Skip to content
FocalIDFocalID

Seven ERP implementation mistakes and how to avoid them

An ERP system connects finance, stock, purchasing, sales and production in one place. Putting one in is among the largest changes a company makes, and the same mistakes derail it again and again. Most have little to do with the software.

By the FocalID team in Colombo, , 2 min read

The interior of a large factory hall with overhead cranes.

The seven mistakes

  • No clear goal. “We need an ERP” is not a goal. “Close the month in three days” or “see stock across all branches” is. Write the goals down and use them to settle arguments.
  • Treating it as an IT project. An ERP changes how people work. It needs an owner from the business with the authority to make decisions, not only the IT team.
  • Customising everything. Each change to the standard product adds cost now and at every upgrade. Adopt the standard process unless there is a strong reason not to.
  • Leaving data migration to the end. Old data is messier than anyone expects. Start cleaning customers, suppliers, items and opening balances in the first weeks.
  • Thin testing. Test whole processes, from order to cash and from purchase to payment, with real data and the people who will use the system.
  • Too little training. People who do not trust the system go back to spreadsheets. Train by role, on your own data, close to go-live.
  • Switching everything on at once. Going live with every module in every branch on one day multiplies the risk. Phase it by module or by site.

Before you choose a system

Map your main processes as they run today and mark the pain points. Decide which processes can follow the product’s standard way and which truly need to stay different. That list is your shortlist test: ask each supplier to show those processes, with your sample data, not a generic demo.

During the project

  • Hold a short weekly decision meeting, so open questions do not block the team.
  • Keep a list of requested changes and their cost, and approve each one deliberately.
  • Rehearse the data migration at least twice before the real one.
  • Agree in advance what must be true before go-live, and stick to it.

After go-live

Expect the first month to be slower than the old way. Keep the project team in place, fix issues daily and close the first month-end together. Once the system is stable, the data it holds becomes the base for reporting, automation and AI agents that act on it.

Common questions

How long does an ERP implementation take?

It depends on size and scope. A phased approach gets the first modules live sooner and spreads the risk.

Should we customise our ERP?

Only where the process is a real advantage. Otherwise, adopt the standard and keep upgrades simple.

When should data migration start?

In the first weeks of the project. Cleaning master data takes longer than almost anything else.

Related