Claim Registry

What it is

A claim registry treats a public statement the way a product catalogue treats a SKU: as a record with fields rather than a sentence buried in prose. Each entry holds the approved wording, a link to the evidence behind it, the person accountable for it, and the date that evidence was last confirmed. The registry is not a content store — the sentence still lives on the page — it is the index that says which sentences are load-bearing.

Why it matters

Other categories of company fact already have registries. Master data management owns entity records, product information management owns product attributes, and regulated industries own promotional claims. Claims written in prose — pricing, policy, capabilities, certifications, metrics — are maintained by whoever last remembered to update the page, and they are also the claims a customer actually reads. The registry closes that gap. Knowledge Company publishes its own at knowledgecompany.ai/claims, including the claims it has not yet re-verified. That register does not yet meet this definition in full: every row carries the wording and both dates, but four rows hold no source — one records that the origin of its claim could not be located, and three record that the page making the claim still has an empty citation slot — and an owner is named only on the rows drawn from this dictionary and on one statement about the company itself, not on the rows drawn from the blog or from competitor comparisons; even there the owner is appended to the source field rather than held in a column of its own.