Step 1 — define the queries that matter

Pick the 5–10 queries where being cited matters most: category ("best X"), comparison ("X vs Y"), and use-case queries. These form your measurement frame. If a "best X" answer in an AI engine could send you a buyer, it belongs on the list.

Step 2 — pick the engines to test

Choose ChatGPT, Perplexity, Google AI Overviews, or all. Don't test everything at once; start with where your buyers are. GeoVisor lets you select a single engine or "All" per audit.

Step 3 — run the audit and capture the score

A tool like GeoVisor returns a 0–100 GEO score plus the top visibility issues for a target query and your URL. Manual method: ask each engine the query and record whether your brand appears, in what form (named, linked, paraphrased), and whether a competitor is cited instead.

Step 4 — log competitor gaps

Note where competitors are cited and you aren't. That gap is your priority list — the specific queries and formats (comparison table, FAQ) where you're absent.

Step 5 — act, then re-measure

Publish the missing asset (a comparison or FAQ page), then re-audit to see if the gap moved. Treat measurement as a loop, not a one-off project.

What a "good" score means

A GEO score is most useful as a relative, time-series measure for your own queries. A rising score suggests your visibility work is moving the gap; treat absolute values as directional, not definitive. No score guarantees a citation.