ERP Off-the-Shelf or Custom Development? The Pragmatic Choice for Small Businesses
When a small company asks for an ERP on a short timeline, the question is not only technical: it is strategic. In this article I explain why starting from an existing solution is often the most sustainable choice.
In freelance groups, it is common to see requests like this: someone needs an ERP for a small company, may already have started generating code with AI, and is looking for help to define the project, get it online, and test it. On the surface, it looks like a standard development job. In reality, before the code, the real question is usually different: does it make sense to build everything from scratch?
When a client expects an ERP in one month, fully working and without problems, the first risk is unrealistic expectations. An ERP is not just another web app with a few extra screens: it requires defining processes, roles, data, permissions, workflows, testing, maintenance, and handling unexpected issues. The smaller the business, the more important it becomes to avoid choices that introduce unnecessary complexity.
That is why, in most cases, the right question is not “can we build it?”, but “should we really build it?”. If there is already a licensed solution that covers most of the needs, it is often much more sensible to start there and customize it. This reduces time, cost, technical risk, and above all operational failure risk. For a small business, the value does not lie in having a unique piece of software in the world, but in having a reliable tool that actually works in the real environment.
Custom development makes sense when the process is truly specific, when the competitive advantage comes from that workflow, or when existing solutions cannot cover the key requirements. In all other cases, starting from an existing base is often the most rational choice. This does not mean giving up quality: it means focusing effort where it really matters, instead of spending months rebuilding features that are already mature in other products.
There is another aspect that is often underestimated: operations. Putting an ERP online does not just mean “deploying the project on a server.” It means managing hosting, databases, backups, tests, permissions, monitoring, and ongoing maintenance. If someone starts from AI-generated code without solid direction, the result can easily become a system that is hard to maintain and fragile over time. AI can accelerate work, but it cannot replace design.
In my approach, I usually start from a very simple evaluation: what is the real problem, how much time is available, what budget makes sense, and how much complexity is the client truly ready to manage? From there, it becomes clear whether a customizable ERP is needed, a module integration is enough, or a broader project is justified. Technology comes after that, not before.
That is why, when I see requests like this, I tend to answer directly: first we should evaluate whether it is better to buy an existing solution or adapt one that is already available. It is a more honest, more sustainable, and often more economical starting point. And above all, it avoids confusing the desire to build with the client’s real need.
In the end, the best software is not the one that sounds most impressive. It is the one a business can actually use, maintain, and grow over time.