Do we have to replace our ERP, or can we connect around it?

Why replacing it is tempting, and risky

A new system promises a clean start. In practice, replacing a core system touches every department at once, takes longer than planned and moves your most important data under time pressure. Many businesses find that the real problem was not the system itself but the lack of connections around it. Before you commit, it is worth looking at the cheaper, lower-risk options.

Four options, from least to most disruptive

1. Connect it. Keep the system and link it to the tools around it, so that data stops being re-typed.
2. Build around it. Put modern portals, reports or workflows on top, and leave the old system as the engine in the background.
3. Replace it in stages. Move one function at a time, such as the warehouse, then pricing, then orders, with the business trading throughout.
4. Replace it outright. Sometimes right, but the costliest and the one with the most risk.

What each option looks like in practice

Illustrative example (connect it): A distributor's order system is stable and accurate, but customers email orders that staff retype, and carriers' tracking details are copied back by hand. Connecting the order system to an order inbox and the carriers removes most of the typing, and the core system stays exactly as it is.

Illustrative example (build around it): A manufacturer's planning system is dependable but has a dated screen that only office staff can use. A simple web portal that reads its data lets sales staff and customers check stock and order status without touching the core system.

Illustrative example (replace in stages): A business's warehouse module is the weak point. It moves warehouse operations to a new scanning-based tool that talks to the existing system, then later retires the old module, while orders, accounts and customers carry on unchanged.

Signs you can probably keep it

• The core records are accurate and the system is stable.
• The pain is mostly manual re-keying and slow reporting.
• The supplier still supports it and issues updates.
• It has an interface other programs can use, or one can be added.

Signs it is time to move on

• It cannot be connected to anything, even with effort.
• Only one person understands it, or the supplier has gone.
• It blocks things you need to do, not just things that are awkward.
• It runs on technology that is no longer supported and has become a security risk.
• The cost of keeping it going is climbing every year.

Why staged beats big-bang

Big-bang rewrites fail because everything has to work on day one. Staged change delivers value early, lets you learn as you go, and can be paused or redirected if priorities change. It also keeps the business running while you modernise, and it spreads the cost over time.

A simple way to decide

1. List the five biggest frustrations with the current system.
2. For each, ask whether it is caused by the system itself or by missing connections around it.
3. Estimate the cost of fixing the connection problems, and compare that with a replacement.
4. Check what you would lose in a replacement: hidden duties, special reports, customer-specific rules.

A good first step

Map what the system does, who relies on it and where the data goes. That map usually shows that a smaller, cheaper change delivers most of the benefit, and it gives you facts to take into any supplier conversation. The cost question is covered in how much it costs to connect two business systems, and if you decide to move gradually, our guide to modernising a legacy system without stopping the business sets out the steps.

Common questions