What actually goes wrong in an ERP implementation
Not “change management.” The failure modes are unclean masters, scope that pretends manufacturing is free, dual running forever, and nobody owning cutover math.
ERP projects rarely fail because the login screen is ugly. They fail because the company tried to buy a future operating model without deciding the present one. The software becomes the place arguments go to die.
Use this as a pre-mortem. If more than two of these already describe your programme, pause sales demos and fix governance first.
Masters are the project
Duplicate customers, vendors without payment terms, SKUs that are really marketing names, and a chart of accounts with every balance as a suspense account. No workflow saves bad masters.
Assign owners: who may create a customer, who may create an item, who may open a G/L account. Write it down before discovery workshops turn into design-by-committee.
Scope cosplay
Teams buy a finance-led system and then insist on full MRP, van routes, and a customer portal in wave one. Each extra surface multiplies master-data and training load.
Omnizone is explicit about waves: finance, HR, stock, and migration first; manufacturing and field depth later. If you need deep shop-floor tomorrow, pick a manufacturing-first product — or wait.
Dual running without an end date
Two systems forever means nobody trusts either. Set a cutover date, a reconciling checklist (AR, AP, bank, stock), and a rule that new documents after cutover live only in the new system.
Keep the old system readable. Do not keep it writable “just in case” for six months unless regulation forces you — and even then, define the exception list.
Nobody owns the cents
Go-live week needs a named accountant who can explain every plug. If the only person who understands opening balances is the vendor consultant flying home Friday, you are not live — you are hoping.
Demand trial balances, open AR/AP ageings, and a sample of customer statements before celebrating. Celebrations are cheap; restatements are not.
Related modules
Related guides
- Moving off accounting software: what breaks and when
- Migrating from Dynamics NAV without losing your history
Related glossary
FAQ
How long should dual run last?
Long enough to prove one full close — often one period — not an open-ended hedge. Extend only with a written reason and owner.
Should we customise heavily before go-live?
Prefer configuration and honest process change. Heavy custom before the first close usually encodes yesterday’s workaround in steel.