How to choose an ERP without running a six-month bake-off

A traditional ERP selection runs three vendors through scripted demos over four to six months, scores them against a 400-line requirements matrix, and produces a decision that everyone can defend and nobody believes in.

The problem is that a demo measures the vendor's demo team. It does not measure how the product behaves against your data, your edge cases, or the three processes that are genuinely unusual about your business.

Score the gaps, not the features

Almost every mid-market ERP does 85% of what you need. The decision lives entirely in the remaining 15%, and a requirements matrix weighted equally across 400 lines drowns that 15% in noise.

We ask clients to name the six processes that are genuinely specific to how they operate. Then we test only those, with real data, in a sandbox.

“A demo measures the vendor's demo team. It does not measure the product against your edge cases.”

Ask what the last three failures looked like

Every implementation partner has failed projects. The ones worth hiring will describe them specifically, name what they would do differently, and give you a reference from a client whose programme went badly.

A partner who says they have never had a difficult project is telling you either that they have not done many, or that they are not being straight with you.

Fix the date after fit-gap, not in the proposal

Any date named before fit-gap analysis is a sales instrument. We name ours in week two, after two weeks inside the business, and we have hit it 22 times out of 22.

If a partner commits to a date in the proposal, ask what happens to it when they discover your data quality. The answer tells you how the next nine months will go.