Google's conversational attributes need a merchandising rule

Set eligibility before writing the fields

Create a conversational-attribute eligibility register before generating or uploading values. For each of Google's six optional fields, name the business purpose, accountable semantic owner, acceptable evidence, permitted source systems, review standard, sensitive or regulated subjects that are out of scope, and the person authorized to submit. A question-and-answer field can carry an explicit brand statement; a document link can import a longer technical record; a related-product field expresses a relationship; and a popularity rank introduces a comparison. Those are different assertions and should not inherit one blanket approval. Keep the primary catalog and these optional contextual fields as distinct records so adding conversational context does not silently redefine the organization's authoritative data.

Define the meaning of every relationship

Write controlled definitions and examples for relationship-bearing values. A related product should state the approved relationship type and evidence behind it; an item-group title and variant option should use a stable naming rule; a popularity rank should identify the inventory population and calculation method; and an FAQ answer should distinguish a documented fact from an editorial explanation. Test ambiguous labels, unsupported superlatives, obsolete manuals, optional accessories, incompatible variants, empty denominators, and language that could imply a sensitive personal trait. The CMO owns the brand meaning and review standard, while merchandising, product, technical, accessibility, privacy, and legal owners retain authority over the underlying subject. Fluency or schema validity alone does not make a contextual assertion appropriate.

Approve a versioned submission package

For each approved package, preserve the item identifier, attribute name, exact value, supporting source, source owner, rationale, reviewer, submission method, and package version. Google's help page says the fields may enter through a supplemental source, the primary source, or the Merchant API, and warns against duplicating details already supplied elsewhere. The release record should therefore show which path is authoritative for the optional context and which team can change it. Use synthetic examples to confirm that separators, identifiers, document links, grouped values, and rank calculations follow the published schema. This is an input-governance artifact; it does not certify what an AI system will say, establish downstream publication parity, or replace the existing change ledger for live marketing assets.

Interpret the provider test narrowly

Do not turn Google's reported 50% incorporation rate from one lululemon test into an accuracy, visibility, or performance promise. Before a limited release, define an eligible attribute set and a question sample, then record whether the submitted context was available, relevant to the question, represented without changing the approved meaning, and distinguishable from information Google supplied from elsewhere. Report missing context, unsupported extrapolation, conflicting fields, and reviewer uncertainty separately. The CMO decision is whether the organization has a defensible vocabulary and evidence process for optional conversational context. Question-to-conversion attribution, booking-stage measurement, commercial-term parity, transaction reconciliation, downstream-asset retirement, and platform-answer correction remain separate decisions.

Turn this source into a reviewable decision

For AI for CMOs, use this briefing as a dated decision record rather than a substitute for the source. Preserve Google, the exact URL, the September 17, 2026 review date, the supported facts above, the editorial interpretation, the limitations, and any buyer-specific evidence. Link that record to the decisions most directly affected: Brand, disclosure, and synthetic-media risk; Commerce and conversion assistance; Customer journeys and personalization. State whether the source changes the scope, evidence requirement, control, sequence, or only the language used to describe the decision.

Before action, name the accountable owner, affected population and workflow, exact offering or configuration, source data and rights, human decision point, exception and appeal path, complete cost, expected benefit, failure and stop conditions, retained evidence, and next review date. Keep official facts, provider statements, buyer observations, representative tests, measured outcomes, editorial inferences, and unknowns visibly separate. Reopen the record when the source, offer, model, integration, data, policy, population, responsible person, or measured result changes.

Limitations and unknowns

Google is the provider source. Its announcement carries JSON-LD datePublished metadata of 2026-09-16T14:00:00+00:00, after the 2026-09-16T12:52:36Z terminal cutoff, and the conversational-attribute capability described is materially new to this run. The announcement and linked official help page support Google's current description of optional conversational attributes, the six listed field types, submission paths, non-duplication guidance, and the reported lululemon test result. They do not independently establish a merchant's access, configured values, source rights, semantic accuracy, review quality, AI use of a field, recommendation accuracy, visibility, customer understanding, conversion, revenue, or legal conclusion. Current Merchant Center configuration, versioned submission packages, authoritative supporting records, controlled vocabulary, representative question tests, and qualified marketing, merchandising, product, analytics, privacy, security, accessibility, intellectual-property, and legal review control.