Exploratory Testing

Overview

Exploratory testing is simultaneous learning, test design, and test execution. It discovers edge-case bugs, usability flaws, and unexpected state combinations that predefined automated tests miss.

Session-Based Test Management (SBTM)

Conduct exploratory testing in structured, timeboxed sessions (45-90 minutes):

1. Charter: Clear scope defining what to explore, with what tools, to discover what risks.
   • Formula: Explore [target/feature] with [tools/data/methods] to discover [risks/vulnerabilities/bugs].
2. Timebox: Focus strictly on the charter without distractions.
3. Session Notes: Log bugs found, areas covered, questions raised, and test ideas.
4. Debrief: Review findings with developers or product managers.

Proven Exploratory Heuristics

SFDIPOT (San Francisco Depot)

• Structure: What is the application made of? (Files, modules, dependencies)
• Function: What does the software do? (Features, calculations, transformations)
• Data: What does it process? (Inputs, outputs, boundary values, corrupted data)
• Interfaces: How does it interact? (UI, APIs, CLI, webhooks, exports)
• Platform: What does it rely on? (OS, browser, hardware, network speed)
• Operations: How is it used? (User personas, unusual habits, peak usage)
• Time: When does timing matter? (Timeouts, race conditions, concurrency, timezones)

Testing Tours

• The Guidebook Tour: Follow user documentation step-by-step; verify reality matches docs.
• The Saboteur Tour: Deliberately disrupt inputs, kill network connections mid-request, supply malformed payloads.
• The Supermodel Tour: Focus entirely on visual layout, aesthetics, typography, alignment, and responsiveness.
• The Intellectual Tour: Stress the system with complex, highly intricate workflows and heavy calculations.