The report card for AI shopping

What grade would your store get from an AI shopper?

Q: Does GradeMY access checkout or customer data?

No. The default scan reads public storefront pages and public metadata only. It does not access checkout, carts, accounts, payments, customer records, private orders, or admin screens.

Q: What does a GradeMY report include?

A completed report shows the grade and evidence behind it, observed blockers, developer remediation when supported, and the report artifacts emitted by the scan.

GradeMY tests what AI systems can actually verify about your products—then shows what lowered your grade, the evidence behind the deduction, and exactly what your developer should fix.

Start your grading↘

✓ No account required ✓ Public storefront only ✓ Read-only ✓ No checkout or admin access

View a full controlled sample report →

YOUR STORE GRADE

CONTROLLED FIXTURE
GRADEMY SAMPLE

85 / 100

Strong public product evidence.
One critical claim is holding the shopping task back.

Grade breakdown	Grade	Points
Product discoveryCan the shopper find the product?	A	20 / 20
Price clarityCan price and currency be verified?	A	20 / 20
Variant clarityCan the requested option be identified?	B	18 / 20
AvailabilityCan availability be verified?	F	9 / 20
Agent task evidenceCan the buyer request be completed?	B	18 / 20

View full sample report →

Observed evidence, not guesses Pass / partial / fail buyer tasks Developer-ready corrections Regrade after the fix

Here’s what GradeMY finds

Every point should have evidence behind it.

The report is not a mystery score. GradeMY shows the public product evidence it observed, the exact claim that failed, and the correction that can change the result.

Controlled fixture product

Trail Runner X

$109.00

Color: Black

Size

910111213

Add to cart

Price verified$109.00✓ public evidence

Variant verifiedSize: 11 · Color: Black✓ matched

Availability not verifiedNo machine-verifiable availability claim found.✕ blocker

The GradeMY rubric

See exactly where your grade comes from.

Each category is tied to evidence collected from the sampled public storefront surface. Untested capabilities are labeled as such rather than silently treated as a pass or failure.

Category	What GradeMY checks	Example grade
Product discovery	Public product paths, collection paths, crawl hints, sitemaps, and accessible product pages.	A
Product identity	Title, description, image, identifiers, brand, and variant evidence.	A
Offer clarity	Price, currency, offer URL, and whether required offer facts are machine-readable.	B
Availability	Whether the requested item’s availability can actually be verified.	F
Buyer-task fit	Whether captured evidence satisfies the constraints in the shopper’s specific request.	B

How it works

Grade. Correct. Regrade.

The workflow is intentionally simple for a merchant and specific enough for the developer who has to fix the defect.

⌨

Submit your store

Enter a public store or product URL and optionally describe a real shopping request.

◎

Test the evidence

GradeMY samples the public storefront and checks what an AI shopping workflow can verify.

✓

Get your grade

See the evidence, deductions, decisive blocker, and developer remediation when supported.

↻

Fix and regrade

Rerun the same public test after the correction to verify what changed.

For merchants and agencies

A grade the client understands. Evidence the developer can use.

GradeMY keeps the business-facing result simple while preserving the technical evidence required to defend the finding and implement the fix.

Merchant view

Why did we lose points?

See the grade, the failed buying claim, and the evidence behind it without reading a schema audit.

Grade breakdown
Observed deductions
Buyer-task result
Regrade after correction

Agency / developer view

What exactly do we change?

Move from the same result into evidence provenance, remediation, acceptance tests, and emitted report artifacts.

Page-level evidence
Narrow developer remediation
Developer tickets when supported
HTML/PDF artifacts when emitted

Technical methodology

The grade is the summary. The evidence is the product.

GradeMY remains deliberately technical beneath the human-facing report card. Agents, engineers, and reviewers can inspect the same visible methodology and boundaries.

Read the full methodology →

Agent Commerce: GradeMY evaluates public product evidence used for product discovery and selection. The default scan stops before cart, checkout, payment, account, and order placement.
Product + Offer data: Structured Product and Offer evidence, pricing, currency, variants, availability, identifiers, images, and public page context are evaluated when present.
AEO / GEO: Adjacent optimization contexts are documented, but GradeMY does not claim general answer-engine rankings, citations, recommendations, traffic, or revenue outcomes.
WebMCP: Reported only when the surface was actually evaluated. Experimental, missing, or untested interfaces do not create unrelated product-evidence defects.
Evidence boundary: Default scans use sampled public pages and metadata. Untested capabilities are labeled not evaluated; missing evidence is not turned into invented certainty.

Before you grade

Two useful boundaries.

Does GradeMY access checkout or customer data?

No. The default scan reads public storefront pages and metadata only. It does not access checkout, carts, accounts, payments, customer records, private orders, or admin screens.

What does the report include?

The grade and evidence behind it, observed blockers, developer remediation when supported, and links to the artifacts the completed scan actually emitted.

Ready to grade your store?

Start with the public storefront. See what passes, what loses points, and what to correct.

Grade my store →