Most businesses should buy. Off-the-shelf software is cheaper, supported by someone else, and improved without you paying for it. The interesting question is how to recognise the minority of cases where that is the wrong call.
Buy when the process is ordinary
Accounting, payroll, email, CRM, HR. These are solved problems where your requirements are not unusual, however unusual they feel. Buying means someone else handles compliance changes, security patches and the features you will want in two years.
Adjusting your process to match good software is usually cheaper than the reverse.
Build when the process is the advantage
If a workflow is genuinely how you compete — how you schedule, price, dispatch or route — then forcing it into a generic tool sands off the thing that makes you better. That is worth building.
Be honest here. "We do it differently" is often habit rather than advantage. The test: would a customer notice if you did it the standard way?
The integration tax
Watch for the case where the licence is cheap and the connection to everything else is not. If the shortlisted product needs six months of middleware to talk to what you already run, the total cost is not what is on the quote.
The customisation trap
Heavily customised off-the-shelf software is the worst of both: you pay licence fees, you pay for development, and upgrades become a project each time because your changes have to be reapplied. If you find yourself customising most of a product, you are building — just expensively, on someone else's foundations.
The usual right answer
A mix. Standard products for standard functions, connected by APIs, with a small custom layer over the part that is actually yours. Less to maintain than building everything, better fitting than buying everything.
How we approach it
We build custom systems — and we will say when a product you can buy this afternoon is the better answer. More in custom CMS vs WordPress and what BK IT does.
Tell us the process and we will tell you which side it falls on.