What is an Evergreen ERP?

An Evergreen ERP is a Business Central that grows with the business instead of being replaced every few years. Not to be confused with Evergreen IT: staying on the current version has long been standard in the cloud, with Microsoft delivering two release waves a year plus monthly updates, automatically and without a migration project. Evergreen ERP, by contrast, means the state of your business, and nobody keeps that current automatically. Four markers make the difference: a data foundation that carries the business, processes with named owners, a methodology that treats change as routine, and requirements that go live in weeks rather than quarters. The state is not reached when a project closes, but when the next change in the business no longer takes the system by surprise.

Why are “up to date” and “outdated” not a contradiction?

The term evergreen has been established in the Microsoft world for years, though in a narrower sense. Evergreen IT describes systems that receive continuous small updates instead of being replaced every few years in a large migration. For Business Central in the cloud, that is solved: Microsoft ships two release waves a year, automatically.

Which is exactly why the version is the wrong question today. It is current. And your system still reflects what your business did seven years ago.

Microsoft publishes the schedule for these waves in advance, so they are plannable and not an event that takes you by surprise. For operations the topic is therefore settled, and that is exactly why the version level no longer works as a yardstick for whether your system fits your business.

Source: Microsoft Learn: Update cycles

An Evergreen ERP therefore means something different: not the state of the software, but the state of your business. Nobody keeps that current automatically. The difference in detail is in Evergreen IT or Evergreen ERP.

How do you recognise it? Four characteristics

1. Strategy is connected. The ERP roadmap follows from your business strategy, not from Microsoft's release plan. If you cannot say what your customers' business will be in five years, you cannot say what the ERP will need to do by then.

2. The data foundation holds. Clean, separated from the application layer, universally usable. Without it, every change becomes an archaeology project, and every AI agent scales your duplicates instead of resolving them.

3. Processes have owners. Documented, and someone is allowed to decide. Without named owners, every new requirement turns into a debate of principle, and the most expensive sentence gets said: "We'll decide that later."

4. Change is routine, not a project A new requirement takes weeks, not quarters. You can't buy that; it's the result of the first three points.

The four characteristics belong together and cannot be bought individually. Whoever wants to check them goes through them in this order, because each one presupposes the previous one.

What is the opposite? The monument

Stable. Patched. On the latest release. And still slowing you down, because it reflects the business of seven years ago. People know it, the processes are settled, everyone is happy. Mostly.

And when the business asks for something new, the waiting starts: alignment, a quote, a delivery date in the quarter after next. Not because Business Central could not do it, but because every change eats time and budget you need elsewhere.

What is treacherous about the monument is that nobody made a mistake. Every individual decision was right at its time, the system was maintained, the updates ran. The distance to the business did not arise through negligence but through time.

That is also why it does not get noticed. There is no outage, no error message and no incident that forces a decision. There is only a response time nobody wanted to measure, and a spreadsheet next to the system that has meanwhile become part of the process.

Why does AI make this urgent?

For years a slow ERP was annoying but survivable. People worked around it, with a spreadsheet on the side and one more workaround.

That is changing. With AI, requirements can be delivered that used to be a project. But only if the foundation holds. An agent works on your data, not instead of it. Point AI at grown master data and you get your mess faster, not cleaner. This is why becoming AI ready and becoming evergreen are the same work.

The difference from before is the price of waiting. As long as adjustments were expensive and slow anyway, a sluggish ERP cost convenience. Now it costs opportunities, because competitors with a clean data foundation implement things that are still a project for you.

The good news in this: it is not two pieces of work but one. Whoever puts the data foundation in order, documents the processes and names the owners becomes AI-ready and Evergreen at the same time. That is why it pays to plan both as one undertaking rather than two.