Test Automation: Strategy, Architecture, and Management
Test Automation: Strategy, Architecture, and Management
This domain tests how to plan, design, and sustain automation. Expect questions on TAS, TAA, TAM, frameworks, ROI, and maintenance-not just tool features.
Core rules
- Test Automation Strategy (TAS) defines goals, scope, risks, and what to automate. Good candidates are stable, repeatable, high-volume, data-intensive, or non-functional tests; poor candidates are volatile, one-off, or exploratory tests.
- Test Automation Architecture (TAA) defines layers: test generation, test definition, execution, adaptation, and reporting. Frameworks include linear, modular, data-driven, keyword-driven, hybrid, and page object model. Separate test logic from data and from the system interface.
- Test Automation Management (TAM) covers process, roles, tools, metrics, version control, CI integration, and maintenance.
- ROI compares costs (tools, development, maintenance, training, infrastructure) with benefits (speed, repeatability, coverage, earlier feedback). Maintenance is usually the largest long-term cost.
- Automation must be risk-aligned: automate high-risk regression checks first.
Worked example
A team has 500 regression tests run weekly at 10 minutes each (about 83 hours). Automation costs 200 hours to build, runs in 1 hour, and needs 10% maintenance per year. Weekly saving is about 82 hours; payback is roughly 2.4 weeks before maintenance. Reasoning: include maintenance and execution time, not only build cost. If the tests change every sprint, maintenance can erase the ROI.
Traps
- Automating unstable features and then blaming the tool.
- Counting only build cost and ignoring maintenance.
- Using brittle locators or hard-coded data.
- Automating everything instead of applying risk and ROI.
- Treating automation as a project rather than a lifecycle.