Entity coherence, Product truth and claims in AI search
Two Products share a similar name. A founder's biography is confused with the company profile. A regional service is described as nationally available. A discontinued offer remains associated with the Brand.
The facts may exist. The entity relationships are unclear.
Entity coherence is the degree to which important organisations, Brands, Products, people, services and locations are consistently identified and correctly related across sources.
Parent and subsidiary information is blended, Product names change without clear supersession, adviser identities are inconsistent, regional availability is omitted, claims appear without their conditions, similar services are merged, and legacy pages remain technically accessible and externally referenced.
The organisation needs approved names and descriptions, clear relationships, current fees, conditions and limitations, owner and effective date, canonical sources, accessible supporting evidence, and correction routes for partner and third-party sources.
A financial group has a parent Brand, a retail bank, a wealth division and several advisory Products. Public sources use the group name and Product names inconsistently. AI answers sometimes attribute the wealth service to the retail bank or combine adviser biographies from different divisions.
The organisation may need a clear entity hierarchy, consistent names and descriptions, relationship statements, distinct canonical pages, current people and role information, structured data that reflects genuine relationships, correction of high-value third-party sources, and clear regional and Product boundaries.
A statement may be factually true but misleading when the qualification is removed. For example, "No monthly account fee" may require a condition about account type or eligibility. If the condition repeatedly disappears in summaries, the source architecture needs a clearer public explanation—not merely a hidden disclaimer. Narrative Integrity therefore examines whether the answer preserves the context needed for the customer decision—not only whether isolated words are accurate.
Entity confusion can leave the customer unsure who provides the service, who is accountable, whether the Product is available to them, which adviser or company they will engage, or which terms and reputation evidence apply — that uncertainty can weaken Trust even when individual facts are mostly correct.
An Entity and Product Truth Register can preserve approved names, relationships, descriptions, markets, owners, effective dates, claims, qualifications and canonical sources. It does not need to be public. It gives Content, SEO, Data, Product and assurance teams one governed reference when sources are created or corrected.
Product names and approved descriptions, parent, Brand and subsidiary relationships, adviser, author and leadership identity, locations and markets, availability and eligibility, claims, evidence and qualifications, legacy names and discontinued Products, and canonical source and supersession.
Clear entity architecture can improve the source environment. It cannot guarantee that every external system will interpret the relationship correctly. The organisation should record the changed source condition and re-observe the representation.
Protect the relationship between fact and entity.
Explore AI Narrative Integrity, or return to the cornerstone framework.
Use search, review, AI and customer evidence to improve Product truth, Brand trust, service…
See how public sources, entity consistency, evidence provenance and AI representation shape…
Search the complete approved SI definitions for architecture, customer decisions, evidence,…