Big refactors need strangler patterns, not big-bang rewrites.
[ essay ]
The refactor that eats a quarter rarely finishes in a quarter. It expands because every opened layer reveals another assumption.
Prefer strangler fig patterns: route new work through the new path, migrate callers incrementally, delete old code when traffic hits zero. Keep main deployable weekly — daily if you can.
Define done as measurable: 80% of checkout flows use PaymentServiceV2, not rewrite payments.
Time-box spikes. If exploration exceeds the box, write findings and choose: continue with funding, or accept current state with documented pain.
Refactoring is not a personality trait. It is a project with scope and a stop condition.
