Why 73% of ERP Replacements Fail — and What UK Manufacturers Do Instead

The project started in January. By June it was six months behind schedule and £150,000 over budget. The consultants were still configuring the test environment. Staff were exhausted from parallel-running two systems. The board was asking hard questions. Sound familiar?

ERP replacement failure is not rare. It is the norm. And understanding why — and what UK manufacturers are doing instead — is the starting point for making a better decision.

The numbers are stark

OECD research on UK SME ERP implementations puts the failure rate at 73% — meaning nearly three in four projects fail to meet their stated objectives. "Failure" does not always mean total collapse. More commonly it means: went live six to twelve months late, cost 40–80% more than budgeted, delivered 60% of the originally scoped functionality, and achieved 30–40% of the productivity improvements projected.

For a business with 50–150 employees, the realistic cost of an ERP replacement is £300,000–400,000 and 9–18 months of business disruption. The promised return on that investment rarely arrives on the timeline projected.

Why ERP replacements fail

1. Scope creep during customisation

No ERP fits a business perfectly out of the box. Every gap between what the ERP does by default and what your business actually needs becomes a customisation — and customisations multiply. Each one is a dependency. Each dependency is a risk. By the time the system goes live, the "standard" ERP has hundreds of custom modifications that will need to be re-validated with every upgrade.

2. Data migration is always harder than planned

Moving data from one system to another sounds straightforward. In practice, legacy data is messy, inconsistent, and full of exceptions that nobody documented. Data migration alone accounts for a significant proportion of ERP project overruns. And a failed migration means you go live on new software with bad data — which is often worse than staying on the old system.

3. Staff do not adopt the new system

Change management is consistently under-resourced in ERP projects. Staff who have spent years learning the workarounds of the old system are asked to unlearn all of it and rebuild their workflows from scratch. Many revert to the old system — or build new spreadsheet workarounds alongside the new ERP. Adoption failure is not a training problem. It is a design problem: the system was not built around how people actually work.

4. The problem was never the ERP

This is the critical insight that most ERP replacement projects miss. For most UK manufacturing and distribution SMEs, the problem is not the ERP system itself. The data is good. The back-end logic handles orders, stock, and finance correctly. The problem is the interface — the way staff experience the system every day. The ERP works. The UI does not.

A third option

If the problem is the interface, the solution is a better interface — not a new ERP. Sysgraft builds modern, purpose-built web applications that connect to your existing ERP via its REST API. Staff see a clean, role-specific interface built around their actual workflows. Customers get a self-service portal. Management gets live visibility dashboards.

The ERP stays exactly as it is. The data stays where it is. There is no migration, no re-training on new back-end logic, and no disruption to the processes that already work. The interface — the part that was causing the friction — is replaced entirely.

Is this right for your business?

The interface layer approach works best for businesses that match most of these criteria: 50–150 employees, running Dynamics 365 Business Central, OrderWise, or Sage 200, experiencing operational inefficiency from poor UI rather than from bad data or wrong system logic, and unwilling to bet the business on a high-risk replacement project.

The entry point is a discovery sprint: 3–5 days, fixed fee, live API audit of your ERP. At the end you have a clear picture of what a modern interface could deliver for your specific system and a fixed-price proposal if you want to proceed. Valuable either way.

---