When custom software is worth it — and when it isn't

We turn down custom builds fairly regularly. Not out of modesty — because a business that could have configured an existing tool in a fraction of the time does not thank you for a six-month project, however well built.

The 80% test

If an off-the-shelf product covers around 80% of the workflow and the remaining 20% is preference rather than differentiation, buy it. Adapt the process to the tool and spend the saved budget on the part of the business that customers actually see.

Custom becomes the right answer when that 20% is the reason you win work, when licence costs scale faster than revenue, or when the real job has quietly become gluing four tools together.

The hidden costs on both sides

Buying has costs people forget: per-seat pricing that punishes growth, data you cannot easily get back out, and roadmap decisions made by someone else. Building has the obvious ones — cost, time, maintenance — plus the one that surprises people, which is that software needs an owner inside the business after launch.

A reasonable middle path

• Buy the commodity parts — accounting, email, payroll, storage.
• Build the part that is genuinely yours — the workflow customers experience.
• Integrate rather than replicate; the best custom systems are usually small.
• Start with one workflow in production, not a full system on paper.

The question is rarely build or buy in the abstract. It is which specific part is worth owning.