Classic SEO measures where a page ranks. AI search has no ranked list, so it needs different measurements: whether the brand is named, how it compares, whether its pages are trusted as sources, and who owns the shopping shelf.
Overall, how visible is the brand inside AI answers?
GEOly's composite 0–100 score for how visible a brand is inside AI answers. It combines whether the brand is named, where it appears in the answer, whether its own domain is cited, and the sentiment of the surrounding language — across every tracked prompt, engine and market.
Formula: A weighted composite of mention rate, answer position, citation rate and sentiment, normalized to 0–100.
Of all the brands AI names in my category, how much of that attention is mine?
The share of all brand mentions in a category's AI answers that belong to your brand. It places visibility in a competitive context instead of counting appearances in isolation.
Formula: SoM = your brand mentions ÷ all brand mentions in the same answer set × 100
How often do AI answers use my own pages as evidence?
The share of tracked AI answers that cite the brand's own domain as a source. Being retrieved is not the same as being cited: an engine can read your page and still choose not to rely on it.
Formula: Citation Rate = answers citing your domain ÷ total tracked answers × 100
Whose products fill the AI shopping shelf in my category?
The share of AI shopping product cards in a category that are occupied by your products. This is the shelf metric of agentic commerce: AI assistants assemble a product shelf before the click, and the card decides which store the buy-now link points at.
Formula: Share of Card = your product cards ÷ all product cards surfaced for the category × 100
Which websites actually feed AI answers in my category?
A source-side authority score for a domain — the AI-era counterpart to domain authority. It ranks how much a given publisher, retailer, forum or reference site actually feeds AI answers in a category, so outreach can be prioritized by influence on answers rather than by traditional backlink value.
Formula: Derived from how often a domain is retrieved and cited across a category's AI answers, and how many brands it funnels into them.
What did the engine search for while writing the answer?
The background searches an AI engine runs while assembling one answer. Reading them shows what the engine was actually looking for — including site-restricted lookups against brand domains to verify claims like warranty, materials or certifications.
Formula: Not a ratio — a captured list of the engine's background queries per tracked prompt.
Start with Share of Model per engine and market, read next to mention rate and Citation Rate. Together they show whether you are present, chosen and trusted. A single composite like AIGVR is useful as a headline, but decisions come from its components.
Rank tracking reports the position of a URL on a results page. AI answers are prose with a few cited sources, so the measurements are presence and position inside the answer, share against other brands, and whether your domain is cited.
You can start by hand: a fixed list of buyer questions, asked in each engine on a schedule, with the answers saved and brands counted. A tool such as GEOly does the same at scale across engines, countries and languages and keeps the raw answers.
Vertrauenswürdig bei führenden Verbrauchermarken
Von Anker SOLIX bis xTool — die oben genannten Marken sehen bereits, wie ChatGPT, Gemini und Google AI sie erwähnen, zitieren und empfehlen. Ihre Marke wird gerade in der KI diskutiert. Sehen Sie es.
Kostenlos starten · Keine Kreditkarte erforderlich