🧪 Software Testing Pyramid

About this simulation

The "test pyramid" is a heuristic for how a healthy automated test suite should be shaped: many fast, cheap unit tests at the base, a smaller layer of integration tests checking how components work together, and a thin cap of slow, expensive end-to-end (E2E) tests that exercise the whole system like a real user would. Each layer here is built from instanced blocks — one per simulated test case — stacked into a pyramid whose width encodes how many tests live at that level.

How it works

• Glowing pulses spawn at the base and climb layer by layer, each one a simulated CI test run.
• A pulse turns red and stops if it "fails" — unit-layer failures resolve almost instantly; E2E failures take far longer to surface because those tests are slower and run later in the pipeline.
• The ice-cream-cone toggle inverts the shape into the classic anti-pattern: few unit tests, heavy reliance on slow, brittle E2E tests — a common real-world failure mode.
• Average feedback time rises sharply as more of the suite shifts toward slower upper layers.

Controls

• Unit share / Integration share — set what fraction of the suite lives at each layer (E2E gets the remainder).
• Run speed — how fast simulated CI runs fire.
• Flakiness / bug rate — probability any given test run fails.
• Ice-cream-cone toggle — flips to the inverted, top-heavy anti-pattern shape.
• ↺ Rebuild pyramid — regenerates the stack from current settings. Drag to rotate, scroll to zoom.

Did you know?

Mike Cohn's original 2009 "test pyramid" argued that because UI/E2E tests are slow, brittle and expensive to maintain, teams should push most of their coverage down into fast unit tests — a principle still central to modern CI/CD and shift-left testing practice.