Custom is not automatically better, and SaaS is not automatically cheaper. The decision comes down to how unusual your process really is, and how much of it you are willing to change.
Every business believes its process is unique. Most are not, and that is good news, because off-the-shelf software is cheaper, available today, and maintained by somebody else. But some businesses genuinely are unusual, and for them, forcing the business to match the software is more expensive than building software that matches the business. Here is how to tell which one you are. ## Buy off-the-shelf when **Your process is ordinary and you are willing to adapt.** If you keep standard records and run a standard workflow, the existing product is better than anything built for you in a month, because thousands of businesses have already found its bugs. **You need it now.** Subscribing takes an afternoon. Building takes weeks at minimum. **The budget is small and uncertain.** A monthly fee you can stop beats a build you have to finish. **Nobody will maintain it.** Custom software needs someone when it breaks. If you have no plan for that, buy. ## Build custom when **The process is genuinely yours, and it is why you are competitive.** If the way you do the work is the advantage, software that forces you into someone else's way is expensive in a way that never shows on an invoice. **Nothing on the market fits without three workarounds.** Watch for the tell: spreadsheets alongside the software, to hold what the software cannot. Those spreadsheets are the specification for what you actually needed. **You are paying per user for a fraction of the features.** Fifty staff on a per-seat plan, using four features out of forty, is a steadily growing bill for software that still does not fit. **Integration is the whole job.** Sometimes the real need is to connect three things you already use. That is a custom build, and usually a small one. **Your data cannot leave, or must live somewhere specific.** Regulatory or contractual constraints sometimes settle the question before anything else. ## The middle road most people miss You rarely have to choose for the whole business. Use standard tools for standard things, accounting, email, payroll, and build only the part that is actually yours. A restaurant does not need custom accounting. It might need its own ordering flow. A school does not need a custom email system. It might need its own results and fees logic, because that is where its rules live. Build the core. Buy the edges. ## The costs people forget on both sides **Off-the-shelf:** per-user pricing as you grow, paying for a tier to unlock one feature, data you cannot easily export, and the price rising on someone else's schedule. **Custom:** hosting, maintenance, the fixes that appear with real users, and dependence on whoever built it. Mitigate that last one by owning the code and keeping it in a language more than one person can read. ## The question that usually settles it What does the business do today that the software would have to stop you doing? If the answer is nothing important, buy. If the answer is the thing that makes the business work, build that part, and buy the rest.