Test Planning, Estimation, and Monitoring
Domain focus
Test planning, estimation, and monitoring test whether you can define scope, choose an approach, estimate effort, set entry and exit criteria, and use metrics to control testing.
Planning essentials
A test plan commonly covers scope, objectives, test items, approach, levels and types, entry/exit criteria, schedule, resources, roles, risks, deliverables, and environment needs; common structures follow IEEE 829 or ISO/IEC/IEEE 29119. Entry criteria define what must be ready before testing starts; exit criteria define when testing can stop. Good exit criteria are measurable: planned coverage achieved, no open critical defects, and agreed residual risk.
Estimation techniques
- Ratio-based estimation uses historical ratios, such as test effort per requirement or per developer-day.
- Extrapolation scales from early data, such as defects found per test hour.
- Expert judgment and Delphi combine estimates anonymously to reduce bias.
- Three-point estimation uses optimistic (O), most likely (M), and pessimistic (P): expected = (O + 4M + P) / 6.
Monitoring and control
Track planned versus executed tests, pass/fail rates, defect discovery and closure, coverage, and environment availability. Compare against the plan, then take control actions: re-prioritize, add resources, reduce scope, or escalate. Metrics must inform decisions, not decorate reports.
Worked example
A team estimates test execution with O = 10 days, M = 14 days, P = 24 days. Three-point estimate = (10 + 56 + 24) / 6 = 15 days. The plan adds contingency for an unstable test environment and sets exit criteria: 100% of high-risk tests executed, no open severity-1 defects, and 95% requirements coverage. During monitoring, execution is 40% after 50% of the schedule; the team re-prioritizes high-risk tests and escalates environment delays rather than simply reporting a red status.
Traps
Estimating without a defined scope or assumptions. Confusing entry with exit criteria. Treating “all tests passed” as sufficient when coverage is low. Reporting metrics without control actions. Ignoring environment lead time in the schedule.