Entity Consistency Index
Also known as: Indice de Cohérence d'Entité
The Entity Consistency Index (ECI) is a proprietary Rankeo score (0-100) that measures how coherently an entity (person, organization, product) is declared across all its digital surfaces. A high ECI signals to AI engines that the entity is a single, verifiable, trustworthy source. A low ECI creates ambiguity and entity confusion, which AI engines penalize by lowering citation priority.
Definition
The ECI is calculated from 7 weighted signals: schema consistency across domains (20%), name uniformity (15%), Wikidata Q-item presence (15%), sameAs cross-linking density (15%), founder/author attribution consistency (10%), logo and visual identity uniformity (10%), and canonical URL hygiene (15%). Each signal can be measured and improved independently.
The ECI exists because AI engines have shifted from keyword-matching to entity-matching since late 2025. They no longer just "read" pages — they ingest them into a Knowledge Graph and ask "is this entity the same as that one?". If your answer is inconsistent, you fragment yourself and lose citation weight.
How It Works
AI engines maintain an internal entity graph where each concept, person, and organization is a node with attributes (name, type, creator, sameAs, identifiers). When the engine encounters new content, it performs entity resolution: "is this Jonathan Jean-Philippe the same Jonathan Jean-Philippe I already know?"
If your schemas across rankeo.io, dealpropfirm.com, thelegalprompts.com and theplanettools.ai all declare the same founder with the same @id and the same sameAs array, the resolver returns "same entity" with high confidence. Your Entity Consistency Index is high. All citations on any of the 4 domains reinforce the same entity.
If your schemas are inconsistent (different founder names, different @ids, missing sameAs) the resolver returns "different entities" or "ambiguous entity". Your ECI is low. Citations fragment across multiple weakly-connected entities, and none of them accumulate enough weight to dominate.
Practical Example
Before measuring ECI, a consultant had 4 sites with 3 different LinkedIn URLs declared across schemas, 2 different photo URLs declared across OG tags, and a founder name formatted 4 different ways ("John Doe", "J. Doe", "John R. Doe", "doe_john").
ECI score: 38/100.
After applying a consistency pass (unified founder schema, single LinkedIn, normalized name, cross-referenced sameAs), ECI rose to 89/100. Within 6 weeks, citations on their main domain increased 47% and their founder became reliably identifiable by ChatGPT and Perplexity with full attribution.