Ban rate

Why "ban rate" is usually marketing

A percentage is useful only with its definition and population. Check whether temporary restrictions are counted, whether accounts that left remain represented, and which dates are covered.

Define the numerator

For an affected-account share, divide distinct restricted accounts by distinct observed accounts in the same window. If repeated incidents are counted instead, label the result as an incident rate per account.

• State which restriction types count and how repeat events on the same seat are handled.
• Disclose the observed population and the handling of accounts that disconnected, churned, or were restricted.
• State the observation window and data freshness.

Why 90 days

A defined window makes observations comparable. The window alone cannot fix missing events, a small population, or inconsistent definitions.

Why the methodology is published before the number

LinkedBoost's restriction-rate percentage is still measuring. A description of a proposed method is not a measured result, and an unpublished number must not be read as zero.

Controls and measurement

Account checks, pacing, review, and explicit exclusions are operational controls. Their presence does not quantify their effect on restrictions or establish a universal collision guarantee. Read the current seat limits and the separately reported measurement status.

How to read any vendor's claim

Ask three questions. What counts as a restriction? Who is in the denominator? What is the window? If a vendor cannot answer all three in writing, treat the number as marketing.

Is a zero percent ban rate possible?

A sample can observe zero restrictions, but that does not prove the true risk is zero. Report the sample and time window; do not infer account protection from an unmeasured rate.