AI language models don't just read the web — they're trained on structured knowledge graphs. When ChatGPT or Gemini mentions a company in an answer, that mention is grounded in an entity: a distinct, recognizable organization with a known name, domain, and set of associations. If your brand isn't a recognized entity in those knowledge graphs, AI models treat you as an anonymous website, not a named source worth attributing.
GEO Auditor's brand entity engine measures how well your brand is recognized across the platforms AI models use as reference sources.
Why entity recognition matters for AI citations
There are two ways an AI engine can cite your content:
- Anonymous citation: "According to one source…" or just quoting the text without attribution
- Named citation: "According to [Your Brand]…" with a link and direct attribution
Named citations are far more valuable for brand building and for driving referral traffic. They require the AI model to have enough confidence in your entity identity to attribute the content by name. That confidence comes from knowledge graph presence.
What the brands engine checks
Platform presence scores
The engine checks your brand's presence and profile completeness across the platforms AI models most commonly use as entity reference sources:
- Wikidata — the most important knowledge graph for AI entity resolution. A Wikidata entry with your organization type, founding date, website, and sameAs links is the highest-confidence signal of entity legitimacy.
- LinkedIn — company page completeness: description, industry, website, founding year, employee count range. LinkedIn is heavily weighted in AI knowledge graphs for business entities.
- Wikipedia — a Wikipedia article about your organization is the strongest possible entity signal, but is only appropriate for companies with significant notability. Most startups won't qualify, but the check is included because it's so impactful when present.
- Reddit — brand mentions and community discussions on Reddit are used by AI models as social proof and topic association signals.
- YouTube — a brand channel or significant number of brand-related videos signals media presence and content authority.
Social presence completeness
Beyond the primary knowledge graph sources, the engine checks your presence on Twitter/X, GitHub, and other platforms relevant to your industry. These contribute to the overall "entity signal density" — the more places your brand appears consistently with the same name, URL, and description, the more confidently AI models can resolve your entity.
Consistency scoring
Entity confidence depends on consistent information across sources. If your brand name is "Acme Corp" on your website, "ACME Corporation" on LinkedIn, and "Acme" on Wikidata, AI models have lower confidence that these all refer to the same entity. The brands engine flags inconsistencies in name, URL, and description across platforms.
How to build brand entity recognition
1. Create a Wikidata entry (highest priority)
Wikidata is free to edit and edit decisions for new organizations are generally approved quickly. A minimal Wikidata entry for a business should include:
- Instance of: business / company / organization (as appropriate)
- Official website URL
- Inception date (founding year)
- Country of headquarters
- Industry classification
Once the entry exists, add its URL to your website's Organization schema as a sameAs property. This creates a bidirectional link between your website and your Wikidata entity.
2. Complete your LinkedIn company page
A complete LinkedIn page means: company description (at least 3 sentences), industry selected, website URL filled in, company size range set, founding year, and a logo. Incomplete LinkedIn pages score significantly lower in entity confidence checks.
3. Maintain NAP consistency
NAP — Name, Address, Phone — consistency across web presences is a traditional local SEO concept that applies to AI entity recognition too. Use exactly the same company name and website URL everywhere. Pick a canonical form and stick to it.
4. Add sameAs to your Organization schema
The fastest way to link your website to your entity profiles is via schema markup. Add a sameAs array to your Organization JSON-LD with links to every authoritative profile:
"sameAs": [
"https://www.wikidata.org/wiki/Q12345",
"https://www.linkedin.com/company/yourbrand",
"https://twitter.com/yourbrand",
"https://github.com/yourbrand"
]Brand entity score in GEO Auditor
The brands engine produces a composite score weighted toward Wikidata and LinkedIn presence, with secondary weight on Wikipedia, Reddit, and YouTube. The paid report shows per-platform scores, specific missing fields on each profile, and a step-by-step setup guide for creating a Wikidata entry if one doesn't exist.
Run a free scan to see your brand entity recognition score and which platforms are suppressing it.