Alert Noise Ratio Calculator

Work out alert noise ratio instantly with clear inputs, formula shown and shareable results.

About Alert Noise Ratio Calculator

The actionable rate is the metric that determines whether alerts are trusted. Google's SRE guidance is that every page should be actionable, novel and urgent; when the actionable rate falls below roughly half, responders start closing alerts without investigating, and the monitoring system stops functioning as a detection mechanism regardless of its coverage.

Formula

Alert quality

noise ratio = (total - actionable) / total; wasted time = noise count x minutes per alert

Worked example

With the default values below, Alert Noise Ratio Calculator returns Alert noise ratio of 85.0%. Change any input above and every figure here is recalculated in your browser.

Inputs used

Alerts firedminimum 1: 1,400
Alerts that required actionminimum 0: 210
Alerts that paged out of hoursminimum 0: 340
Time spent per alertminimum 0.1 min: 6 min

Results

Alert noise ratio: 85.0%
Actionable alert rate: 15.0%
Hours spent on non-actionable alerts: 119.0 h
Estimated out-of-hours pages that were noise: 289

Step by step

1Values used
2Alert quality
3Alert noise ratio= 85.0
4Actionable alert rate= 15.0
5Hours spent on non-actionable alerts= 119.0 h
6Estimated out-of-hours pages that were noise= 289

Assumptions and limits

The result is derived only from the 4 inputs listed above. Anything not among them is not part of the calculation.
In particular, this calculator does not take inflation, tax, fees or charges or compounding frequency into account.
Figures are rounded for display to 1 decimal place, so totals built from several rounded parts can differ slightly from the unrounded arithmetic.

How these calculators are built and tested is described on the Methodology page. If a figure looks wrong, please report it.

Frequently Asked Questions

What actionable rate should be targeted?

Above 80 percent for anything that pages. Alerts that are merely informative belong on a dashboard or in a ticket queue, not on a phone at 3 am.

How is noise reduced quickly?

Raise thresholds to symptom level rather than cause level, alert on burn rate against an SLO instead of on raw metrics, and group related alerts into a single incident.