One product. Many channels. One source of truth.
Product Truth is a structured, sourced representation of the product—what it is, what is being claimed, where each fact came from, what remains uncertain, and how the information can be expressed across commerce channels.
Product information breaks when every team rebuilds it independently.
Packaging, PDPs, feeds, ads, marketplaces, and AI shopping systems ask different questions about the same product. Without a shared representation, facts drift, claims lose context, sources disappear, and every channel review starts over.
What Product Truth carries
- Canonical product identity
- Taxonomy and normalized attributes
- Ingredients, materials, quantities, or composition
- Exact claims and their qualifiers
- Evidence or substantiation supplied for those claims
- Applicable category and channel context
- Provenance for each product fact
- Confidence and uncertainty
- Conflicts between sources
- Version and change history, where supported
Every important fact should be able to answer: “According to what?”
A product record should not merely say that a product contains 500 mg of an ingredient. It should identify whether that information came from the label, a facts panel, a certificate, a manufacturer record, a feed, or an inference—and whether the available sources agree.
When sources conflict, Product Truth surfaces the conflict instead of quietly choosing one.
Govern the knowledge. Change the expression.
Preflight
Check the supplied product record before launch.
Channel expression
Adapt product language and fields to the selected destination without changing the underlying facts.
Suppression recovery
Connect a notice, affected content, correction, and supporting evidence.
Structured data
Produce reusable product fields and schema where the sources support them.
AI shopping readiness
Make the product easier for machine systems to parse and compare.
Important distinction
Product Truth does not mean that every statement is independently verified. It means that facts, claims, sources, inferences, conflicts, and recommendations are represented distinctly and transparently.