Off-the-shelf or custom software: a decision you can actually make
Buy versus build is not a philosophy question. It comes down to four practical checks. Here is the framework we use with clients before writing a line of code.
Founder·30 July 2026·7 min read
Buy or build is one of the first real decisions a growing business faces, and it is usually argued as a matter of taste. It should not be. It is a set of specific questions with specific answers.
Start with the process, not the product, and run it through four checks.
1. Is this process your advantage, or a commodity?
Payroll, email and accounting are commodities. Everyone does them roughly the same way, and a packaged product will almost always beat a custom one on price and polish. Buy those.
The process that makes you different, the one your competitors cannot copy from a manual, is the one worth building around. That is where custom software earns its cost.
2. How much does the tool bend, and how much do you?
Every off-the-shelf product assumes a way of working. If it fits yours, wonderful. If it does not, you will spend years bending your business to fit the software, and paying in small frustrations every day.
Custom software bends the other way. The question is simply how much your process differs from the standard one, and how much that difference matters.
3. What does leaving cost?
A tool you cannot export from is a tool you cannot leave. Before committing, find out what it takes to get your data out and move on. Lock-in is a cost even when you never intend to leave, because it removes your leverage in every future conversation about price.
4. What is the honest three-year cost?
Per-seat subscriptions look cheap on the first invoice and grow with every hire. A custom build costs more upfront and then belongs to you. Compare them over three years, with your real headcount, not month one.
The honest answer is often both
Most businesses should buy the commodities, build the one or two things that are genuinely theirs, and connect the two. The mistake is not choosing wrong. It is choosing everything from one column because it felt like a rule.