Off-the-shelf or bespoke? A straight test for SMEs

The wrong question

“Should we buy something or have something built?” cannot be answered for a whole business, because a business is not one process. Payroll and your quoting method have nothing in common except that they both happen on a computer.

Ask it one process at a time and the answer usually falls out in a couple of minutes.

Where off-the-shelf wins

Buy when the process is a solved problem that works the same way in every business, and especially when somebody else is obliged to keep it current.

• Anything regulated or filed. Accounts, payroll, VAT, tax. Someone else tracking every rule change is worth paying for, and building it yourself buys you a maintenance obligation forever.
• Commodity infrastructure. Email, file storage, video calls, password management. There is no competitive advantage available here.
• Genuinely standard workflows. If a product fits your process without contortions, use it. Fitting is the test, not features.

The failure mode of buying is not the licence fee. It is buying a large product for one function and then reshaping the business around software that was designed for someone else’s.

Where bespoke wins

Build when the process is the business, or when the shape of it is genuinely unusual.

• The thing you do differently. The pricing method, the scheduling logic, the way you handle a job. If a competitor’s version of this process would look nothing like yours, no product can fit it well.
• Trades the software industry skipped. Some sectors have never had a product built for how they actually work, so they run on generic tools plus a lot of human glue. Steel stockholding was one, which is why we ended up building MetlSys.
• The joins between systems. The most valuable bespoke work is often small: the automation that stops an order being typed in three times.
• When the workaround has become the job. If people spend their day moving data between screens, that work is the specification.

The five-minute test

Take one process and answer four questions.

1. Would a competitor’s version of this look basically the same? Yes leans buy. No leans build.
2. Is anyone re-keying data that already exists somewhere else? That is automation work regardless of which way the rest goes.
3. Does it break the law or the accounts if we get it wrong? Leans buy, for the maintenance obligation alone.
4. Is there a spreadsheet propping up the software we already pay for? That is a bad fit, already costing you money.

Three answers pointing the same way is a decision. A split usually means the process needs separating into the standard part and the part that is yours.

The answer is usually both

The pattern that works for most SMEs is to buy the commodity, build the differentiator, and connect them. Keep the accounts package. Build the order and job system that fits how you actually work. Automate the handover so the invoice raises itself from the despatch rather than from somebody re-typing it.

This is usually the cheapest option as well as the best-fitting one, because the expensive part of a bespoke build is the ground everybody else has already covered.

What a bad fit costs

The cost of software that does not fit never appears on an invoice, which is why it survives for years. It shows up as the shadow spreadsheet, the export-and-fix routine every Monday, the report assembled by hand because the built-in one is not quite right, and the new starter who takes three months to learn the workarounds.

Put a number on it with the arithmetic in what bespoke software actually costs before you renew anything. Some of the strongest cases for building start as a licence renewal nobody had questioned in five years.