Main idea

Fixing a defect straight away has a visible price: a feature you don't finish, a sprint goal disrupted, a roadmap subject pushed back. It is unpleasant, and it is normal — quality consumes capacity like everything else.

A team that sets defects aside to preserve its velocity gets exactly what it measures. The number holds. Meanwhile usage friction, support load, debt and distrust accumulate — none of those four lines enters velocity.

Why it matters

Velocity measures a delivery throughput, not a state of the product. Making it a goal creates the precise incentive not to fix: the only way to protect the number is to take quality out of the measured scope.

Nuances and limits

The opposite excess exists. A team that switches all its capacity to fixing protects the product and ships nothing, which also has a cost.

Open questions

• Which measure puts the cost of defects in the same unit as velocity, so the trade-off becomes legible?