Concrete feedback on ERP projects is rare. Software vendors publish marketing testimonials, integrators share sanitized success stories, and companies that struggled prefer to move on in silence. The result: when an SME launches its own project, it navigates blind.
This article offers a detailed ERP project example, based on a real case in a French industrial SME. Names have been changed, but the figures, timelines, mistakes and results faithfully reflect what happens in the field. If you are in the planning or scoping phase, this case study will give you a realistic picture of what to expect.
This article complements our complete ERP implementation guide 2026 and follows on from our article about the 5 phases of a successful ERP project.
The context: MétalRhône, a growing industrial SME
MétalRhône (fictitious name) is a plastics and metalworking SME based in the Rhône-Alpes region. Here is its profile at the time the project was launched:
| Indicator | Value |
|---|---|
| Headcount | 120 employees (2 production sites, 1 headquarters) |
| Revenue | EUR 18M |
| Sector | Industrial subcontracting (automotive, aerospace) |
| Existing IT | Sage 100 (accounting, sales management) + Excel (planning, inventory, reporting) |
| Key customers | 4 major clients representing 65% of revenue |
The company had been growing at 12% per year for three years. This growth exposed the limitations of the existing system.
The problem: data silos holding back growth
Management identified four major pain points:
- Systematic double entries: customer orders were entered in Sage, then re-entered in an Excel planning spreadsheet. Re-entry errors generated an average of 3 customer disputes per month.
- No real-time inventory visibility: the logistics manager ran a weekly rolling inventory count on Excel. The gap between computerized and physical stock regularly reached 8%.
- Manual reporting: the management controller spent 3 days per month consolidating data from Sage, production Excel files and the HR spreadsheet to produce the monthly management report.
- Compliance under pressure: the FEC export (Fichier des Écritures Comptables — the mandatory French accounting entries file) was functional, but traceability between sales management and accounting relied on manual reconciliations. The DSN for URSSAF (France’s social security contribution reporting) was managed via a separate payroll software with no interface to Sage.
The CEO summed up the situation: “We’re managing an EUR 18 million company as if it were doing 5 million. If we land a new major client, we won’t cope.”
Phase 1 — ERP project scoping (2 months)
Management launched the project in January 2025. Rather than rushing into vendor selection, they devoted two months to scoping. This was a wise choice: as we explain in our article on how to write an ERP requirements document, poor scoping is the number one cause of project overruns.
Setting up the steering committee
The steering committee comprised 7 people:
- Sponsor: the CEO (budget decisions and arbitration)
- Internal project manager: the supply chain manager, chosen for his cross-functional knowledge of business flows
- Members: CFO, production manager site 1, production manager site 2, quality manager, outsourced IT director (service provider)
The steering committee met every two weeks. The internal project manager devoted 60% of his time to the project from this phase onward.
Requirements gathering and specifications document
The project team conducted 14 business workshops over 6 weeks, covering: purchasing, sales, production, inventory/logistics, quality, accounting/finance and HR-payroll. Each workshop produced an AS-IS (current state) and TO-BE (target state) process sheet.
The final specifications document was 85 pages. It covered:
- Target business processes (28 processes identified, 12 prioritized)
- Regulatory constraints: compliant FEC export, DSN, Chorus Pro invoicing for public contracts, IFRS standards for group reporting
- Technical requirements: sovereign cloud hosting, interfacing with the existing MES (phase 1) then replacement (phase 2)
- Volumes: 4,500 item references, 350 orders/month, 12,000 stock lines
Planned budget stated in the specifications: EUR 250K to EUR 300K (excluding internal costs).
Phase 2 — Vendor and integrator selection (3 months)
The shortlist: 3 solutions evaluated
After market research and exchanges with other industrial SMEs through the UIMM network, MétalRhône shortlisted three vendors for demonstrations:
| Criterion | Odoo (Enterprise) | Sage X3 | Cegid XRP Flex |
|---|---|---|---|
| Estimated annual license | EUR 28K | EUR 45K | EUR 38K |
| Estimated integration cost | EUR 140K | EUR 180K | EUR 160K |
| Strengths | Flexibility, modern UX, community | Very established in French industry, native FEC | Strong finance/HR coverage, sovereign cloud |
| Weaknesses | Less mature production module | Aging interface, high cost | Fewer references in plastics manufacturing |
| Similar industry references | 2 | 5 | 3 |
The scoring grid
MétalRhône used a scoring grid out of 100 points inspired by the methodology we detail in our article how to choose an ERP integrator. The criteria and weightings:
- Functional coverage: 30 points
- Industry fit: 20 points
- Total cost of ownership over 5 years (TCO): 20 points
- Integrator quality: 15 points (references, proposed team, geographic proximity)
- Scalability and ecosystem: 15 points
Each vendor gave a 2-day on-site demonstration based on a real business scenario provided by MétalRhône (a complete order-production-delivery-invoicing cycle).
