SAMPLE: fictional store, not a client result. Store, figures and findings are invented to show the format of a readiness review. Example Outdoor Co. does not exist.

Sample readiness review: Example Outdoor Co.

Read-only scope: public pages and an exported feed. Sample prepared October 2026 by Kevin Hobbs, who designs and runs his own Medusa stores. A real review cites the store's own pages and lists the pages it sampled.

SAMPLE: fictional store, not a client result.

1. Summary

Agents can read most of this catalog but cannot reliably match variants or answer return and shipping questions. Three fixes close most of the gap: identifiers on variants, return and shipping markup, and feed-to-page parity.

Structured product data Amber
Product markup present; identifiers missing on variants
Policy markup Red
No return or shipping markup anywhere
Product feed Amber
Availability differs between feed and page
Agentic checkout programs Not assessed
Programs are gated; readiness only
Crawlability Green
AI-relevant robots.txt rules are explicit
Checkout predictability Amber
Tax shown late for one scenario

Green: no action. Amber: fix within 60 days. Red: blocks reliable reading or comparison. No numeric score is given because none could be sourced as a benchmark.

SAMPLE: fictional store, not a client result.

2. Findings

SAMPLE: fictional store, not a client result.

Structured product data

S1 · Product and Offer markup present (name, image, price, priceCurrency)

Observed (fictional): Present on 5 of 5 sampled pages.
Impact: Low
Fix: None
Effort: S
Confidence: High
Blind spot: Pages not sampled; markup injected by JavaScript after load may differ from raw HTML.

S2 · gtin present and valid

Observed (fictional): Missing on 5 of 5 sampled pages; only an internal sku is present.
Impact: High
Fix: Add a GTIN per variant from supplier or GS1 data where one is assigned. Mark custom-made items as having no identifier. Never invent identifiers.
Effort: M
Confidence: High for sampled pages; medium store-wide
Blind spot: Identifiers may exist in the platform or Merchant Center but not in page markup.

S3 · mpn and brand

Observed (fictional): brand present; no mpn.
Impact: Medium
Fix: Add mpn where the manufacturer has one.
Effort: S
Confidence: Medium
Blind spot: Whether a manufacturer mpn exists for private-label goods.

S4 · Per-variant Offer

Observed (fictional): One Offer for a 3-size, 2-colour product; variants not distinguishable.
Impact: High
Fix: Emit one Offer per variant with its own sku and availability.
Effort: M
Confidence: Medium
Blind spot: Theme or app behaviour at runtime; variant data behind JavaScript selectors.

SAMPLE: fictional store, not a client result.

Policy markup

P1 · hasMerchantReturnPolicy

Observed (fictional): Absent from sampled product and site-wide markup; a human-readable returns page exists.
Impact: High
Fix: Add return policy markup, preferably site-wide, consistent with the returns page.
Effort: S
Confidence: High
Blind spot: Return settings configured only in Merchant Center, which public inspection cannot see.

P2 · shippingDetails

Observed (fictional): Absent from sampled markup; the shipping page says free over a threshold.
Impact: High
Fix: Add shippingDetails matching the shipping page.
Effort: S
Confidence: High
Blind spot: Shipping rules configured in Merchant Center or by carrier at checkout.

P3 · Markup matches the policy pages

Observed (fictional): Cannot be compared until P1 and P2 exist.
Impact: Medium
Fix: Re-check after fixes.
Effort: S
Confidence: Low
Blind spot: Policies that change by region or promotion.

Google's merchant-listing documentation recommends gtin, brand, mpn, hasMerchantReturnPolicy and shippingDetails. The ChatGPT product feed specification has return fields and reads an omitted shipping price as unknown, not free.

SAMPLE: fictional store, not a client result.

Product feeds

F1 · Feed vs page availability

Observed (fictional): 4 of 20 sampled items are in stock in the feed export and out of stock on the page.
Impact: Medium
Fix: Sync feed refresh with inventory changes.
Effort: S
Confidence: Medium
Blind spot: Feed timing: a snapshot can differ from steady state.

