You can tell something is wrong with the CRM. You cannot easily say what, how much, or which thing to fix first.
Reads your portal through the API and scores it across seven operational domains: portal configuration, users, CRM structure, data quality, workflows, marketing and reporting. Each finding carries the count behind it, the population it was measured against, and what to do about it.
A full scan once a week, including the duplicate detection pass, which is the expensive one. A lighter scan every day for the figures that move. Every run is stored as a snapshot, so a number can be compared with the same number a week ago rather than recalculated from scratch and hoped about.
The hygiene half of the audit, and the half most people are looking for. Duplicate contacts and companies, both exact matches and near matches. Records missing fields that downstream processes depend on. Contacts with no logged activity. Ownership recorded in more than one field, where two reports can disagree about who owns a record. Configuration that has drifted from what the rest of the portal assumes.
How duplicates are counted. Two passes. An exact pass on normalised values, and a near-match pass that compares names and requires the email domains to corroborate the match. The near-match pass is approximate by design, and it is deterministic: the same records produce the same pairs on every run. When the comparison budget is reached, it says the scan was truncated rather than quietly reporting a smaller number.
Nothing. The audit is read-only. It needs read access to your portal and writes no field, no property and no record. Findings reach you through the RevOps Health digest and the audit itself.
Where a figure cannot be measured this period, the audit says why instead of showing a zero. A duplicate count that is older than the rest of the report is labelled as such. A metric whose calculation changed is shown as a fresh baseline rather than compared against a number produced a different way.