SearchEngineAnswer browser tool

AI Visibility and Referral Lag Auditor

Check whether a page-level visibility series is complete enough to compare with referral sessions, then calculate descriptive Spearman rank correlation across a declared lag grid.

Browser-local: this page makes no network request and does not use cookies, local storage, or session storage. Your pasted CSV stays in this tab. Remove confidential prompts, personal data, raw IP addresses, tokens, and private URLs before use.

1. Paste the paired observation grid

Required headers: observation_date, platform, canonical_url, visibility_score, referral_sessions, and coverage_status. The full 36-field SearchEngineAnswer ledger is accepted; extra columns are preserved only in the source textarea.

CSV observations

Maximum lag (days)

Minimum complete pairs

Minimum usable coverage (%)

The default example is intentionally too small for the 20-pair decision gate. It demonstrates the calculations without pretending to be a publishable result.

How to interpret the output

A row marked hold fails the declared pair, coverage, or variation gate.
A coefficient describes rank alignment for that page, platform and lag. It does not identify the cause of either series.
Report every lag you tested. Selecting the largest value after inspection inflates the chance of a misleading pattern.
Analyze page-level groups before pooling. A sitewide result can reflect page mix rather than within-page alignment.

Method boundary: descriptive diagnostics only. No p-value, confidence interval, traffic classification, attribution model, forecast, or causal claim is produced.