Custom Software vs Off-the-Shelf: A Decision Framework for Growing Businesses

Every growing business reaches the point where the spreadsheet, the subscription tool, or the collection of apps stitched together with exports and re-typing starts to strain. The question that follows — should we build something of our own? — is asked far too early by some and far too late by others. Here is how we think it through with clients.

Off-the-shelf is the right default

For anything that is not specific to how your business competes — accounting, email, document storage, generic project tracking — a well-established product will be cheaper, more secure and better supported than anything built for you. It benefits from thousands of other customers finding its bugs. Start here and stay here as long as it fits.

The signs a ready-made tool is costing you

• Your team keeps a parallel spreadsheet because the tool cannot represent how you actually work.
• Data is re-typed between two systems that do not talk to each other, and the re-typing is where errors come from.
• You pay for seats and modules you do not use because the plan that fits one need drags the rest along.
• The process you are known for — the thing customers choose you for — has to be bent to fit the software rather than the other way round.
• Reporting means someone exporting three CSVs every Monday and building the same sheet by hand.

One of these is an annoyance. Three or more, growing with every new hire, is the case for building — or, more often, for building the connective piece that lets the existing tools work together.

Often the answer is integration, not replacement

The most valuable custom software we build is frequently small: a service that takes leads from the website, enriches them, and creates them in the CRM with the right owner; a dashboard that reads from three systems so nobody exports anything; an automation that moves a job through its stages and notifies the right person. These keep the tools people already know and remove the manual work between them.

When to build the whole thing

Build when the workflow is central to how you make money, when no product models it without contortion, and when you expect it to change as the business grows. A clinic's patient journey, a hotel's front desk and kitchen, a lender's onboarding — these are places where owning the software is owning the process.

What to insist on if you commission custom software

1. Discovery before a quote: a written scope, the integrations it must speak to, and the acceptance criteria for 'done'.
2. You own the code and the accounts it runs on. No lock-in through a vendor's hosting or a licence you cannot leave.
3. Reviewable stages with working software at each one, not a big reveal at the end.
4. Security treated as an acceptance criterion: access controls, encrypted data, audited dependencies, no secrets in the code.
5. A handover that includes documentation, a maintenance agreement, and a realistic answer to 'what happens when you are not around'.

The question is not 'can it be built' — almost anything can. It is 'which parts of how we work are worth owning'.