Define an edition
Freeze the question sample, provider surfaces, collection window and metric definitions.
Research
AI visibility research with published methods and review gates.
MonitorMyGEO's research program uses eligible, derived observations to study AI visibility, citation and recommendation patterns. A result is published only after dataset, methodology, privacy and editorial review. This page describes the framework; it does not present a market statistic without a verified edition.
The problem
Why this matters
AI visibility numbers are easy to misread when the prompt sample, provider surface, collection period or failed observations are omitted.
Workflow
How to use it
Check eligible data
Use only observations allowed by workspace data policy and apply disclosure thresholds to every cohort.
Review before publication
Retain the evidence snapshot and require data, privacy, methodology and editorial approvals.
Observable signals
What you can inspect
- Prompt and provider sample size
- Observed mention, recommendation and citation rates where publishable
- Provider variation and known comparability limits
Limits to keep in mind
No numerical result is published here before a reviewed edition passes the existing disclosure and evidence controls.
Questions before you start
Is the AI Visibility Index live?
Its methodology is available. Numerical editions require verified data and review before publication.
Are customer responses public?
No. The research surface uses eligible derived observations and excludes raw answers and customer identity.
Related resources
Next step
Inspect evidence for your own brand.
Choose relevant prompts and review the underlying provider answers before drawing conclusions.