AI answer metrics

Visibility metrics

• AI Share of Voice. How often the models name your brand compared with competitors across the same scenario pool. Computed per AI system, market and prompt segment.
• Rank Position. Where the brand appears in the lists: whether you make the top-3 options the model offers first.
• AI Recommendation Rate. The share of scenarios where the brand is not merely mentioned but directly recommended for purchase or use.
• Prompt Coverage. Which funnel segments you are present in, and where the answers belong entirely to competitors — a gap map.

Context and sentiment

• Sentiment Score. Mentions split into positive, neutral and negative — you can see whether a label like “expensive” or “hard to set up” has stuck to the brand.
• Association cloud. The words models most often put next to your brand, and their dynamics from wave to wave.
• Factual Accuracy. An error register: outdated prices, discontinued models, wrong distribution claims, pre-rebrand names — each with the exact quote and measurement date.

Sources and content

• Source Share. The sites, media and review platforms the models cite when answering about your category — a ready-made priority list for PR and link building.
• Semantic Match. How close your site’s content is to what the models cite. A low score explains why AI takes facts about you from other people’s pages.
• Volatility. How much the answer changes across repeated runs — separating a stable brand position from random mentions.
• Model / Market Variance. Differences between AI systems and markets: a brand can lead in ChatGPT for the US and be absent from answers about Russia.

AI answer metrics are not market share and do not predict sales. Methodology and repeated runs are described on the How it works page.