Walk most operations floors and you will find two systems running side by side. There is the ERP, which holds the data and cost a fortune. And there is the real one: the whiteboard, the paper order slips, the weights phoned in and written down, the orders printed on A4 and sorted into trays. The ERP records the end result. The actual work happens in the margins around it.
Replacing the ERP is the wrong instinct
The instinct is to fix this by replacing the ERP. It is almost always the wrong instinct: expensive, slow, risky, and beside the point. The system of record usually isn’t broken. The problem is everything it cannot see. So we don’t replace it. We build a layer on top.
A layer on top, three ways
For a make-to-order manufacturer, that meant connected views sitting over the existing ERP, replacing the printed orders and the trays, capturing real assembly times as a by-product, and never once touching the system underneath. For a recycling operator running 600 transport orders and 350 invoices a month on a whiteboard and paper slips, it meant a platform where the order itself is the system: customer to order to driver to weight to invoice, in one flow. For a distributor drowning in 30-50 purchase orders per container, it meant goods receiving that collapsed to one receipt per truck, 80-90% less receiving time.
None of those projects ripped out the ERP. They stopped the business running in the gaps around it.
You rarely need a new system of record. You need to stop running the business in the margins of the old one.
Lower risk, shipped in weeks
There is a quieter benefit too. When the new layer leaves the system of record untouched, it is reversible, it is lower-risk, and it ships in weeks instead of years. You find out whether it works before you have bet the company on it.
The system of record usually isn’t the problem. Everything it cannot see is. Build for that, and leave the ERP where it is.