An ERP project that is over budget rarely fails in one moment. It drifts: a change here, a delay there, data that takes longer than planned, a go-live date that moves again. By the time someone asks what went wrong, the money is spent and the team is tired. The good news is that most troubled ERP projects can be recovered. This guide is the recovery plan we use.
Step 1: Stop the overrun
Before fixing anything, stop making it bigger.
- Freeze new scope. No new requirements enter the project until it is re-baselined. Good ideas go on a written list for later.
- Pause custom development that is not essential for going live.
- Keep the data and configuration work going. Those are needed whatever happens next.
Step 2: Find out where the budget went
Overruns feel general, but they have specific causes. Go through time and cost by area:
| Area | What overrun usually means |
|---|---|
| Customisation | The software is being made to copy the old process |
| Data migration | The data was dirtier than planned, or cleanup started late |
| Rework | Requirements changed after configuration, or were never written down |
| Waiting | Decisions on the client side took days or weeks |
| Integrations | Systems were added that were never in the scope |
The pattern tells you what to fix. If most of the money went into customisation, the fix is scope. If it went into waiting, the fix is decision-making.
Step 3: Re-baseline phase one
Write a new, smaller phase one that can go live:
- Core processes only: usually sales, purchasing, inventory, invoicing and accounting.
- Standard flows wherever the software already does the job.
- A clear definition of done for each process, agreed in writing.
- Everything else moved to a dated phase-two backlog, with an estimate.
A smaller phase one that goes live beats a complete phase one that never does. Once the business runs on the system, phase two is far easier to plan.
Step 4: Fix the data in parallel
Data is the most common hidden cause of delay. Give it its own workstream with a template, an owner on your side and a verification step. Migrate only what the new system needs: open balances and active records, not ten years of history.
Step 5: Put decisions back on a schedule
Name one decision-maker on each side with the authority to decide, and hold a weekly decision meeting. Anything that waits longer than a week goes to a sponsor. Slow decisions are one of the most expensive, least visible causes of overruns.
Step 6: Train before go-live, by role
Projects in trouble often cut training. That makes the first month after go-live worse. Train each role on its own processes with its own data before cut-over.
Warning signs a project needs a rescue now
- The go-live date has moved more than twice.
- Nobody can say exactly what phase one must do.
- Change requests outnumber completed tasks.
- Users are still keeping the old spreadsheets "just in case".
- The partner and the business describe the problem differently.
When to get a second opinion
An independent audit is worth it when the team cannot agree on the cause, or when you are considering changing partner. At MFY Innovatech we audit existing ERP systems, especially Odoo, as an official Odoo partner: configuration, data, customisations and scope. We tell you what to keep, what to fix and what a realistic phase one looks like. For companies in Kuwait, see our Odoo partner in Kuwait page.
To avoid the overrun next time, read why ERP projects go over budget and how to plan an ERP budget.
Frequently asked questions
- Should we restart our ERP implementation from scratch?
- Rarely. Most troubled projects have usable configuration and data underneath the problems. Audit first: keep what works, fix what is wrong and restart only the parts that cannot be saved.
- Should we change our ERP implementation partner?
- Only after an audit shows where the problem is. If the issue is scope, data or slow decisions on your side, a new partner inherits the same problems. If the configuration is poor or the partner cannot staff the project, a change can be the right call.
- How do we know if an ERP project can still be saved?
- If the core configuration is sound, the data can be cleaned and the business still agrees on what phase one must do, it can be saved. Warning signs are a scope that keeps growing, nobody able to say what done means, and go-live dates that move every month.
- How long does an ERP rescue take?
- It depends on how far the project has drifted. The goal is to get a reduced, standard phase one live as soon as it is safe, then deliver the rest in planned phases rather than one large push.
Keep reading
Related guides
- Why ERP Projects Go Over Budget, and the Fixed-Scope Approach That Avoids ItWhy most ERP projects run over budget, and the written-scope, phased approach that keeps an Odoo implementation on the number you signed.
- How Long Does an ERP Implementation Really Take? Timelines by Company SizeSmall companies go live in 2 to 4 months, mid-sized in 4 to 9, large multi-company rollouts in 9 to 18. Phase-by-phase timelines and what makes them slip.
- How to Plan an ERP Budget: What Really Drives the CostWhat actually drives ERP implementation cost: licences, scope, data, integrations, customisation, training and support. A practical way to build an ERP budget.
Talk to us directly
Prefer a quick conversation?
Message us on WhatsApp or call. You'll reach the team, not a call centre.

