AGS (AI Grading System) is an open method for scoring brand visibility in generative AI answers. For each answer, a judge from a pool of independent models scores presence, influence and quality, aggregated into a single 0-100 score. A sample is double-scored to measure inter-judge reliability (GRC).
What is the AGS?
The AGS (AI Grading System) is the scoring system that measures a brand's visibility in the answers of ChatGPT, Claude, Gemini, Perplexity and other AI tools. Rather than an opaque grade, it is an open method: it can be understood, verified and challenged. The full method is published on our AGS Methodology page.
The three dimensions
Every AI answer is scored across three composite dimensions, combined using a geometric mean:
- P — Presence: is the brand cited, in what position, and to what extent? (see mention rate)
- I — Influence: is what is said informative, unique and relevant?
- Q — Quality: is the answer accurate and pitched at the right sentiment?
Why a panel of several AI models?
The AGS relies on an AI panel drawn from different providers to stop any single model from favouring itself. The score comes with a confidence interval and an inter-rater reliability coefficient, so that its uncertainty is shown rather than hidden. The acknowledged limits of the measurement are documented too.
Going further
Agencies, consultants and in-house teams: compare the answers generated for your own scenarios, languages and selected models. Our infrastructure, your scenarios, your expertise.