Legacy monoliths are often blamed for slowing teams down, and the temptation to rewrite everything from scratch is strong. Yet big-bang rewrites are among the riskiest projects in software, frequently running over budget and failing to ship. There is a better way.
Why Big-Bang Rewrites Fail
A complete rewrite freezes feature development, underestimates hidden complexity, and delays value for months or years. Meanwhile the business keeps moving, and the target keeps shifting. The risk compounds with every month of delay.
The Strangler Approach
Named after a vine that gradually envelops a tree, the strangler pattern replaces a monolith incrementally. New functionality is built as services around the old system, which slowly shrinks until it can be retired. This approach delivers:
- Continuous value throughout the migration
- Reduced risk through small, reversible steps
- Early validation of the new architecture
- The ability to pause or adjust course anytime
Start With the Edges
We begin by extracting well-defined capabilities at the boundaries, where coupling is lowest. Each extraction teaches us more about the system and builds momentum for the harder core.
Keep the Business Running
Throughout the migration, the product keeps evolving and customers keep being served. That continuity is the whole point. Modernization should strengthen the business, not put it on hold.
