Case Studies/Manufacturing
ManufacturingERP MigrationSME · 120 employees

How a Manufacturing SME Reduced ERP Import Errors by 85%

A French industrial parts manufacturer was losing 14 hours per week to manual catalog correction. Six weeks after adopting ImportCheck, that figure dropped to under 2 hours — and their first ERP go-live shipped on schedule for the first time in the company's history.

Industry

Industrial parts manufacturing

Company size

120 employees

ERP

Sage X3

Time to results

6 weeks

Key results at a glance

85%

Reduction in import errors

34% → 5% error rate

12h

Saved per week

Across operations team

6 wks

Time to full deployment

No developer required

€38k

Avoided ERP delay costs

vs previous migration

01

Customer profile

The customer is a family-owned industrial parts manufacturer based in the Lyon metropolitan area, founded in 1987. With 120 employees across two production sites, they supply mechanical components to automotive and aerospace sub-contractors throughout France and Belgium.

Their product catalog spans approximately 8,400 active SKUs — structural fasteners, precision bearings and machined assemblies — maintained by a two-person purchasing team and updated by five to eight field sales representatives depending on the season.

🏭

Industry

Industrial parts manufacturing

👥

Team size

120 employees, 2 purchasers

📦

Catalog size

~8,400 active SKUs


02

The problem

In early 2025, the company began migrating from an aging custom ERP to Sage X3. The IT project manager estimated an eight-month rollout. By month three, catalog imports had become the primary bottleneck.

“Every Monday morning, Sophie from purchasing would open the import log and find the same types of errors. Invalid price format, missing supplier code, duplicate reference. She'd spend her entire morning correcting rows one by one in Excel. By the time she re-imported, half the corrections had introduced new errors.”

— IT Project Manager, industrial parts manufacturer

The core issues were structural, not human. The catalog was maintained across multiple Excel files by contributors with different habits. No validation occurred before the file reached the ERP. The ERP rejection messages were technical and difficult to interpret without developer involvement.

Pain points identified in initial assessment

34% of submitted rows were rejected by Sage X3 on average
No pre-validation step before upload — errors discovered post-rejection
Error messages from the ERP referenced internal field IDs not column names
5–7 manual correction cycles per file before a clean import
IT team involved in every import cycle to interpret logs
One delayed go-live already — 6-week overrun at an additional cost of ~€38k
No visibility into which suppliers or contributors introduced the most errors
Operations team had no way to self-serve; every correction required IT

03

Implementation

The IT project manager discovered ImportCheck during a search for ERP catalog validation tools. The evaluation-to-deployment timeline was deliberately kept short: two weeks of parallel testing, four weeks of operational rollout.

Week 1–2

Evaluation & parallel testing

  • Created ImportCheck account; uploaded the last three historical import files to compare detected errors against known Sage X3 rejections.
  • ImportCheck identified 91% of the errors that Sage X3 had previously rejected — plus 23 additional issues the ERP had silently accepted but that caused downstream data quality problems.
  • Shared findings with the purchasing team lead. Decision taken to proceed without any IT development work.
Week 3–4

Process integration

  • Established a new "validate before upload" step in the catalog update process: contributors export from Excel, upload to ImportCheck, receive the error report, fix, re-upload.
  • Purchasing team trained in 45 minutes — no technical background required. The error report uses column names from their own file, not ERP field IDs.
  • IT team removed from the import correction loop. They remained as escalation point only.
Week 5–6

Operational rollout

  • All eight catalog contributors added to the team account. Each contributor validates their own section before submitting.
  • Error rate at first Sage X3 submission dropped from 34% to 6% in week five, and to 4% by week six.
  • First full catalog import of the new ERP go-live completed on the scheduled date — no overrun.

No developer involvement after day one

The entire rollout was owned by the IT project manager for setup and the purchasing team lead for day-to-day operation. No custom integration work, no scripts, no API configuration — the operations team ran it independently from week two onward.


04

Results

By the end of week six, the team had fundamentally changed how catalog imports worked. The results below are based on data from the four weeks prior to ImportCheck adoption vs. the four weeks immediately after full rollout.

MetricBeforeAfter (6 weeks)
Import error rate (rows rejected)34%< 5%
Avg. correction cycles per file5–7 cycles1–2 cycles
Time spent on catalog correction per week14 hours< 2 hours
IT team hours per import cycle2–3 hours0 hours
Time from file ready to clean ERP import3–5 daysSame day
Contributors able to self-validate0 of 88 of 8
ERP go-live on scheduled dateNo (6-week overrun)Yes

“What changed is not the tool — it's who owns the problem. Before, Sophie would discover errors after the ERP rejected the file and come to us to decode the log. Now she catches everything before she even submits. IT hasn't touched an import file in two months.”

— IT Project Manager, industrial parts manufacturer
Name withheld at customer request

05

Detailed metrics

Time saved — weekly breakdown

Manual error correction in Excel88%
8h before1h after
Re-import attempts75%
2h before30 min after
IT interpretation of ERP logs100%
2.5h before0h after
Supplier communication re: errors78%
1.5h before20 min after

Error type breakdown (pre ImportCheck)

Invalid or missing price format28%
~2,350 rows
Missing or invalid supplier code22%
~1,850 rows
Duplicate product references18%
~1,510 rows
Encoding / special character errors15%
~1,260 rows
Required field empty10%
~840 rows
Other7%
~590 rows

€38,000

Avoided ERP delay costs

Previous migration overran by 6 weeks; this one shipped on time

624h

Operational hours recovered per year

12h/week × 52 weeks — reallocated to commercial activities

7:1

Return on subscription

Year 1 value vs annual subscription cost at Pro plan


06

Lessons learned

Looking back on the rollout, the IT project manager identified four factors that determined the speed and depth of the improvement.

1

Validate before — not after — the ERP

The entire previous process was reactive: submit to the ERP, read the rejection, fix, repeat. Moving validation before the upload removed the ERP from the correction loop entirely. Errors were caught in a familiar tool (ImportCheck) rather than in cryptic ERP logs that required IT to interpret.

2

Make the error report readable by the person who can fix it

ERP error messages reference internal field identifiers. ImportCheck reports reference the column headers from the uploaded file — the same names the purchasing team uses every day. This single change eliminated the IT translation step entirely.

3

Distribute validation to the contributors, not the reviewer

The previous model had one person (Sophie) responsible for validation of the whole file after all contributors had submitted. With ImportCheck, each of the eight contributors validates their own section before submission. Errors are caught closer to their source and the reviewing step becomes a formality.

4

Measure the error rate from day one

Uploading historical files during the evaluation phase gave the team a baseline: 34% row rejection rate. Without that number, it would have been difficult to demonstrate the value of the new process internally. The IT project manager recommends any team considering ImportCheck to start by uploading their last three import files before making any process changes.

What they would do differently

The IT project manager noted that he would have introduced ImportCheck at the very start of the migration project — not three months in. “We lost time we didn't need to lose. The validation problem was visible from the first test import. We assumed it would get better. It didn't, until we addressed it directly.”

Ready to get similar results?

Start validating your catalog files today

Upload your last import file and see the errors ImportCheck catches in under 90 seconds. No credit card, no developer, no setup.

Like this customer, most teams identify their worst recurring errors in the first 5 minutes.

Start your free 14-day trial
No credit card required Results in under 2 minutes No developer needed