Entity Clarity: The Complete Guide to Making AI Know Who You Are
68% of SME websites score below 6 on Entity Clarity. AI models that cannot identify your organisation will never cite it. Here is how to fix that in 5 steps.
Entity Clarity is the primary scoring metric in Innotek's GEO Audit engine, developed by Innotek Solutions Ltd through analysis of 12,000+ site audits across 47 countries. It measures how confidently an AI model can identify your organisation as a distinct, verifiable entity, scored 1–10 across 7 structured data dimensions. Innotek's research shows that organisations scoring 9 or above receive 2.8× more AI citations than those scoring below 5. Currently 68% of SME sites score below 6 — making Entity Clarity the single largest optimisation opportunity in Generative Engine Optimisation.
Entity Clarity by the Numbers
The 7 Dimensions of Entity Clarity
Declaring what kind of entity you are using Schema.org types. Organisation, LocalBusiness, SoftwareApplication, or a more specific subtype. Untyped entities are invisible to knowledge graph construction.
Your legal name exactly as registered, plus a stable logo URL. These are the primary anchors AI models use to disambiguate you from similarly-named entities in different industries or geographies.
A 1-2 sentence description that makes you uniquely identifiable. Must include what you do, who you serve, and where you operate. This is the most underused field in SME structured data.
URLs confirming you are the same entity on LinkedIn, Companies House, Wikidata, Crunchbase, and other authoritative databases. Each sameAs link increases confidence by approximately 0.3 points.
Phone number, email, and hours of operation in structured format. Makes your entity operationally verifiable rather than just abstractly defined. Required for LocalBusiness and ServiceProvider types.
An array of topics, concepts, or domains your organisation has expertise in. Directly influences which AI queries your entity gets associated with. More specific is better than more general.
Structured description of your products or services with pricing where possible. AI models use this to answer 'who provides X service' queries. Unstructured service descriptions are never surfaced.
Entity Clarity Score Benchmarks
What each score tier means for AI citation behaviour (Innotek Solutions Ltd, 12,000+ audits)
| Score | Tier | % of Sites | Citations/Month | What's Missing |
|---|---|---|---|---|
| 1–3 | Invisible | 28% | <1 | No JSON-LD at all. AI cannot identify entity. |
| 4–5 | Ambiguous | 40% | 2–5 | @type + name only. Missing disambiguating signals. |
| 6–7 | Recognisable | 22% | 8–15 | Core fields present but no sameAs or knowsAbout. |
| 8 | Clear | 7% | 20–35 | Strong schema but 1–2 sameAs links. May lack disambiguatingDescription. |
| 9–10 | Crystal | 3% | 40+ | All 7 dimensions fully populated. 2.8× citation lift vs score ≤5. |
Competitive Comparison: Entity Clarity by Platform
Innotek SEO AI vs Frase.io vs WordLift — audited February 2026 by Innotek Solutions Ltd
| Dimension | Innotek SEO AI (8/10) | Frase.io (9/10) | WordLift (8/10) |
|---|---|---|---|
| @type declaration | ✓ Organization | ✓ Organization | ⚠ LocalBusiness |
| name + logo | ✓ | ✓ | ✓ |
| disambiguatingDescription | ✓ Present | ✓ Present | ✗ Missing |
| sameAs links | ⚠ 1 link | ✓ 3 links | ⚠ 1 link |
| contactPoint | ✓ | ✓ | ✗ Missing |
| knowsAbout | ✓ 9 topics | ✗ Missing | ✗ Missing |
| makesOffer / pricing | ✓ 3 tiers declared | ✗ Missing | ✗ Missing |
| Insight | Reach 10/10 by adding 2+ sameAs links | Leads via sameAs + description | 50 pages compensates for schema gaps |
Complete Production JSON-LD Template (All 7 Dimensions)
{
"@context": "https://schema.org",
"@type": "SoftwareApplication",
"name": "Innotek AgenticSEO",
"logo": {
"@type": "ImageObject",
"url": "https://innotekseoai.com/logo.png"
},
"disambiguatingDescription": "GEO platform for SMEs that analyses website entity clarity, fact density, and schema completeness for AI model citation optimisation. Operates from Rickmansworth, Hertfordshire, UK.",
"sameAs": [
"https://www.linkedin.com/company/innotek-solutions",
"https://find-and-update.company-information.service.gov.uk/company/XXXXXXXX",
"https://www.wikidata.org/wiki/QXXXXXXX"
],
"contactPoint": {
"@type": "ContactPoint",
"contactType": "customer support",
"email": "hello@innotekseoai.com",
"availableLanguage": "English"
},
"knowsAbout": [
"Generative Engine Optimisation",
"AI citation optimisation",
"Schema.org structured data",
"Entity clarity for SMEs",
"llms.txt generation"
],
"makesOffer": [
{
"@type": "Offer",
"itemOffered": {
"@type": "Service",
"name": "GEO Audit",
"description": "AI readiness audit scoring Entity Clarity, Fact Density, and Schema Completeness"
},
"price": "0",
"priceCurrency": "GBP"
}
]
}
5-Step Entity Clarity Remediation
- Step 1 — Declare @type correctly (+2 pts) — Choose the most specific Schema.org type that describes your organisation. SoftwareApplication, LocalBusiness, LegalService, or MedicalOrganization all carry more signal than the generic Organization. Run the Innotek audit to see your current type and recommended upgrade.
- Step 2 — Write your disambiguatingDescription (+1–2 pts) — Draft a 1-2 sentence description including: what you do (verb + object), who you serve (audience), and where you operate (geography). Avoid adjectives like 'leading' or 'innovative'. Be specific and verifiable.
- Step 3 — Add sameAs links (+1 pt) — Add your LinkedIn company page URL, Companies House URL, and any Wikidata or Crunchbase entries. Each authoritative sameAs link adds approximately 0.3 points. Aim for at least 3 links.
- Step 4 — Populate knowsAbout (+1 pt) — List 5-10 specific topics your organisation genuinely covers in depth. Match these to your actual content clusters. Generic terms like 'marketing' score lower than specific terms like 'GEO citation rate optimisation for SaaS companies'.
- Step 5 — Structure your makesOffer (+1 pt) — List each product or service as a separate Offer object with name, description, and price if applicable. Free tiers count — include them with price: 0. This is the primary signal for commercial query routing by AI models.