ERP projects fail for predictable reasons, and succeed for predictable ones. The software is rarely the cause either way. What separates a system people rely on from one they work around is a small set of habits, most of them visible in the first weeks of the project.
In short
- Scope, sponsorship, data, phasing and adoption decide the outcome.
- The most expensive mistakes are made early and discovered late.
- Standard first; customize where it earns its keep.
- Go-live is the start of the value, not the end of the work.
What drives success
A scope tied to outcomes
"Implement an ERP" is not a goal. "Close the month in five days instead of fifteen" or "know stock in real time across two warehouses" is. Outcomes make trade-offs possible: when a new request appears, you can ask whether it serves them.
A sponsor who decides
Every project reaches points where two departments want different things. Someone with authority has to choose, quickly. Projects without that person do not fail dramatically; they slow down until they stop.
Key users with time
The people who know the processes are also the busiest. If their project work comes on top of a full day, it will not get done. Freeing part of their time is the single most concrete commitment a company can make.
Clean data
Duplicate customers, products without units, balances nobody has reconciled. Loaded as they are, they make a new system look unreliable in its first week. Data preparation starts at the beginning of the project, not at the end.
Phased delivery
A first phase that covers the core and goes live builds confidence and teaches everyone how the system behaves. Later phases are faster and better scoped because of it.
Training and adoption
Role-based training, on the company's own data, close to go-live. Then support on the floor in the first weeks. People adopt what they understand and what makes their own work easier.
Common pitfalls
| Pitfall | What it looks like | What to do instead |
|---|---|---|
| Over-customizing | Every screen rebuilt to match the old system | Try the standard process first; customize for real advantage |
| Scope creep | "While we're at it" added every week | A backlog for later phases, decided by the sponsor |
| Dirty data | Old errors migrated into the new system | Clean before loading; test with a trial migration |
| Thin testing | Only the implementation team tests | Real users, real scenarios, end to end |
| Deadline-driven go-live | Launching because the date arrived | Launch on readiness, with clear criteria |
| No change management | Users hear about the system at training | Communicate early: what changes, why, and for whom |
| Go-live as finish line | The team disbands the next day | Plan the first three months of support and improvement |
Warning signs during the project
- Decisions wait for weeks.
- Key users miss workshops.
- Nobody outside the project team has seen the system.
- The list of customizations keeps growing.
- Data migration has not been rehearsed.
- Status is reported as "on track" without anything to show.
None of these is fatal alone. Together, and unaddressed, they are how projects stall.
A go-live readiness checklist
- End-to-end scenarios have been run by the users who own them.
- Opening balances and stock match the old system.
- Every user has an account, the right access and has been trained.
- Printed documents and statutory outputs have been checked.
- There is a cutover plan with owners and a fallback.
- Support for the first weeks is scheduled, with named contacts.
The mindset
Treat the project as a business change made possible by software, not as an IT installation. The questions that matter are about how the company wants to work: who approves what, when stock is counted, how prices are set. The system then makes those decisions consistent. Companies that use the project to settle such questions end up with more than new software.
After go-live
- First week: close support, quick fixes, daily check-ins.
- First month: the first month-end close, done together.
- First quarter: review against the original outcomes; plan the next phase.
Frequently asked questions
What is the most common reason ERP projects struggle?
Most often, slow decisions and unavailable key users. Technical problems are usually solvable; an absent decision-maker is not.
How much customization is too much?
There is no fixed number. A useful test: would you still want this change if you had to pay for it again at every upgrade? If yes, it probably earns its place.
Should we go live with everything at once?
Rarely. A focused first phase lowers risk and gives the team a working system to learn from before the scope expands.
Book a free executive diagnostic or explore our Odoo, AI and cybersecurity services.