The final choice: Sage X3
Sage X3 won with a score of 76/100, versus 71 for Odoo and 68 for Cegid. The decisive factors:
- 5 verifiable references in comparable industrial SMEs in the Rhône-Alpes region
- Local integrator with a dedicated team of 4 consultants, including a MES expert
- FEC, DSN and Chorus Pro managed natively without custom development
- Consultant daily rate: EUR 950/day (within the market range for a Sage X3 integrator of this size)
Odoo was cheaper but the proposed integrator lacked industrial references. The CEO made the call: “You don’t cut corners on experience when you’re an SME that can’t afford to get it wrong.”
The contract was signed in May 2025. Contracted budget: EUR 280K broken down as follows:
| Item | Amount | % of total |
|---|---|---|
| Sage X3 licenses (3 years) | EUR 42K | 15% |
| Integration and configuration (185 days) | EUR 176K | 63% |
| Custom developments | EUR 32K | 11% |
| Training (8 sessions) | EUR 18K | 6% |
| Data migration | EUR 12K | 4% |
| Total | EUR 280K | 100% |
Phase 3 — ERP deployment in the company (8 months)
This is the longest and most demanding phase. Deploying an ERP in a company requires the mobilization of both the integrator’s consultants and internal teams. For a comprehensive overview of this phase, read our article on ERP implementation budgets.
Actual implementation timeline
June 2025 ████ Detailed design (configuration workshops)
July 2025 ████ Finance + purchasing module configuration
Aug 2025 ██ Summer break (reduced team)
Sept 2025 ████ Sales + inventory module configuration
Oct 2025 ████ Production module configuration (MES)
Nov 2025 ████ Custom developments + interfaces
Dec 2025 ████ Data migration + unit testing
Jan 2026 ████ User acceptance testing (UAT)Custom developments
Despite a specifications document that aimed for 80% standard functionality, three custom developments proved necessary:
- EDI interface with the main automotive client (OFTP2 protocol): 12 days of development
- Shop floor dashboard on a touchscreen in production: 8 days
- Custom export for group reporting in IFRS format: 5 days
Total: 25 days of custom development, i.e. EUR 23,750 (below the planned budget of EUR 32K). The surplus was reallocated to additional training.
Data migration: the underestimated workstream
Migrating data from Sage 100 was the main friction point of this phase. The problems encountered:
- Customer duplicates: 340 customer records in Sage 100, but 87 duplicates identified (same customer with different spellings). Cleanup took 6 days instead of the planned 2.
- Stock history: 18 months of history to migrate, with discrepancies between computerized and physical stock on 15% of references. Decision made: migrate only actual physical stock after a full inventory count.
- Chart of accounts: migrating the Sage 100 chart of accounts to Sage X3 required a complete remapping to ensure FEC compliance during the transition fiscal year.
Lesson learned: data migration consumed 18 days instead of the planned 10. The EUR 12K budget was met thanks to the mobilization of an internal intern for data cleansing, but the timeline slipped by 2 weeks.
Change management
MétalRhône applied several best practices detailed in our ERP change management guide:
- Business ambassadors: 6 key users identified (1 per department), trained ahead of schedule and tasked with cascading to their colleagues
- 8 training sessions: 2 days per group of 8 to 10 people, on the test environment with their own data
- Internal communication: monthly project newsletter, break room demonstrations, online FAQ on the intranet
- Major friction point: the production department (35 people) resisted the change. The former shop floor manager, accustomed to “his” Excel planning spreadsheet, actively hindered adoption. The CEO had to personally intervene to address the situation.
Phase 4 — Go-live and stabilization (3 months)
The phased rollout approach
MétalRhône opted for a phased rollout rather than a big bang. The sequencing:
- February 2026 — Finance, purchasing and sales (HQ + site 1)
- March 2026 — Inventory and logistics (sites 1 and 2)
- April 2026 — Production / MES (sites 1 and 2)
This choice extended the deployment phase by 4 weeks compared to a big bang but allowed the team to smooth out the support workload and fix issues module by module. To understand the pros and cons of each approach, read our article on ERP implementation failures.
Go-live incidents
The first weeks were not without turbulence:
- Week 1: web interface response times hit 8 seconds on the purchasing module. Root cause: a missing index on the supplier orders table. Fixed in 2 hours by the integrator.
- Week 2: the FEC export generated an error on intra-community VAT entries. The integrator corrected the configuration in 1 day. The FEC check using the tax authority’s tool then passed without issues.
- Week 3: the production module displayed manufacturing orders with incorrect bills of materials. Root cause: an error in the routing data migration file. Fixed in 3 days, including 1 day of cross-checking 45 routings.
- Week 4: spike in internal support calls (35 tickets in one week, versus an average of 8 in normal operation). Mainly usage questions, not bugs. The integrator deployed a consultant on site for 2 additional weeks.
Enhanced support (1 on-site consultant + priority hotline) was maintained for 6 weeks after the go-live of the last module, at an additional cost of EUR 14K not included in the initial budget.
