When a feature lands with no artifacts, I start by decomposing scope into testable slices — API, UI, integrations, performance, security — and score each by complexity and risk to produce a time-boxed estimate. For a typical medium feature I’d estimate 40–80 test hours across 2–3 testers in a two-week sprint, with an initial smoke suite in 8 hours. My strategy doc highlights objectives, risk-based priorities, environments (dev, staging with prod-like data, performance lab), and automation targets: aim for 90% regression automation for critical flows and 60–70% for new scenarios. Deliverables: test plan, prioritized test cases with traceability to requirements, test data matrix, CI pipeline jobs, and a checklist of exit criteria — 95% pass rate on critical tests, zero open Severity 1, and performance within 5% of baseline. I iterate automation after release to drive down manual effort.
Get AI-powered feedback on your answer and improve your skills
Takes 5-10 minutes