EI-301i · Module 3

Preventing Alert Fatigue

3 min read

Alert fatigue occurs when recipients receive so many alerts that they stop paying attention to any of them — including the critical ones. Alert fatigue is the most common failure mode for alerting systems, and it is almost entirely preventable through design discipline. The symptoms: recipients stop reading alerts (observable through declining acknowledgment rates), recipients create email filters to hide alerts (observable through delivery vs. open rate gaps), and recipients fail to act on P1 alerts that they received (the most dangerous symptom).

Do This

  • Set a per-recipient alert budget: no individual should receive more than 5-8 alerts per week — exceed this and fatigue is guaranteed
  • Monitor alert acknowledgment rates — a declining rate is an early indicator of fatigue
  • Regularly audit alert actionability: what percentage of alerts led to an action or decision? If below 40%, reduce volume
  • Provide a "snooze" or "not relevant" feedback mechanism — recipient feedback improves trigger calibration and shows recipients their input matters

Avoid This

  • Add new alert types without retiring old ones — alert volume should stay within the budget, not grow indefinitely
  • Send identical alerts to multiple channels (email AND Slack AND SMS) for non-P1 items — multi-channel delivery multiplies perceived volume
  • Ignore complaints about alert volume — complaints are the last signal before complete disengagement

The alert budget is the single most important design constraint. If a recipient should receive a maximum of 5 alerts per week, the triggering system must be calibrated to produce that volume for that recipient. When a new alert type is added, something else must be tightened or removed to stay within budget. The budget constraint forces continuous trigger optimization and prevents the gradual volume creep that causes fatigue.