Every system decision is at its core a make-or-buy question. Do we build, do we buy, or do we start fresh? Build, buy or greenfield are just three names for the same underlying decision — and it is almost never settled where you would expect: on cost.
How a defensible number comes about
In an assessment for the move to a new ERP core system, we put the decision at roughly 18 months and three million. The number was not a gut feeling. Together with the developers and functional consultants we went through the business blueprint, module by module, process by process, and estimated for each building block the effort all the way into production — build, test, sign-off, deployment. In the end came a recommendation for greenfield: the old system was heavily modified, and the data had to be cleaned up for the move anyway.
The factor beyond cost
What tipped it in the end was the legacy. Brownfield — carrying the existing system forward and evolving it — is entirely unproblematic as long as the degree of modification is low, the master and transactional data are consistent, and the custom developments stay manageable. Then the focus is on the product, because the history has a narrow footprint. Here it was the other way round: unclean master data, extensive modifications in the old release. That shifts the focus from the new release back onto the legacy. And then it becomes an effort question after all: if I have to look into everything anyway — straightening data, resolving or reconciling modifications — is that more expensive or cheaper in a fresh system?
Often the industry decides too. Where operating in-house is mandatory, the cloud drops out as an option before the discussion even starts. Such constraints are not details. They narrow the solution space long before the economics are calculated.
The most common fallacy at board level is the search for the royal road. There is none. The decision hangs on the analysis, and its quality on the skill of the analysts. Beforehand you can hardly say seriously what makes more sense — only on a solid basis does the question become decidable at all. Greenfield or brownfield is really not an architecture question. It is a bet on time, money and your own ability to steer. So do not ask first about the target picture. Ask: how well do we know our own legacy?
