6 of 19 schema-page-evaluable storefronts had a deterministic public-surface conflict.
The report measures disagreement between store-controlled public surfaces. It does not judge legal sufficiency, merchant trustworthiness, or whether AI agents can answer a policy question.
For this report, StoreSteady evaluated a frozen corpus of 100 Shopify storefronts. Freeze: 2026-05-21. Evaluation window: 2026-05-26T02:47:00.000Z to 2026-05-26T02:48:44.361Z.
Run a store-specific scan for product, offer, shipping, return, contact, business identity, and trust-surface consistency. The result is evidence for your store, not a projection from the benchmark.
See whether your schema and pages disagree- Frozen corpus tranche
- 100
- Rank-stratified Shopify storefronts selected from the frozen external_top_500 corpus.
- Blocked or noindex
- 75
- Stores blocked/noindexed before enough public field evidence could be compared.
- Not evaluable after access
- 6
- Stores reached by the crawl but without enough field-level counterparts for a store-level classification.
- Schema-page evaluable
- 19
- Stores with enough visible and/or structured public evidence for at least one field comparison.
- Deterministic conflict
- 6
- Stores where two public store-controlled surfaces deterministically disagreed.
Published metric quality
| Metric | Tier | Evaluable N | Finding N | Flagged N | Non-finding N | Rate | Wilson CI | Quality | Caveat |
|---|---|---|---|---|---|---|---|---|---|
Deterministic schema-vs-page conflict deterministic_schema_page_conflict | Headline | 19 | 6 | - | - | 31.6% | 15.4% - 54.0% | deterministic | Deterministic field comparison only; missing counterpart states are reported separately and are not legal sufficiency or trustworthiness claims. |
Visible evidence missing structured counterpart missing_structured_counterpart | Supporting | 19 | 11 | - | - | 57.9% | 36.3% - 76.9% | deterministic | Counts stores where visible public evidence existed for at least one field but the structured counterpart was missing. |
Structured evidence missing visible counterpart missing_visible_counterpart | Supporting | 19 | 1 | - | - | 5.26% | 0.94% - 24.6% | deterministic | Counts stores where structured public evidence existed for at least one field but the visible page counterpart was missing from the scanned public evidence. |
Blocked or noindex before field comparison blocked_or_noindex | Limitation | 100 | 75 | - | - | 75% | 65.7% - 82.5% | deterministic | Separated from findings. Blocked/noindex stores are not counted as schema-vs-page conflicts. |
Not evaluable after access not_evaluable_after_access | Limitation | 100 | 6 | - | - | 6% | 2.78% - 12.5% | deterministic | The crawl reached public evidence but did not collect enough field-level counterparts to classify a store-level finding. |
The useful next step is checking whether your own public schema and page evidence agree.
Run a store-specific scan for product, offer, shipping, return, contact, business identity, and trust-surface consistency. The result is evidence for your store, not a projection from the benchmark.
See whether your schema and pages disagreeDetector primitives
Product visible-vs-structured counterpart
15 consistent, 3 missing structured counterpart, 1 not evaluable field row
Field-row counts are among the 19 schema-page-evaluable stores. Compares visible PDP evidence with Product/ProductGroup markup where a PDP was collected.
Offer price, availability, URL, and image consistency
6 deterministic conflicts, 6 consistent, 3 missing structured, 3 missing visible, 1 not evaluable field row
Field-row counts are among the 19 schema-page-evaluable stores. Counts deterministic conflicts only when public page and schema/URL/image signals disagree under exact parser rules.
Return policy counterpart
2 consistent, 12 missing structured, 1 missing visible, 4 not evaluable field rows
Field-row counts are among the 19 schema-page-evaluable stores. Compares visible return/refund policy pages with MerchantReturnPolicy or nested return policy markup.
Shipping counterpart
3 consistent, 14 missing structured, 2 not evaluable field rows
Field-row counts are among the 19 schema-page-evaluable stores. Compares visible shipping policy pages with OfferShippingDetails markup.
Contact and business identity counterpart
21 consistent contact/identity field rows, 10 missing structured counterparts, 3 missing visible counterparts, 4 not evaluable field rows
Combines two field rows per evaluable store: contact evidence and business identity evidence. Compares contact-page/footer evidence with Organization, LocalBusiness, Store, and OnlineStore markup.
Methodology summary
Selection rule: Deterministic rank-stratified 100-store tranche from the frozen external_top_500 Shopify storefront corpus; no merchant-identifying examples are published.
What StoreSteady Measured
StoreSteady compared public machine-readable commerce data against visible public page evidence across product, offer, shipping, return, contact, and business identity fields.
The state contract was consistent, missing_structured_counterpart, missing_visible_counterpart, deterministic_conflict, blocked_or_noindex, and not_evaluable.
- Deterministic conflicts require two store-controlled public surfaces to disagree.
- Missing structured or missing visible counterparts are reported separately from conflicts.
- Blocked/noindex and not-evaluable stores are not treated as findings.
What StoreSteady Did Not Measure
The report does not decide whether a policy is legally sufficient, whether a merchant is trustworthy, or whether an AI agent can answer a shopper question.
It also does not use private Merchant Center, Shopify admin, customer, order, checkout, or feed evidence.
Quality Gates
All published rows are deterministic parser rows. Judge alpha and calibration accuracy are not applicable because no LLM judgments were used.
The 1-3 store smoke validated the state contract before the rank-stratified 100-store tranche. The headline denominator is schema-page-evaluable stores, while access limitations use the full scanned corpus denominator.
Limitations
- Dynamic pricing, localization, sale timing, member pricing, and storefront personalization can make some fields not evaluable without being conflicts.
- The widened tranche still had high blocked/noindex exposure, so the headline denominator remains narrower than the full frozen corpus.
- A missing counterpart is not a policy correctness claim and is not counted as a deterministic conflict.
- The report publishes aggregate counts only and omits store names, domains, raw URLs, source excerpts, exact policy quotes, promo codes, and merchant-identifying examples.
Artifacts
Artifacts are aggregate-only and include no store names, raw domains, raw URLs, raw crawl text, raw AI responses, exact promo codes, private feed/admin evidence, or merchant-identifying examples.
Correction policy
Methodology gaps, reproducible errors, or source corrections can be sent to research@storesteady.com. Corrections are published with the version history in the methodology page and report JSON.