26 May 2026
Most ERP projects that go wrong share one trait: the vendor consultation started before the company knew precisely what it wanted. As a result, the most seductive demo ends up defining the need. And you discover the real scope mid-project, at change-order rates.
Upstream scoping is not paperwork. It is what turns a gut-feeling purchase into a comparable decision. What you should have in hand before sending out a single request for proposal comes down to six workstreams.
Describe your processes as they are, not as they should be
Take your five critical processes, from order to invoice, from purchasing to receipt, and describe what actually happens: who enters what, in which tool, with which exceptions. The exceptions are what matters. They are where standard ERPs bend and where budgets explode.
A two-hour workshop per process with the people who do the work is often enough. No idealised diagrams: facts.
Sort the scope: essential, useful, superfluous
Sort every requirement into three columns: what must work on day one, what can wait six months, what gets talked about but nobody will use. This sorting is your best protection against scope creep, and your best negotiation asset.
A useful rule: if nobody can name the user of a feature, it goes in the third column.
Look your data in the eye
Data migration is the most underestimated workstream of ERP projects. Before consulting, know the answers to three questions: which data must be migrated, what state it is really in (duplicates, empty fields, diverging references), and who will be able to clean it. A polluted client file does not become clean by changing tools.
Set the non-negotiable constraints
The real maximum budget, not the one displayed for negotiation. The periods when the business cannot be disturbed: closing, high season, regulatory deadlines. And the actual availability of your teams. An ERP project consumes business time, and the half-time internal project manager who does not exist is the leading cause of drift.
Define the selection criteria before seeing the offers
Decide calmly what matters: coverage of the essential scope, full five-year cost (licences, integration, maintenance, changes), proximity and responsiveness of the integrator, reversibility. Weigh the criteria. Then, and only then, consult.
Companies that do it the other way round choose on charm, and regret it by month twelve.
What this work changes in the offers you receive
A precise consultation file has an immediate effect: offers become comparable, serious integrators respond better, and the others disqualify themselves. Price differences can be explained instead of endured.
Allow a few weeks for serious scoping, alone or with support. Set against the cost of an ERP project going off the rails, it is the cheapest insurance on the market.