Entity Coherence

Entity coherence is the degree to which the signals an entity controls fit together into one unambiguous identity: the same name, a clear type, matching attributes and plausible relationships across its website, markup, social profiles and listings. It is the internal side of entity work. Entity consensus is the external side, asking whether independent sources agree. A brand can be coherent and still lack consensus if nobody else writes about it, and it can have plenty of coverage while being incoherent if its own channels contradict each other.

Coherence covers more than spelling the name consistently. It includes:

• Type clarity: is this a company, a product, a person or an event? An entity that looks like several types at once is harder to classify.
• Attribute agreement: founding year, headquarters, headcount band, leadership and category match everywhere the entity speaks for itself.
• Relationship sense: products link to the company that makes them, people link to the organizations they work for, and none of those links point at a namesake.
• Description stability: the one-line description of what the entity does is the same on the entity home, LinkedIn, app stores and directory profiles.

There is evidence that type clarity matters to Google. Kalicube, which tracks Google's Knowledge Graph, reported that the June 2025 updates removed more than 3 billion entities, many of them carrying the vague 'thing' type, and raised the share of person entities with a single clear type from 70.16% to 76.78%. Google has not confirmed the reasons, but the direction is toward entities it can classify confidently.

Relationships also matter for retrieval. Google's May 2026 guide to generative AI in Search describes query fan-out, where the model issues several related sub-queries for one prompt. A prompt about 'tools for automating invoices in Xero' might fan out into sub-queries about integrations, pricing and reviews. An entity that is clearly connected to the right integrations, categories and use cases can match more of those sub-queries than one described only in its own marketing language.

Incoherence shows up in AI answers as mixed facts, a product attributed to the wrong company, an old name, or a category the brand left years ago. To audit it, list the core facts and the one-line description, then compare them across every property the brand controls, including old press pages, app store listings and partner directories that are easy to forget. Fix the entity home first, then the schema markup, then the profiles. Coherence is one of the inputs to brand depth and to entity SEO generally.