Build or Buy? A Practical Test for Business Software

We build custom software, so it would be convenient to argue that everyone needs it. Most businesses do not. Off-the-shelf products are extraordinarily good value when they fit, and the fastest way to waste money is to commission a bespoke version of something you could have subscribed to on Monday.

Start by asking whether the process is an advantage

The core question is whether the process in question is a source of competitive advantage or simply a cost of doing business. Payroll, accounting and email are cost centres — everyone does them roughly the same way, and off-the-shelf products are mature and cheap. But the specific way you route a job through your workshop, or price a quote, or manage a client relationship may be exactly why customers choose you. Forcing that into a generic tool erodes the thing that makes you different.

Four signals that buying is right

• A mature product exists that covers 80% or more of what you need without heavy configuration.
• The process is standard across your industry and unlikely to change much.
• You need it working in weeks, not months.
• Your team is small enough that per-seat licensing stays comfortably cheaper than a build.

Four signals that building is right

• You are paying for many seats of a product where you use a fraction of the functionality.
• Critical work still happens in spreadsheets that get emailed around, because no product matches your process.
• You are running three systems that must agree on the same number, and they never do.
• Your process is genuinely unusual, and the workarounds needed to fit it into a product are themselves a source of errors.

The option most people miss

It is rarely all-or-nothing. The strongest arrangement is usually buying the commodity layer — accounting, payroll, email, CRM — and building only the thin custom layer that encodes your actual advantage, integrated into the rest. You get mature software for the boring parts and bespoke fit where it matters, at a fraction of the cost of building everything.

Run the honest sum

Compare five years, not one. On the buy side: subscription costs at your expected headcount, implementation and configuration, integration work, and the cost of the workarounds you will live with. On the build side: development, hosting, and roughly 15–20% of build cost annually for maintenance. Then add the hardest number to estimate and the one that usually decides it — the hours your team currently spends on manual work that the right system removes.