Why does the GEO Score not include web search?

Because mixing training knowledge with real-time search would make the score inconsistent — the same brand could get different values depending on whether the model searched at that moment. Separating the two layers keeps the score stable and comparable over time.

Can the score vary between runs of the same prompt?

Yes. LLMs are not deterministic, so the same prompt can produce slightly different answers across runs. The GEO Score should be read as a trend across multiple analyses.

Does a low GEO Score mean bad SEO?

Not necessarily. GEO Score measures the models' organic knowledge, which depends on training data — not the site's current technical SEO. A brand can have excellent SEO and still a low GEO Score if its digital presence is recent.

How the GEO Score is calculated - full transparency