ERP modernization without stopping the business

Almost every business we talk to has the same story about their ERP. It was implemented years ago, it was customised heavily by someone who has since left, and it now sits at the centre of operations in a way that makes it both indispensable and impossible to change. Finance depends on it. Nobody wants to touch it.

The instinct is to replace it. A new platform, a clean implementation, a go-live date. That instinct is what turns an eighteen-month programme into a four-year one, and it is worth understanding why before committing budget to it.

The problem is almost never the technology. Modern ERP platforms are capable and the migration tooling is mature. What breaks these programmes is sequencing: the decision to change everything at once, in a business that cannot stop while you do it.

The first move is not migration. It is to stop the ERP being the only door into your data. That usually means putting a thin integration layer in front of it — an API surface, an event feed, or at minimum a reliable synchronised copy of the data other systems need.

This sounds like a detour. It is the opposite. Once other systems can read and write without going through the ERP's own screens, you can build new capability outside it without waiting for the migration. A supplier portal, a mobile capture app, an approval workflow — none of these need the ERP replaced first, and each one removes load from it.

It also does something quietly important: it turns the ERP into one system among several rather than the system. That is the state you need to be in before replacing it is safe, and many businesses discover that once they are there, the urgency to replace drops considerably.

A phased programme is not automatically cheaper in total. What changes is when you spend and what you get for it. Instead of committing a large budget to a single outcome eighteen months away, you commit to a first phase that is live in roughly three months and either pays for itself or tells you something important.

The first phase is usually the most expensive per unit of work, because it includes the integration layer that every later phase reuses. That is the part worth funding properly. Businesses that try to save money by skipping it end up rebuilding a point-to-point integration for every subsequent process, and pay for it several times over.

The other honest point: this approach needs someone on your side who can make decisions about process, not just about software. If every question has to go through a committee, the phased model loses its main advantage, which is speed of learning.