Every engineering team eventually proposes building something it could buy — a deployment tool, a monitoring layer, an internal wiki, a ticketing system, a feature flag service. The arguments are consistent and often sincere: the commercial option is expensive, it does not fit our workflow, and we could build a simpler version in two weeks. The two weeks is the part that is almost never true.
The reason is that the visible build is a fraction of the commitment. What follows is authentication, permissions, an audit trail, backup, the edge cases discovered in use, the person who maintains it, and the documentation for the next person. And because it is internal, none of that work will ever be prioritised against customer-facing features — so it gets done badly or not at all, and the tool degrades into something the team tolerates rather than uses.
The question that resolves most of these decisions is whether this thing is your business. Software that differentiates you, that customers pay for, that encodes something you know and competitors do not, should be built. Software that runs a process every company runs should be bought unless there is a specific, nameable reason. The internal tool that runs your unusual delivery workflow may genuinely be worth building. Your wiki is not.
Where you do build, be honest about the ongoing cost and assign it. A named owner, a small standing allocation of time, and a decision about what happens when they leave. Internal tools without an owner become systems everyone depends on and nobody understands, and the discovery usually happens during an incident when the person who wrote it is unreachable.
One caution about the cost comparison, which is usually done wrong. The commercial product's price is visible and the internal tool's is not, so the comparison flatters building. Cost the build at loaded engineering rates, include a realistic maintenance percentage annually, and compare over three years. Teams that do this arithmetic honestly buy considerably more often than teams that do it by instinct — and the ones that still choose to build have a case they can defend.
