Every company hits this moment eventually: the warehouse manager quotes one figure, accounting quotes another, and sales quotes a third. None of them are lying — each is simply working from their own file.

That is exactly when the question "do we need an ERP" comes up.

This article answers it: what ERP is, which problems it solves, when it's needed, when it isn't, what it costs, and how implementation actually runs.

What ERP is, and what it isn't

ERP stands for Enterprise Resource Planning. The name is a little dated — modern ERP does not just plan, it actively runs day-to-day operations.

In practical terms, ERP means this: one database, with several connected modules on top of it.

When a salesperson enters an order, stock drops immediately. When goods leave the warehouse, an invoice is generated automatically. When an employee clocks in, attendance is logged, and payroll is calculated from that at month end. Nothing gets entered twice.

What ERP is not:

• Not accounting software. Accounting handles statutory reporting; ERP handles operations. They connect through integration — neither replaces the other.
• Not a CRM. A CRM manages clients and deals. ERP may include a sales module, but running a sales pipeline is a CRM's job.
• Not a magic fix. Feed a disorganised process into an ERP and what comes out is digitised disorder.

Which problems ERP solves

The table below covers the most common situations. Left is where things stand now, right is after ERP.

The biggest shift is in the last two rows. The rest saves time; these two stop the business depending on any one person.

Five signs you need ERP

• Headcount has passed 30 and departments cannot see each other's data
• The same data gets entered by hand several times a day
• Preparing a monthly report takes several people several days
• A dispute over stock or receivables comes up multiple times a month
• You are opening a new branch or line of business and the current setup will not transfer

Three of these mean it's time to think about it. Four means you are already losing money.

Do the maths. Say payroll takes three days a month to calculate, and the accountant earns $700. Three days is roughly $100. That's $1,200 a year. Add stock disputes, delayed decisions and missed orders on top. In a 50-person company this figure typically lands between $8,000 and $20,000 a year.

When you don't need ERP

Most articles skip this section, because it doesn't help anyone selling ERP. But it's worth knowing.

Under 20 employees. In a company that size, people talk to each other directly. A system becomes extra overhead for limited benefit. A well-organised spreadsheet and a simple inventory tool are enough.

Your processes haven't settled yet. If the way you work changes every month, fixing it into a system is premature. Stabilise the process first.

There's only one problem. If all you need is attendance tracking, a full ERP is unnecessary — a standalone system does the job and costs a tenth as much.

Leadership isn't ready for change. ERP changes how work happens. If the ask is "keep everything exactly as it is, just on a computer," the project will fail.

ERP modules: which ones you actually need

You don't need every module at once. Most companies start with two or three.

Off-the-shelf or built from scratch

This is the most common question, and the answer depends on your process.

The most common mistake runs backwards. A custom system gets built for simple inventory tracking, while a genuinely distinctive production process gets squeezed into an off-the-shelf template. Both waste money.

The right question is: does this process set me apart from competitors? If not — go with a ready system. If it does — build from scratch.

What it costs and how long it takes

Real ranges for GCC business:

Factors that move the price:

Costs that sit outside the budget: server or hosting ($50-300 a month), third-party licences, SMS services. Ask about these separately when you receive a proposal.

How implementation runs

A good project has six stages.

1. Discovery and scope - 2 weeks. The process is studied, a written document is produced, and a firm budget comes out of it. It's normal for this stage to be paid — it determines what system you actually need, and the document belongs to you.

2. Architecture. Screens, data model, technology choices.

3. Sprints. A working version ships every two weeks. You open it and give feedback. This is the most important condition — not reading a report, but opening the system and seeing it.

4. Testing. Checklist review, load and security testing.

5. Go-live. Deployment, data migration, staff training.

6. Support. The warranty period, then an SLA agreement.

Why ERP projects fail

In our experience there are three main causes, and none of them is technical.

Scope was never written down at the start. It begins with "we'll clarify as we go," then a new requirement gets added every week. Each one adds two weeks. Add ten and that's twenty weeks gone, with the budget doubled.

No internal champion. The system arrives from above, and staff see the new screen one day and resist it. The fix is simple: involve one or two people from each department throughout the project. They give feedback, then train their colleagues.

Migration left until the end. The new system is ready, but five years of data is still sitting in the old one. Migration is separate work and it's expensive. Plan for it from day one and run a trial migration in month one.

Questions to ask before choosing an ERP

When you receive a proposal, ask these:

• Do you have a client in my industry, and can I talk to them?
• Can you show me a live system - not a demo, with a real user?
• Is data migration included in the price, and how long does it take?
• Who will own the code and the database?
• Is there support after handover, and what does it cost?
• What do you do if staff don't end up using the system?

The last two matter most, and they're the ones most often left out.

In summary

ERP isn't software — it's a decision about process. It pays off from around 30-50 employees, and it only works once your processes have settled.

Three steps to start:

1. Identify your two most time-consuming processes and calculate their monthly cost
2. Decide whether your process is standard or distinctive - that answers the off-the-shelf-or-custom question
3. Start with a discovery stage. Two weeks and a written document turn the price and timeline from a guess into a calculation