ERP Off-the-Shelf or Custom Development? The Pragmatic Choice for Small Businesses
Someone asks for an ERP in a month, with AI-generated code already ready. Before saying yes, let's do the math. And the answer usually isn't the one they want to hear.
Freelance groups always have the same request: "I need an ERP for my small business. I already had AI generate the code. Looking for someone to put it online and test it."
Sounds like a developer job. Normal.
But the real question comes before the code: does it make sense to build it from scratch?
An ERP isn't a web app with two extra screens
"An ERP in a month, working, no problems." Well, the first problem is the expectations.
An ERP is processes, roles, data, permissions, tests, maintenance, surprises. And the smaller the company, the more complexity you need to remove. Not add.
Not "can we build it?" but "should we?"
If licensed software already covers most of the needs, starting there and customizing is almost always the right call. Less time, less money, less risk of the project dying after six months.
A small business doesn't need one-of-a-kind software. It needs software that works.
Custom makes sense when:
- the process is genuinely unusual;
- the competitive edge comes from exactly that;
- what exists doesn't cover the essentials.
Otherwise, rebuilding features that already exist, mature and tested, is throwing months away.
AI speeds things up. It doesn't design
Putting an ERP online isn't "uploading it to a server". It's hosting, database, backups, permissions, monitoring, maintenance. Every day.
Starting from AI-generated code with no direction? You end up with a fragile system nobody can maintain. And the client calls you every time it breaks.
Problem first
I always start from four dumb questions:
- what's the real problem?
- how much time is there?
- what budget makes sense?
- how much complexity can the client actually handle?
That tells you whether you need software to adapt, a module to integrate or a big project. Technology comes after.
So to requests like this I answer straight: first let's see if buying something ready makes more sense. It's more honest, and often cheaper.
The best software isn't the one that looks impressive. It's the one the company can actually use and maintain. The rest is marketing.
Have you ever seen an ERP "built in a month" survive its second year?