Definition
Error Rate measures the proportion of user requests or operations that fail, encompassing both server errors (5xx) and client errors (4xx) in web applications. It is the most direct measure of reliability from the user's perspective. High error rates erode trust, increase support costs, and directly impact conversion and retention. Error rate should be tracked per service, per endpoint, and per user segment to quickly identify and isolate issues.
Formula
How to measure
Calculate (5xx responses + significant 4xx responses) ÷ total requests. Exclude expected 4xx responses (e.g., 404 for missing pages that are handled gracefully). Track per service or endpoint. Set real-time alerts for error rate spikes. Correlate with deployment events to catch regressions.
Industry benchmarks
Production targets: below 0.1% for critical paths, below 1% for non-critical endpoints. During incidents, error rates above 5% typically warrant immediate response. Payment/checkout flows should maintain error rates below 0.05%.