Off-the-shelf ERP or custom build: which one fits you

This question comes up at the start of almost every ERP project, and it is usually answered the wrong way round.

The most common mistake runs in reverse: standard inventory accounting gets built from scratch, while the company's genuinely distinctive process is squeezed into an off-the-shelf template. Both cost money and time.

The question that decides it

Split your processes into three categories.

Standard processes. Payroll, cash handling, document flow, attendance. These are almost identical across companies and nobody competes on them. Building them from scratch is money spent for nothing.

Industry processes. Production norms, weighbridge control, warehouse structure. Similar within an industry, different outside it. A system built for your industry works here.

Your advantage. How you serve a client, how you calculate a price, how you assemble an order. If that process is what sets you apart in the market, do not force it into a template.

The mixed option, which is the most common

In practice, for most companies the right answer sits in the middle.

The core comes from a ready system: inventory, cash, payroll, reporting. These are standard, and spending time on them makes no sense. On top of that, your own distinctive module is built from scratch.

Take a metal forming plant. Inventory, sales and payroll are ready-made. Cutting plan optimisation and waste accounting are custom, because that is exactly where the plant makes or loses money.

This approach cuts the price by a factor of two or three and shortens the timeline.

The decision matrix

Assess four factors. If three point towards ready-made, take the ready system.

Budget and timeline are often decisive. If the system is needed in three months, the custom option does not exist.

The hidden costs of a ready system

Ready-made does not always mean cheaper. Costs appear in three places.

• Adapting your process to the system requires retraining staff
• Every additional feature you need is a separately paid change request
• Dependency on the system owner if they raise prices or shut down
• In a multi-tenant system your request sits in a shared queue
• Difficulty of migrating data out later if you change direction

So when you evaluate a ready system, ask: who owns the code and the database, what happens if the system stops, and what an additional module costs.

The hidden costs of building from scratch

The other side has costs too, and they are usually left out.

Support. In a ready system it is priced in and shared across other clients. In a custom system support is yours alone - $300 to $2,000 a month.

Defects. New code always ships with defects. A ready system has been tested across dozens of companies; yours is tested only at yours.

Time. Four months is not simply waiting. Throughout it you keep working the old way, and the losses continue.

How to test the decision

There are two practical steps before committing.

1. See the ready system running. Not a demo - a live system with real users. Talk to a client in your industry. Often at this point it becomes clear that the system covers eighty per cent of your process and the remaining twenty per cent does not matter.

2. Run a discovery stage. Two weeks, process analysis and a written document. At the end of it the choice becomes a calculation rather than an assumption. It costs $3,500 and it is needed either way.

In summary

The question is not technical, it is commercial: does this process set you apart or not.

A practical order:

1. Split your processes into three categories: standard, industry, your advantage
2. Consider a custom build only for the third category
3. Cost the mixed option - a ready core with a custom advantage module
4. Before choosing a ready system, see it running in real conditions
5. Start with a discovery stage either way