9 May 2026 · 4 min read
The rip-and-replace trap
ERP replacement projects are long, expensive, and disruptive to exactly the processes that keep revenue moving. Most of the frustration comes from a handful of workflows, not the core.
Identify the specific workflows that hurt. Building those as custom modules alongside the ERP is usually an order of magnitude cheaper.
Interface design matters more than technology
An integration layer with a defined contract, idempotent writes, and reconciliation is what makes this safe. Direct database writes into an ERP are how integrations become unmaintainable.
Treat the ERP as a system of record with an API boundary, even if you have to build that boundary yourself.
Roll out module by module
Big-bang cutovers concentrate all the risk on one weekend. Phased rollouts let you learn cheaply.
Run old and new in parallel for one cycle and reconcile. The discrepancies you find are the ones that would have been an incident.
Where to go from here
If you are weighing ERP integration for an upcoming project, the useful next step is usually a short scoping conversation rather than a longer article. Bring your constraints — budget, timeline, existing systems, and the team you already have — and the decision usually resolves itself in an hour.
We run these as a fixed-scope discovery session with a written recommendation at the end, including the cases where our advice is that you do not need us.
Working on something like this?
Send us the brief. We will tell you what we would build, what we would cut, and roughly what it costs.
Talk to us