ERP projects fail before kickoff — here is exactly where
The costliest ERP decisions happen before any system is chosen: unowned scope, undocumented processes, unpriced expectations. An executive checklist for what must happen before signing any contract.
When an ERP project stalls, blame usually lands on the implementer or the software. Yet in most situations we've reviewed, the causes were present and visible before kickoff — nobody had read them out loud.
Cause one: nobody owns the scope. When "implementing the system" is itself the goal, scope grows with every meeting. Correct goals are always operational: "a five-day financial close," "inventory accuracy above 98%," "order-to-invoice without manual touches." Statements like these automatically decide what enters scope and what gets refused.
Cause two: processes documented on paper, not on the ground. The gap between "how we're supposed to work" and "how we actually work" is where timelines die. Before any configuration, walk one real order from receipt to collection and record every stop and every exception — half your "requirements" will change.
Cause three: customization without a cost register. Every deviation from standard has two prices: building it today, and dragging it through every future upgrade. Disciplined projects run a customization register that forces a written business case from whoever requests a change — and rejects everything else by default.
The pre-signature checklist: a measured operational goal per business unit; field documentation of the three highest-impact flows; an executive sponsor who attends rather than delegates; written exit criteria per phase; and a data plan (who cleans it? who reconciles it?) before discussing a single screen.