Do I Need Custom Software, or Is Off-the-Shelf Fine?

I build custom software. The honest answer to this question is usually no, and I would rather say so here than waste your time on a call.

Buy the ready-made thing when

Your problem is a normal problem. Accounting, payroll, email, storage, card payments. Thousands of businesses need these the same way, and companies have spent years being told what is wrong with theirs. You will not beat that, and trying is expensive.

It is regulated or it touches money in a way that has rules. Somebody else keeping up with those rules is worth paying for.

You are small and still changing. If how you work is likely to be different in six months, building software around today's version of it is building the wrong thing carefully.

The gap is annoying rather than expensive. Every tool has friction. Friction alone is not a reason to build. It has to be costing you something you can name.

Build when

The workaround has a name. When there is a spreadsheet everybody knows about, or a step people do in a specific order because the software will not, that is the tell. Not that something is imperfect, but that you have already built a system by hand and are maintaining it with human memory.

You are paying for several tools that do not talk. The cost is not the subscriptions. It is somebody retyping the same information into the second tool and getting it slightly wrong sometimes.

How you work is actually the advantage. Some businesses win because of a specific way they do things. Software that forces you into the generic version of that is quietly taking the thing you are good at.

The per-seat price punishes growth. When hiring someone means a bigger software bill, and it is a real number, the maths changes.

You have been told no. You asked the vendor for the thing you needed and were told it is on the roadmap. It has been on the roadmap for a year.

The test I would use

Not a feature list. This:

What do you or your staff do every week that a computer should be doing, and how long does it take?

Put an honest number of hours on it. That number, times what an hour is worth to you, times fifty two, is what the gap costs you a year. Compare it to what a build costs.

If that comparison is close, do nothing. It is only obvious when it is obvious.

The answer nobody offers

Both, and it is the most common right answer.

Keep the accounting package. Keep the payment provider. Build the one piece in the middle that is genuinely yours, and have it talk to the rest. You are not replacing your whole stack, you are filling the hole in it that your staff currently fill by hand.

That is most of what we build, and it is why the build is broken into pieces with a price on each one rather than sold as one lump. Take the two that matter and leave the rest.

If you are not sure

Then it is probably not time. Genuinely.

The businesses that should build usually know exactly which weekly job is costing them, and they can describe it in a sentence without being prompted. If you cannot, wait until you can. The problem will get clearer and the build will be better for it.