Why "Share of Model" is the New "Share of Search"
In traditional SEO, you tracked rankings (e.g., "Position 1 for 'best running shoes'"). In the AI era, rankings don't exist. You are either the cited answer or you are invisible. This metric is called Share of Model (SoM).
How AI "Sees" Your Brand
When a user asks Perplexity or Gemini about your brand, the model performs a "Retrieval Augmented Generation" (RAG) process:
- Retrieval: It searches its training data and live web indexes (Bing/Google) for mentions of your entity.
- Synthesis: It combines these mentions to form an answer.
- Citation: It credits the most authoritative source.
If your "Entity Confidence" score is low (meaning the AI finds conflicting data about your pricing, location, or services), it will often choose not to mention you at all to avoid hallucinating.
The "Invisible Brand" Problem
We frequently see brands with great Google rankings but zero AI visibility. Why?
- Blocking Bots: They inadvertently block GPTBot or ClaudeBot in their robots.txt.
- Unstructured Data: Their pricing is hidden in PDFs or images, which AI struggles to parse.
- Lack of Consensus: Their Trustpilot reviews say one thing, but their website says another.
The Fix: EZY.ai 's dashboard monitors your "Sentiment Score" and "Citation Frequency" across major models, alerting you instantly if your brand becomes invisible to the AI.