Final budget vs. planned budget
| Item | Planned budget | Actual budget | Variance |
|---|---|---|---|
| Sage X3 licenses (3 years) | EUR 42K | EUR 42K | 0% |
| Integration and configuration | EUR 176K | EUR 183K | +4% |
| Custom developments | EUR 32K | EUR 24K | -25% |
| Training | EUR 18K | EUR 26K | +44% |
| Data migration | EUR 12K | EUR 12K | 0% |
| Enhanced go-live support | EUR 0K | EUR 14K | N/A |
| Total | EUR 280K | EUR 301K | +7.5% |
A 7.5% overrun on an ERP project of this scale is actually a good result. The industry average is between 15% and 25% cost overrun.
18-month assessment: did the ERP project deliver on its promises?
Eighteen months after the first module went live, MétalRhône took stock. Here are the before/after KPIs:
| Indicator | Before ERP | After ERP (18 months) | Change |
|---|---|---|---|
| Customer disputes (entry duplicates) | 3/month | 0.3/month | -90% |
| Computerized vs. physical stock gap | 8% | 1.2% | -85% |
| Monthly reporting close time | 3 days | 4 hours | -83% |
| Average invoicing delay | 5 days after delivery | 1 day (auto-generated) | -80% |
| Customer order processing time | 45 min | 12 min | -73% |
| Manual re-entries per day | ~35 | ~4 | -89% |
| FEC / DSN compliance | Manual, error-prone | Automated, built-in checks | Secured |
| Chorus Pro invoicing | Manual (web portal) | Automated from the ERP | 2 days/month saved |
Financial ROI
The ROI calculation includes direct and indirect gains:
- Direct gains: reduction of 1.5 FTEs on administrative tasks (data entry, reconciliation, reporting), approximately EUR 65K/year in loaded cost
- Dispute reduction: EUR 18K/year reduction in customer credit notes
- Inventory savings: 12% reduction in dormant stock thanks to real-time visibility, approximately EUR 40K in freed-up cash
- Total project cost: EUR 301K + internal costs estimated at EUR 80K (team time) = EUR 381K
Return on investment achieved in 3.5 years, counting only recurring direct gains (EUR 83K/year). If indirect gains are included (improved customer satisfaction, ability to absorb growth without additional administrative hires), the actual ROI is likely shorter.
The 5 lessons learned by MétalRhône
After 18 months of hindsight, MétalRhône’s steering committee formalized a lessons-learned review. Here are the five main takeaways, useful for any SME preparing its own ERP project.
1. Invest in scoping — it pays off throughout the project
The 2 months of scoping represented 15% of the total project duration but prevented costly change requests during the implementation phase. MétalRhône estimates it saved between 30 and 50 integration days thanks to precise specifications. Don’t skimp on this phase: read our guide on how to write an ERP requirements document.
2. Budget twice as much for data migration
The “data migration” line item is systematically underestimated. Not because of the technical side, but because of the quality of source data. At MétalRhône, cleaning duplicates, making inventory reliable and remapping the chart of accounts took almost twice the planned time. The recommendation: budget data migration at 8-10% of the total budget (instead of the initial 4%).
3. Change management is not optional
The production department nearly derailed the project. Resistance from a single influential shop floor manager blocked adoption for 3 weeks. Training alone is not enough: you need to identify resistance early, involve the reluctant in the design process, and if necessary, commit to managerial intervention. Our change management guide details the approaches that work.
4. The phased rollout is worth the extra cost
Choosing to deploy module by module cost 4 extra weeks and approximately EUR 15K more than a big bang. But it allowed the team to fix configuration issues (bills of materials, indexes) before they impacted the entire company. For an SME that cannot afford a production shutdown, phased rollout is almost always the best choice.
5. Plan a post-go-live stabilization budget
The 6 weeks of enhanced support (EUR 14K) was not in the initial budget. This is a classic mistake: go-live is not the end of the project — it is the beginning of the critical phase. MétalRhône recommends provisioning 5 to 8% of the total budget for post-go-live stabilization.
What MétalRhône would do differently
In hindsight, the steering committee identified three things it would handle differently:
- Integrate payroll from phase 1: payroll remained on a separate system, which maintains a manual interface for DSN reporting. The integration cost seemed high, but the cost of maintaining two systems is high too.
- Hire a full-time internal project manager: the supply chain manager handled the project at 60% of his time, which created tensions in his operational role. A project manager dedicated 100% for 12 months would have streamlined decision-making.
- Start training earlier: training sessions took place 3 weeks before go-live. That was too late. Some users had not absorbed the processes by the time they switched over. Ideally, training should have started 6 weeks before.
Conclusion: an ERP project example that reflects reality
This ERP project at MétalRhône is neither a textbook success nor a disaster. It is a realistic industrial project, with its compromises, controlled delays and tangible results. The budget overran by 7.5%, the timeline slipped by 2 weeks, change management was complicated in one department, but ultimately the business objectives were met and the ROI is on track.
If you are at the beginning of your planning, start with our complete ERP implementation guide. If you are in the selection phase, our integrator scoring grid will save you time. And if you fear cost overruns, our article on ERP implementation failures will give you the warning signs to watch for.
An ERP project means 16 months of intense effort for 10 years of benefits. MétalRhône does not regret theirs.
