9.3Flaky Tests as Organizational Smell

Anti-Pattern: Flaky tests are treated as individual test bugs to be fixed one at a time. The flaky rate stays persistently high because the root causes are systemic, not local.

Pattern: Treat persistent flakiness as a signal about the organization, not just the test suite.

What Flaky Tests Signal

Architecture signal — Tight coupling between services, shared mutable state, missing API contracts. When services depend on each other's internal behavior rather than defined interfaces, tests that cross service boundaries become flaky. The fix is not better tests — it is better contracts.

Process signal — No ownership of flaky tests, no time allocated for test maintenance, no accountability for test infrastructure health. When flakiness is "nobody's job," it persists indefinitely. The fix is assigning ownership and allocating sprint capacity for test health.

Leadership signal — Quality infrastructure is underfunded, test environments are unreliable, CI resources are insufficient. When leadership treats testing infrastructure as a cost to minimize rather than a capability to invest in, flakiness is the predictable result. The fix is making the business case for infrastructure investment.