Almost every business we speak to arrives with the same question phrased two different ways. Either “which system should we buy?” or “can you build us something?” Both skip the step that actually decides the answer.
That step is working out which parts of what you do are genuinely distinctive, and which parts are the same as everybody else’s.
Buy the ordinary parts
Payroll is payroll. Accounting is accounting. Email is email. These are solved problems with mature products, competitive pricing and someone else carrying the maintenance burden. Building your own is almost always a mistake, and the rare exception involves a regulatory requirement no vendor has met.
The test is simple: if a competitor could use the identical tool with no disadvantage, buy it.
Configure where the shape is close
Between buying and building sits a large middle ground. A good platform with thoughtful configuration will cover a surprising amount of unusual process. Configuration is cheap, reversible and does not need a developer every time something changes.
The warning sign is when configuration starts to feel like programming with the wrong tools: fourteen automation rules referring to each other, custom fields named after a workaround, a spreadsheet that exists to fix what the platform gets wrong. That is the point where configuration has become more fragile than code.
Build the part that is actually yours
Every business has a handful of processes that are genuinely its own. The way you price. The way you schedule. The rules that decide what happens when a customer asks for something unusual. That is where custom software earns its cost, because it is the part no vendor will ever build for you.
Building there is also lower risk than it sounds, because the scope is narrow. You are not replacing your accounting system. You are building the one thing that sits in the middle and makes the rest work together.
A practical order of operations
- Write down the process as it actually runs today, including the exceptions people handle by memory.
- Mark each step as ordinary, close-to-standard, or genuinely yours.
- Buy the ordinary. Configure the close. Only then ask what is left.
- Cost the remainder honestly, including the five years of maintenance that follow.
Done in that order, the build is usually smaller and more obviously worth it than anyone expected. Done backwards, you end up paying to rebuild things that were already solved.