F2 · Identifier coverage in feed

Observed (fictional): 62% of exported rows have a gtin (invented figure).
Impact: High
Fix: Fill gaps from supplier data where identifiers exist.
Effort: M
Confidence: Medium
Blind spot: Rows excluded from the export; Merchant Center diagnostics need read access.

F3 · Required fields in the ChatGPT product feed specification

Observed (fictional): All required fields populated in the sampled rows.
Impact: Low
Fix: None; add recommended fields over time.
Effort: S
Confidence: Medium
Blind spot: Whether this store has any ChatGPT feed relationship at all.

F4 · Merchant Center warnings

Observed (fictional): Needs read access to Merchant Center.
Impact: Unknown
Fix: Review with access.
Effort: S
Confidence: Low
Blind spot: Entirely invisible from public pages.

SAMPLE: fictional store, not a client result.

Agentic checkout programs

C1 · ChatGPT product feed and checkout

Observed (fictional): Not enrolled. Onboarding is limited to approved partners.
Impact: Informational
Fix: Clean data and policies first; apply only if the store chooses to.
Effort: S
Confidence: High on the gating wording as of the access date
Blind spot: Whether the program's terms have changed since the access date.

C2 · Google checkout in AI Mode (Universal Commerce Protocol)

Observed (fictional): Not enrolled. Limited to selected merchants through an interest form.
Impact: Informational
Fix: Same as C1.
Effort: S
Confidence: High on the gating wording; treat the US as the only safe assumption
Blind spot: Merchant Center eligibility attributes are not publicly visible.

No outcome is promised: a review does not enable, apply for or guarantee approval into any program.

SAMPLE: fictional store, not a client result.

Crawlability for AI agents

R1 · robots.txt rules

Observed (fictional): Product and policy paths allowed; cart and account blocked.
Impact: Low
Fix: None.
Effort: S
Confidence: High
Blind spot: Firewall or CDN bot rules that never show in robots.txt.

R2 · Product data in raw HTML

Observed (fictional): Product markup present without running JavaScript.
Impact: Low
Fix: None.
Effort: S
Confidence: Medium
Blind spot: Rendering behaviour of specific agents is unknown.

R3 · Sitemap, canonicals, status codes

Observed (fictional): Sitemap lists products; 3 sampled URLs return 200 with correct canonicals.
Impact: Low
Fix: None.
Effort: S
Confidence: Medium
Blind spot: Pages outside the sample.

R4 · llms.txt (unofficial)

Observed (fictional): Absent. llms.txt is an unofficial community proposal.
Impact: Low, informational
Fix: Optional; no benefit is claimed.
Effort: S
Confidence: Low
Blind spot: No confirmed consumer of the file.

SAMPLE: fictional store, not a client result.

Checkout predictability

K1 · Agreed purchase scenarios run end to end

Observed (fictional): In 1 of 3 invented scenarios (international address), tax appeared only at the final step.
Impact: Medium
Fix: Show tax and shipping earlier, or document the behaviour.
Effort: M
Confidence: Medium
Blind spot: Payment behaviour in production; real card flows are not exercised.

SAMPLE: fictional store, not a client result.

3. 30/60/90 plan

30 days: S2, S4, P1, P2
Owner: Store team (PrimeCommerce can apply approved fixes as store operations)
Verify by: Re-sample pages; identifiers and policy markup present
60 days: S3, F1, F2, R1
Owner: Store team and feed owner
Verify by: Feed export matches pages on a re-sample
90 days: K1, C1, C2 review, monitoring cadence
Owner: Store team and PrimeCommerce
Verify by: Scenarios re-run; program pages re-read and dated

SAMPLE: fictional store, not a client result.

4. Notes

Everything above is invented. A real report cites URLs and snippets from the store's own pages.
Findings come from public pages and an exported feed. Merchant Center settings and JavaScript-injected markup can change any row; each row names its blind spot.
Program and specification wording was checked against the providers' own pages in October 2026.

Want this for your store?

Request a readiness review

Opens your email app with a few short questions filled in. Or write to kevin.hobbs@primecommerce.com