INDEPENDENT MERCHANT INTELLIGENCEPUBLIC BETA / SEPTEMBER 2026 BASELINE
Original research / Methodology first

Product data, under inspection.

A study design for identity, offers, variant clarity and public policy evidence.

Research draft / No market-wide findings

Research question

Which product facts survive from the catalog to public evidence?

Proposed method

Inspect stable identity, Product/Offer fields, variant relationships and policy links within the pilot’s bounded sample. Retain a denominator for each check and record static-rendering limitations.

What this cannot establish

Field presence is not field accuracy. Sampled pages do not establish full-catalog health, channel acceptance or causal impact on discovery.

Publication gate — Research HOLD

Validate source scope, check unresolved conflicts, document the sample and review conclusions against the underlying observations. A finding may be published only at the level its evidence supports. Incomplete studies remain drafts, noindex and outside the production sitemap.

Evidence ledger

Go to the source.

Primary documentation records what is publicly stated. It does not independently prove outcomes or your account’s eligibility.

  • Product vocabulary

    Schema.org · official primary · Checked 2026-09-28

    Claim scope: Product properties

    Vocabulary presence is not validation, eligibility or certification.

    Source ID: schema-product · Published: Not stated

  • Product structured data

    Google Search Central · official primary · Checked 2026-09-28

    Claim scope: Product markup and merchant listings

    Markup is one input and does not guarantee search appearance.

    Source ID: google-schema · Published: Not stated

  • File upload overview

    OpenAI · official primary · Checked 2026-09-28

    Claim scope: Full snapshot feed delivery

    At least daily delivery recommended; not a scanner protocol endpoint.

    Source ID: openai-feed · Published: Not stated

Dispute a fact or suggest a source