Glossary

Entity optimization (for AI answers)

Entity optimization is the discipline of shaping how machines recognise your brand as a structured object — IDs, facts, disambiguation — so answers stay consistent across models.

Same spirit as Knowledge Graph SEO, but tuned for generative retrieval and citations.
Pairs with schema.org, Wikidata-style references, and authoritative bios.

Definition

Entity optimisation focuses on the identity layer of your brand: official name variants, stock tickers, founder names, HQ locations, product lines, and how those facts appear in structured data and trusted third-party profiles. When entities are messy, models merge you with homonyms, invent subsidiaries, or swap pricing tiers. Cleaning entities raises the odds that RAG pipelines retrieve the correct chunk and that evaluators mark answers as wins.

How it's computed

There is no public formula — progress is observed indirectly via mention detection, duplicate entity counts in answers, and hallucination rates on identity prompts. Practitioners track consistency scores across fanout queries once canonical facts are published.

How it works in practice

Concrete tasks

Publish Organization / Product JSON-LD with stable @id URLs.
Align Wikipedia/Wikidata, Crunchbase, App Store, and marketplace listings.
Maintain a “do not confuse with” note for similarly named competitors.

How to read it

If LLM-Score is volatile but content is strong, run an entity audit before rewriting articles.

When to use

After mergers, rebrands, or multi-brand portfolios.
When models keep inventing offices you do not have.
When launching in a new language market with duplicate transliterations.

AEO (Answer Engine Optimization)RAG (Retrieval-Augmented Generation)Hallucination LLM-Score

Check identity prompts in a scan