Claim Accuracy

Answer engines behave like accuracy and consensus machines: they favor claims they can verify and corroborate, and they deprioritize unsupported or overstated assertions. Accuracy is therefore not just ethics — it is an AEO signal. This skill makes a brand's content verifiable, internally consistent, and free of overclaims, and fixes the cases where engines repeat something false about the brand. It works for any business, B2B or B2C, and pairs tightly with evidence-and-citations (adding proof) and entity-optimization (consistent facts).

When to use this skill

• Fact-checking content meant to be cited by AI.
• Removing overclaims and unsupported assertions that engines discount.
• Reconciling facts that differ across a brand's pages or profiles.
• Correcting inaccurate or outdated things AI assistants say about the brand.

Before you start

Read aeo-foundation.md for the canonical facts, claims, and any compliance constraints (especially YMYL — health, finance, legal). If it is missing, point to the aeo-foundation skill.

Principles

• Accuracy is a trust signal. Verifiable, correct content is favored; a detectable error or overclaim erodes how readily engines cite the source.
• Support every strong claim. Unsupported confidence backfires — pair claims with a source, a statistic, or a clear basis (see evidence-and-citations), or soften them.
• Be consistent everywhere. The same fact should read the same across your pages and your off-site profiles; contradictions inject doubt and weaken entity recognition.
• No hype inflation. "Best," "leading," "guaranteed," and invented numbers without backing are liabilities, not assets.
• Correct, transparently. When something changes or was wrong, fix it and date it.

Workflow

1. Inventory claims. Pull every checkable assertion — numbers, superlatives, comparative statements, dates, capabilities, guarantees.
2. Verify each. Confirm against a primary or authoritative source. Cite it, correct it, or soften it. Anything you cannot support gets reworded or removed.
3. Check internal consistency. Ensure the same facts (pricing, dates, capabilities, results) match across all of the brand's own pages.
4. Check external consistency. Reconcile against the brand's other profiles and listings — hand the cross-web check to entity-optimization (and its consistency tool).
5. Strip overclaims. Replace unbacked superlatives and vague boasts with specific, supported statements — or cut them.
6. Detect and fix AI misrepresentations. Test what the major engines currently say about the brand (see ai-visibility-audit); for each error, fix the upstream source the engine is likely drawing on (your page, a profile, a third-party listing) and, where possible, the authoritative record.
7. Keep a source of truth. Maintain the canonical facts in aeo-foundation.md so future content stays consistent.

Output format

• A claims audit: each checkable claim marked verified / needs-source / overclaim, with the fix.
• Inconsistencies found across pages or profiles, and how to reconcile them.
• AI misrepresentations observed, with the upstream source to correct for each.
• (assumption) tags where a fact could not be verified.

Common pitfalls

• Superlatives and invented numbers with no backing — discounted by engines.
• Stale statistics or dates that make a page demonstrably wrong.
• The same fact stated differently on different pages.
• Ignoring what AI already says incorrectly about the brand.
• "Fixing" content while leaving the third-party source of the error untouched.

Related Skills

• evidence-and-citations — the sources and data that turn a claim into a supported one.
• entity-optimization — consistent canonical facts across the web.
• ai-visibility-audit — surfaces what engines currently say (and get wrong).
• aeo-foundation — the source of truth for the brand's facts and claims.