Methodology

How accurately each AI model describes your brand compared to your competitors.

Summarizes relative visibility when models draft recommendation lists for your category search queries.

Each search term you choose is expanded into roughly ten synthetic sub-queries across seven angles — comparison, price and value, use-case fit, alternatives, how-to, trust and reviews, and attribute-specific. Every sub-query goes to three models with web browsing switched off, so this measures what those models already hold, not what a live search returns today. Scoring is presence-based: for each answer we record whether your brand and each competitor were named at all, taken from the model's own report and cross-checked against the answer text. We report the share of answers that named each brand alongside the share of cells that returned; a term whose cells failed or timed out is reported as not measured. Sub-answers are prose rather than ranked lists, so there is no position to score, and these percentages are not comparable with the Search Term Visibility scores.

How well your website is set up for AI to find, read, and reuse.

Which topics your competitors cover that your site doesn't — and how completely you cover the ones that matter.

GEO Brand and GEO Web establish your baseline across the signals that drive AI visibility. These KPIs define what to measure as you implement recommendations over time.

After your audit, these are the numbers worth watching: how often AI mentions and recommends your brand, whether the three major models agree about who you are, and how often your pages get used as the direct answer to a question. Each is measurable, and each one improves as you implement the report's recommendations.

Full definitions for each KPI are available in Key Terms.

Longitudinal tracking and benchmarking tools are on the Yes AI roadmap.