Why do many ERP projects still fail in SMEs?

The real problem of ERP projects in SMEs:

An ERP almost never failed "because of the software package".
It fails because the company is launching an organization site thinking about buying a tool. In an SME, the ERP affects all at the same time: finance, sales, purchases, stock, production, HR. If the method is not solid, the project quickly slides to delays, additional costs, demotivated teams... Then a return to Excel files.

The most common symptoms of an ERP that goes wrong:

A blurred perimeter that changes every week.
Data that are not ready or unreliable.
Users who do not understand "why we change".
Arbitrations that are never settled.
A controlled "function by function" parameter instead of "end-to-end process".

1) An ERP decision made too early, without a real diagnosis:

Many SMEs choose a solution before clarifying three key points:
What processes should be standardised?
What indicators should be controlled?
What level of change can the company absorb?

Result: we configure, we correct, we add bypasses... and the project becomes more complex than expected.

2) Insufficient scoping and drifting perimeter:

Without framing, everything becomes urgent and priority. The project becomes a list of internal, sometimes contradictory, requests.

Good framing is concrete:
A validated perimeter.
Clear roles (who decides what).
A realistic schedule.
Acceptance criteria (which means "it's ready").

3) Unprepared data:

The ERP does not "clean" the data: it amplifies them.

Typical causes:
Reference items and inconsistent customers.
Units, taxes, poorly structured families.
Unreliable stocks.
Incomplete history.

When the data are fragile, the ERP appears "bugged" while the problem is upstream.

(4) Neglected adoption:

The number one risk in an SME: teams continue to work "as before", in parallel.

What is often missing:
A simple change management plan.
Training by role, oriented tasks of the daily.
Test scenarios close to reality.
A structured start support.

Why do ERP projects fail most often?

The software package is selected before diagnosis.
The framing is too light or not locked.
The data is not ready.
Adoption is underestimated.
Pilotage has no clear decision rules.

(5) A project management without governance:

Without governance, the project becomes "personal": everyone pushes their need, nobody slices.
It takes a simple mechanics: a sponsor, a project manager, a steering committee, quick arbitrations.

Practical Tips for Successing an LES Project in SMEs:

1) Start with a short but serious diagnosis:

Mapping key processes.
Identify major irritants.
Set KPIs expected.
Validate priorities.

2) Frame before setting:

Perimeter, planning, responsibilities, validation rules.

3) Work the data as a full site:

Cleaning, harmonization, import tests, quality control.

(4) Putting adoption at the centre:

Train, test, accompany, correct quickly after go-live.

Conclusion:

An ERP project in SMEs is not a software purchase. It's an operational transformation.
With a simple method (diagnostic, framing, data, adoption, governance), the risks are greatly reduced and an LES is actually used, reliable, and useful for piloting.