We have built custom systems for bakeries, multi-branch retailers and security companies, and the first question is always the same: "what will it cost?" The honest answer is a range — but ranges are only useful when someone explains what pushes a project from one end to the other. That is what this guide does.

The realistic ranges

What actually moves the price

Integrations move it most. "It must talk to our accounting system / our tills / our cameras" can be a third of a project on its own — that is not padding, it is where the risk lives. User count and roles come second: software for three trusted staff is simpler than software for forty users with different permissions. Polish comes third: an internal tool your team uses daily can be plain; anything your customers touch cannot.

What does not move the price the way people expect: the number of screens. A twelve-screen system with no integrations is often cheaper than a three-screen system that must sync with QuickBooks Desktop in real time. (We know — we built exactly that.)

When you should NOT build custom

The costs after go-live

Budget for hosting (from a few hundred rand a month — or your own on-site server, which we often deploy for clients who want their data at home), and for support. A sensible support arrangement for an SME system is a small monthly retainer or an hourly rate against a response-time promise. Ask any developer two questions before you sign: "who owns the source code?" (you should) and "what happens if you disappear?" (there should be a real answer — code in your repo, documentation, a handover path).

If you are weighing a build, we will give you a straight answer about which tier you are in — including "don't build this" when that is the truth. Get in touch, or see what we have already built.