A food packaging manufacturer in Sheffield spent 14 months and just over 120,000 pounds implementing an ERP system that their production team refused to use. Within six months of go-live, the factory was back on spreadsheets. The software still runs. Nobody logs in. That story isn't unusual. Research from Panorama Consulting Group puts the ERP implementation failure rate at 70% or higher, with cost overruns averaging 215% in manufacturing environments.

What "Failure" Actually Means

Most failed implementations don't crash spectacularly. The software works. The data migrates. The system goes live. Then nothing changes.

Production planners keep their spreadsheets open in a second tab. The warehouse team enters data into the ERP because they have to, then checks their own stock sheet because they don't trust it. Sales calls the warehouse directly instead of checking the system. Within three months, the ERP is a reporting tool that nobody reports from.

That's the typical failure. Not a technical disaster. A human one.

The Five Reasons SME Implementations Go Wrong

1. Wrong System for the Business

This kills more implementations than anything else, and it happens before a single line of data moves. A manufacturer picks an ERP designed for service businesses, retail or distribution. It handles invoices and purchase orders fine. But it doesn't understand bills of materials. It can't route a product through production steps. It has no concept of work-in-progress.

So the implementation team builds workarounds. Custom fields. Manual processes. Export-to-spreadsheet reports. Six months and 50,000 pounds of consulting later, the manufacturer has an expensive system that still doesn't match how their factory operates.

The fix is straightforward. Choose a manufacturing ERP purpose-built for manufacturing. Not adapted. Not configured. Built.

2. Data Migration Treated as an Afterthought

Every implementation guide lists data migration as a key step. Every project plan allocates time for it. And nearly every project underestimates it by a factor of three.

Here's what actually happens. The manufacturer has customer records in one system, supplier data in another, product information in a spreadsheet last updated by someone who left two years ago, and stock counts that don't match physical inventory. Merging that into a clean dataset takes weeks of manual work. Duplicate records. Missing fields. Conflicting formats. Data that nobody can explain but nobody wants to delete.

Start cleaning data three months before go-live. Not three weeks. Map every field. Run test migrations. Verify the output against physical reality. If the stock count in the new system doesn't match what's on the shelf, fix it before go-live, not after.

3. The People Problem

According to research published by TechTarget, inadequate change management is among the top three causes of ERP failure. The phrase sounds corporate. The reality is personal.

Dave in the warehouse has used the same stock spreadsheet for eight years. It works for him. He knows where everything is. A new system means learning new screens, new processes, new ways of doing a job he's already good at. Nobody explained why the change is happening. Nobody asked for his input. Nobody trained him properly. So Dave enters the minimum data required and keeps his spreadsheet running underneath.

Multiply Dave by every person in the factory. That's change management failure.

The fix isn't a training day. It's involvement. Get the warehouse team, production planners and sales staff into the selection process. Show them what the system does for their specific job. Train them with their own data, not sample data. And give them someone to call when they get stuck in week two.

4. Big Bang Instead of Phased Rollout

Go-live day. Everything switches at once. Sales, procurement, production, inventory, fulfilment, HR. All on the same Monday morning. By Wednesday, three critical processes have broken, the support team is overwhelmed, and the factory manager is asking whether they can switch back.

Phased implementation is slower on paper. Faster in practice. Start with the module that solves the most painful problem. Usually production management or inventory. Get it stable. Get the team comfortable. Then add procurement. Then sales. Then fulfilment. Each phase builds on the last. Each phase has a smaller blast radius when something goes wrong.

5. No Clear Definition of Success

"We need an ERP" is not a business case. It's a feeling. Without specific, measurable targets, there's no way to know whether the implementation worked.

Before signing anything, write down what success looks like. Order processing time drops from 4 hours to 30 minutes. Stock accuracy exceeds 98%. Production schedule generated in minutes instead of days. Supplier on-time delivery tracked and reported monthly. These aren't aspirational goals. They're the minimum return on a significant investment.

What the 30% Get Right

Successful implementations share three characteristics. They chose software built for their industry and scale. They implemented in phases with clear milestones. And they invested in people, not just technology.

Notice what's not on that list. Budget size. IT expertise. Company size. The 30% who succeed aren't bigger or richer. They're more deliberate.

How Long Should ERP Implementation Take?

Traditional enterprise ERP: 12 to 18 months. Sometimes longer. That timeline includes hardware procurement, custom development, extensive testing and iterative configuration changes driven by scope creep.

Cloud-based ERP designed for SME manufacturers: weeks to a first operational module. The system is configured, not customised. There's no hardware to procure. Data migration runs in parallel. Training starts before go-live.

The difference isn't just speed. It's risk. An 18-month project has 18 months to go wrong. A phased rollout delivering value in the first month maintains momentum and builds confidence.

Getting Implementation Right

Arcflow was built for manufacturers who can't afford the 70% failure rate. No-cost implementation removes the largest financial risk. Phased rollout starts with the module that matters most. Monthly licensing means the manufacturer isn't locked into a multi-year contract before the system has proved itself. And AI-powered automation across 110+ input metrics delivers production planning, supplier analysis and order automation that works from day one, not month 18.

Book a demo to see how Arcflow makes ERP implementation work for SME manufacturers.