I85PREP
⌘ K
QAI CAST study materials
QAI CAST study resources

QAI CAST: Testing Tools and Automation Basics

QAI CAST · Work through your exam, one topic at a time.

All guide sections

Testing Tools and Automation Basics

Testing Tools and Automation Basics covers the tool landscape and the economics of automation. CAST expects you to categorise tools by purpose, know what record/playback can and cannot do, understand the main automation frameworks, and reason about when automation pays back rather than assuming it always does.

Tool categories

  • Test management: cases, runs, traceability, reporting.
  • Defect tracking: workflow, assignment, triage history.
  • Static analysis and review support: coding rules, complexity, duplicate code.
  • Test design and data generation.
  • Execution: record/playback, data-driven, keyword-driven, behaviour-driven.
  • Performance and load, security, coverage, and continuous integration orchestration.

Automation realities

  • Automate stable, repetitive, high-volume, deterministic checks: regression suites, data setup, performance loads.
  • Do not automate one-off exploratory checks or volatile interfaces.
  • Record/playback is quick to create and expensive to maintain; it breaks on dynamic identifiers, timing waits, and layout change.
  • Data-driven frameworks separate test data from scripts; keyword-driven frameworks separate action words from implementation; hybrid frameworks combine both.
  • Stubs and drivers simulate missing components during integration testing.

Return on investment

Break-even runs = automation cost / (manual cost per run − automated cost per run, including maintenance). Skipping maintenance in that formula is the most common analytical error.

Worked example

A regression suite takes 8 hours to run manually. Automation costs 120 hours to build, an automated run takes 1 hour, and maintenance averages 2 hours per run. How many runs to break even?

Reasoning: saving per automated run = 8 − (1 + 2) = 5 hours. Break-even = 120 / 5 = 24 runs. If the team runs the suite monthly, payback takes two years, which may fail the business case. If continuous integration runs it nightly, payback arrives in under a month. Frequency and maintenance cost, not scripting speed, usually decide the outcome.

Traps that cost marks

  • Believing automation replaces test design or exploratory judgement.
  • Omitting maintenance cost from the ROI calculation.
  • Using record/playback against dynamic data or unstable interfaces.
  • Treating a tool purchase as a process improvement in itself.
  • Forgetting that a tool's output still needs human interpretation.