Fachbeitrag zu ERP- und Business-Central-Projekten
Article · Data foundation

Never starting over again: how an ERP grows with the business

Why ERP projects keep starting from zero, what Evergreen IT has to do with it, and what a system needs to grow with the business.

Frank Maier21.07.20266 min read

What this article is about

  • The expensive pattern: starting from zero every few years
  • Evergreen IT keeps the version current, an Evergreen ERP keeps your business current
  • Four markers: strategy, data foundation, owners, change as routine
  • The counter-image: the monument ERP
  • The test: how long does a requirement take to reach the system?

Why do ERP projects keep starting over?

ERP projects start over because, for years, changes get filed at the edge of the system instead of inside it. Almost every mid-market company knows the pattern: a system is selected, implemented, customised and run, and for the first few years it fits. Then new products arrive, a subsidiary, a different pricing model, a customer with their own supply chain requirements. Each of those changes gets represented somehow, usually alongside: a side list here, a workaround there, a report someone builds by hand at month end. At some point it is no longer the system carrying the business, but the collection of workarounds.

After five to seven years, the gap between what the system represents and what the company actually does has grown so wide that nobody wants to bridge it any more. Then comes the sentence we have heard countless times in over 25 years: "We need a new system." And the next project begins. With an analysis phase, migration, training, back to zero.

The cost of this pattern appears on no invoice. A new project budget every time. Knowledge that disappears with the old system every time. A year of attention from the people who should be developing the business every time. And the uncomfortable truth: it will happen again next time if nothing about the causes changes.

It is not the software

The obvious suspect is the software. It is rarely the culprit. Business Central, SAP and comparable systems are mature, and in the cloud they solve a problem that used to be a real project: the version level. Microsoft applies two release waves a year automatically, security updates come with them. Technically your system is more current than an on-premises system ever was.

And this is exactly where a confusion arises that we need to clear up before the rest of this article makes sense.

Evergreen IT keeps the version current. An Evergreen ERP keeps your business current.

Evergreen IT is an established term and describes exactly what the cloud delivers: components that are always up to date, continuous small updates instead of a big-bang migration. That is good, that is solved, and you do not need us for it.

An Evergreen ERP means something else. Not the version level, but the state of your business. The question is not "are we on the latest version?" but "does the system represent what we do today?". Microsoft solves the first for you, automatically. No vendor can take the second off your hands, because it depends on your data foundation, your processes and your decision paths.

That is why the following sentence is not a contradiction but the norm: your Business Central is fully up to date. And outdated all the same.

Being current is not a property of the software. It is a property of the organisation running it.

Four markers of an ERP that grows with you

1. Strategy is connected

The roadmap for the system derives from the company strategy, not from the vendor's release plan. A company that wants to sell in three more countries in two years decides differently today about entity structure and master data ownership. Where that does not happen, ERP planning is a pure IT topic, and the business only shows up when something does not work.

2. The data foundation holds

Master data and rules have a single, owned truth and are cleanly separated from the application layer. That sounds technical but it is the commercial core: the data foundation is what remainswhen systems change. Software is a rented present, your data is lasting capital. And every AI function you use in future will decide on exactly this basis.

3. Processes have owners

Processes are documented, and for each one there is a person allowed to own changes. The second part is the harder one. Most systems do not age because documentation is missing, they age on deferred decisions: a requirement is on the table, three departments have an opinion, nobody is allowed to decide, so it waits.

4. Change is routine, not a project

This is the result of the first three markers and at the same time the only proof that counts: a new requirement from the business takes weeks, not quarters. Where that holds, no backlog builds up, and the point where only a new system will help never arrives.

The counter-image: the monument ERP

Stable. Patched. On the latest version. The staff know it, the processes are established, everyone is satisfied. In principle. Because when the business raises a new requirement, the waiting begins: coordination, quote, implementation slot in the quarter after next.

A monument ERP never draws attention because nothing breaks. You notice it when the questions stop. When nobody in the company raises requirements any more because "that takes forever here anyway" has become accepted truth, the system is already slowing the business. And nobody sees that cost.

The test you can run today

Take the last ten change requests from the business and measure a single number: the time from "the business needs X" to "X is live". Not the implementation time, but the whole distance, including coordination and decision.

If the honest answer is weeks, your system is growing with you. If it is quarters, it is lagging behind, regardless of how current the version is. The place where most of the time is lost also reveals the next step: is it in the data, because nobody knows which figure applies? In the processes, because how things run today is undocumented? Or in the decision, because nobody is allowed to make it?

Why AI makes this urgent

Until now a monument ERP was inconvenient but manageable. People compensated for the gaps: they knew which of the three customer lists was right and which report you would rather not put in front of the management board. An AI compensates for nothing. It reads the data as it is and decides on that basis, fast and at scale.

Gartner expects that by the end of 2026 around 60 % of AI projects will be abandoned where an AI-ready data foundation is missing. That is not an AI forecast, it is a finding about data. Meeting the four markers above therefore does not only make your company able to change, it makes it AI-ready along the way. Both are the same piece of work.

The first step

An Evergreen ERP is not a product you buy and not a maintenance contract. It is a state a company reaches, and the road there starts with an honest assessment: where exactly does your organisation lose the time between requirement and live system? That can be answered in days, not months. Let us look at it together in a conversation, senior-led and with Business Central as home.

Frequently asked questions

Is an Evergreen ERP the same as Evergreen IT?
No. Evergreen IT keeps the version current, which the cloud handles automatically. An Evergreen ERP keeps the state of your business current, and that depends on the data foundation, processes and decision paths.

Do I need a new system for this?
In most cases, no. The four markers can be built in your existing Business Central. Changing systems without doing this work only creates the next fresh start a few years later.

Is this a maintenance contract?
No. Applying updates is operations and automatic in the cloud anyway. This is about how quickly your company gets changes into the system.

Where do I start?
With the measurement: time from requirement to live, for the last ten changes. The result shows which of the three levers to address first.

Related

More insights

Article · Data foundation

Three revenue levers that sit in your ERP

Delivery reliability, quote speed, service quality: where revenue is made in the ERP, and why all three depend on your data.

08.09.2026Read
Webinar · Data foundation

AI needs order

Data and processes as the foundation. Including a live look at AI-assisted duplicate checking.

12.08.2026Watch
Article · Data foundation

Business Central vs. SAP in the mid-market

Business Central or SAP in the mid-market: an honest question of fit rather than a matter of principle.

08.06.2026Read

Where does your project really stand?

Talk to a senior, not to a sales rep.

Arrange a conversation