What we exclude from Wi-Fi “health” dashboards

Every vendor ships a rainbow health score. We remove metrics that reward high channel utilization driven by legitimate broadcast traffic, because leaders misread them as congestion crises.

We also exclude raw client counts on guest SSIDs from the same view as operations SSIDs. Mixing those curves trains people to optimize the wrong problem.

Latency samples without confidence intervals get dropped. A single bad minute should not dominate a weekly digest aimed at regional directors.

Exclusions are documented beside the dashboard, not in footnotes. If a metric returns later, it returns with an owner and a refresh cadence so ghost tiles do not resurrect quietly.