no bundle offer on the product page, no add-ons under it.
Rating summary below the fold on product pages
Copilotmedium priorityTrust signals score 9 of 15; reviews exist but are not visible before the first scroll.
A report you can forward, a queue you can act on
The scan is the entry point. Its result is written to a public report page anyone can read without a login, and — once the store is connected — to the ranked action queue inside Comver.
The score, the breakdown and the leaks, at their own link
The CRO score ring, the six-part breakdown, the top leaks with their estimated lift, detected opportunities, missing conversion signals, what is already working, and an estimated recoverable range labelled directional. Useful for a supplier review with a buying committee that will not install anything.
AI Commerce Copilot Scan report
beautyskin.com
Scanned Aug 16, 2026 · homepage, catalogue, sample product page, cart
64
CRO score
Six weighted parts scored from public storefront pages. Target 70+.
Estimated recoverable revenue
$12k–$17k / mo
Score breakdown
This is a directional estimate based on public storefront signals. Full revenue recovery model requires analytics, Shopify and traffic data.
Top revenue leaks
Detected opportunities
- Add Bundling & Upsells
- Add Sticky ATC
- Improve PDP trust system
Missing conversion signals
- Free-delivery threshold
- Reviews above the fold
- Payment trust marks in cart
What’s already working
- Shopify detected · OS 2.0 theme
- Technical basics 9 / 10
- Clear homepage hero and navigation
- Product feed reachable
From URL to the first fix
Nothing to install for the public scan. Connect the Comver app when you want the score to feed Copilot, the Action Plan and Reports.
- 1beautyskin.comScan
Read-only · public pages only
Enter your storefront URL
The scan reads public pages only — the catalog, the product feed, a sample product page and the cart. It never signs in, so anything behind a trade login is invisible to it.
- 2Connect to the storefrontRead the storefrontScore the funnelQuantify and report
58% · usually 60–90 seconds
Four passes, 60–90 seconds
Connect to the storefront, read it, score the funnel, quantify and report — with screenshots captured along the way. The stage rail shows exactly where the scan is.
- 3Daily brief · after the scan
PDP structure scores 15 / 25 — mobile shoppers reach add-to-cart below the fold. Two low-risk fixes cover most of the gap.
CRO 643 actions in queue≈ $12,850 upsideCopilot writes the brief
The result becomes a score, top issues, opportunities, missing signals and a leak range. Inside Comver, Copilot reads the connected data and writes the daily brief and up to five recommendations.
- 41Sticky add-to-cart bar on mobile$6,400 / mo est. · Low changeOpen fix
Sticky ATC module · Prepare change · reversible
Open the fix
Each leak links to the module that closes it. Prepare the change, review its evidence and measurement plan, apply it reversibly, and watch the next scan move the score.
A public scan reads the public storefront
The most important limitation for a wholesaler: most of what matters is behind a login. Everything gated — the agreed price list, the account portal, the quote history — is invisible to a read-only scan, and the twin will not pretend otherwise.
Public only
What is read
Catalog, product feed, a sample product page and the cart. The scan never signs in, so a gated trade area returns nothing rather than a guess
DTC weights
How it scores today
The six parts and their weights are tuned for shopper behaviour; a B2B weighting and the B2B areas on this page are roadmap
≈
Revenue at risk
A directional estimate from public signals. Leaks without a monetary basis read “Not estimated” instead of carrying an invented figure
Once a store is connected, revenue at risk uses one canonical estimator everywhere in the product: the larger of the latest audit’s leak range and the sum of open recommendation estimates, capped by measured monthly revenue when it is known. If neither exists, the interface says “Awaiting data”.
Frequently asked questions
Yes — if the catalog is a public Shopify storefront, the scan will read it, score it and produce the report. What is not there yet is B2B scoring: price-list pages, quote forms, MOQ and pack-size tables and the reorder path are not among the areas it grades. Those are on the roadmap with no date announced.
Nothing. It is read-only: it fetches your public pages, takes screenshots and scores what a visitor would see. Nothing is applied, no theme is changed and no app permission is needed for the public scan.
Six weighted parts: homepage clarity (0–15 points), product page structure (0–25), cart and checkout friction (0–20), trust signals (0–15), offer and AOV system (0–15) and technical basics (0–10). Healthy starts at 70, and your last six scans are trended against that target.
No, and it should not. The scan never signs in, so anything behind a trade login — agreed prices, the account portal, quote history — is invisible to it. That is a real limit on how much a public scan can tell a wholesaler, and it is why the B2B twin also needs the connected account layer.
From the public scan alone it is a directional range based on storefront signals, and it is labelled that way. Once the store is connected, one canonical estimator is used across the product, capped by measured monthly revenue when known. If neither exists, the interface says “Awaiting data” rather than inventing a figure.
Not for the public scan — enter a storefront URL and you get the report. To open the fixes, see the recovery funnel, receive the daily brief and let the score feed the Action Plan and Reports, install the Comver Shopify app and connect the store.
Yes. Every scan has its own public report page — score, breakdown, top leaks, opportunities, missing signals and what is already working. Anyone with the link can read it; no login is required and nothing is installed on their side.
On demand, and the score history keeps your last six completed scans. Separately, the daily brief refreshes automatically every morning for every connected store.
Keep exploring
See the score before you change a thing
The scan reads a public Shopify catalog today, read-only and shareable. Join the B2B waitlist to hear when price-list pages, quote forms and MOQ tables become part of what it grades.
14-day free trial · Live on Shopify in minutes · Checkout stays in Shopify