> From what I've seen / heard from others, it's generally better to adapt processes to the off-the-shelf tool
I have never seen this work even once. I've actually built entire businesses on the concept that people are impossible to change, but software is easy to change.
Customizing ERPs is where consulting firms make money but the ERP vendors themselves advise against this because it becomes expensive maintaining the customizations as new versions of the software and more features are released.
Did you use much off the shelf software? I mean, I'm guessing you used databases and frameworks, but were the business process and UI components off the shelf or custom?
I use as much off-the-shelf as possible. But there is real work involved in adapting software to a company's culture.
If you do it right, 99% of the work has already been done for you by an existing SaaS or FOSS project, and "all" you have to do is customize it. And generally this doesn't mean "change the config" but rather assemble different building blocks into the right shape.
This can still be a multi-year project, even when you're writing almost no code. It takes a long time to understand an existing company's culture, processes, and needs.
Makes sense, this is just my personal experience so it's always interesting to hear what others have seen. I recently learned about Mechanical Orchard[1], which seems to have the same thesis (better to update legacy custom software with a modern custom solution rather than migrate to a modern off-the-shelf solution).
I have never seen this work even once. I've actually built entire businesses on the concept that people are impossible to change, but software is easy to change.