Definition
The likelihood that a software release will cause a production incident, degraded user experience, or data integrity problem. Release risk is a composite score derived from multiple signals: engineering activity (schema changes, commit frequency, CI events), cross-signals (table diff results, expectation failures, SEO checks), and historical patterns (recurrence risk from the Memory Graph).
Why it matters
High release risk means the team is flying blind into production. Low release risk means the evidence says this release is safe — but only if the right checks were run. A release risk score without evidence is noise; a score backed by schema diffs, CI signals, and recurrence history is a real decision tool.
How Well Tested handles it
Well Tested computes a release risk score (0–100) that blends engineering event severity, cross-signal concern (table diff failures, expectation failures, SEO regressions), and recurrence risk from the Memory Graph. The score comes with specific reasons — 'schema change in billing.invoice_items detected, 2 incidents correlated with prior changes in last 90 days.'
No commitment. 20 minutes. We review your release workflow and show you what's possible.