The Michai Index.
A privacy-safe benchmark for what public business websites make visible to search engines, AI answer systems, and ready-to-act buyers.
No fake benchmark.
The pilot cohort has not reached the minimum sample, so Michai is not publishing a misleading average or pretending a handful of scans represent a market.
Scans count only when the operator explicitly opts into anonymous aggregates. The scanner has processed 2 stored reports in total.
Run a free Signal scan →Visible rules. Conservative claims.
Inputs
Public homepage HTML and response headers, plus standard robots.txt, sitemap.xml, and llms.txt locations. No credentials or private analytics.
Weights
Search 30%, AI answer visibility 25%, demand conversion 25%, and technical delivery 20%. Each check exposes its available points.
Privacy
Reports are unlisted. Aggregate consent is opt-in. Emails, companies, hostnames, and report-level findings never appear in the Index.
Thresholds
Overall results require n≥10. Industry and market cuts require n≥3. Thin cohorts remain unpublished rather than overinterpreted.
Limitations
Observed response time is one server-side sample. The Index does not measure rankings, backlinks, demand, field Core Web Vitals, CRM outcomes, or revenue.
Versioning
Method v1.0 is frozen for the pilot. Material scoring changes will create a new version rather than silently rewriting historical comparisons.
Scan the public surface.
The report is useful immediately. Anonymous benchmark participation is always optional.
Run Michai Signal →