Off-the-shelf or bespoke: how to tell which you need

Every business that outgrows its software goes through the same argument. Someone says the current system is holding everyone back. Someone else points out that a new one costs money the business would rather spend elsewhere. Both are right, and neither has the number that settles it.

Here is how to work that number out, and how to tell which side of the line you are on.

The cost you can see, and the one you cannot

Off-the-shelf software has an obvious price: the monthly licence, times the number of people who need a seat. It is on an invoice, so it gets discussed.

The second cost is the work your team does to keep the software happy. Nobody invoices for it, so it rarely comes up. It looks like this:

• Exporting from one system so it can be pasted into another
• A spreadsheet that exists only to hold what the software will not
• Someone re-keying the same order into two places
• A weekly report assembled by hand because the built-in one is close but wrong
• Steps in your process that exist because the software insists on them

That last one is the expensive one. A tool you cannot change slowly reshapes how you work, and after a couple of years nobody remembers which parts of the process are your business and which parts are the software's opinion.

Put a number on it

You do not need an audit. Pick the two or three people who touch the system most and ask them one question: how much of your week goes on getting data into or out of this thing?

Then do the arithmetic:

hours per week  ×  52  ×  loaded hourly cost  =  annual cost of the workaround

Loaded cost means salary plus employer's National Insurance, pension and overhead — roughly 1.3× gross pay for most UK employers. On a £35,000 salary that is £45,500 a year, and about £26 an hour once statutory holiday is taken out of the working year. Six hours a week is 312 hours, so a shade over £8,100. Two people doing it is £16,200. Add the licences on top.

Now compare that against building the thing once. If the workaround costs more per year than the build costs in total, the argument is over — and it has been over for a while.

Four questions that decide it

Cost is not the only test. These four settle most cases faster than a spreadsheet does.

Is the process the thing you are good at? If the way you do this particular job is why clients choose you, software that forces you into someone else's version of it is working against you. Bespoke is right. If it is payroll, or accounting, or email, buy it — those are solved problems and you will not do better.

How many systems have to agree? One tool doing one job is usually fine off the shelf. The pain starts at the joins. If your quoting system, your job tracker and your invoicing do not talk, someone is being the integration, by hand, every day.

Whose data is it? Ask what happens if you leave. If the answer is a CSV export and good luck, and the data is central to your business, that is a real risk with a real cost attached.

Will this still fit in three years? Not five — three. If the honest answer is "only if we do not grow", you are buying the problem again later, plus a migration.

When off-the-shelf is the right answer

It usually is, and any developer who tells you otherwise is selling.

Buy it when the job is common, the tool is mature, and your version of the process is not a competitive advantage. Accounts, payroll, email, calendars, storage, video calls. You will never out-build Xero at being Xero, and you should not try.

Buy it when you need it next week. A bespoke build has a lead time. If the business is on fire, put it out first.

Buy it when nobody can yet describe what they need. Software written from a vague brief is expensive and wrong. Use something adequate, watch where it hurts for six months, then build from what you learned. The waiting is not wasted — it is the specification.

What "bespoke" should actually mean

Custom software has a reputation for going badly, and it earns it whenever any of these are missing:

• You own the source code. Written into the contract, not implied. If the relationship ends, the software does not.
• Your data stays yours, in a standard format, exportable without asking anyone.
• A fixed quote for a defined first phase. Open-ended day rates are how a £15,000 project becomes a £60,000 one.
• Something working early. Not a document describing what will work later — the actual thing, in front of you, within weeks. If you cannot see it running, you cannot tell whether it is going wrong.
• A plan for the boring parts. Backups, someone to call, and what happens when the person who built it is unavailable.

That last one matters more than it sounds. Plenty of bespoke systems are perfectly good software that became a liability because nobody could maintain them.

A reasonable way to start

You do not have to decide the whole thing at once, and you should not.

Take the single worst workaround — the one everybody complains about — and cost it out using the arithmetic above. Then get a fixed quote to remove just that one. Small enough to say no to, big enough to prove whether the approach works with your business and with whoever is building it.

If it pays for itself, you have your answer and a working relationship. If it does not, you have spent a fraction of what a full rebuild would have cost finding out.

That is the honest test, and it works whichever way it comes out.

---

We build bespoke software for UK businesses — internal tools, client portals and the integrations between systems that will not talk to each other. Fixed quotes, source code yours.

If you want a second opinion on whether your problem is worth building for, tell us what is broken and we will tell you straight — including if the answer is "keep what you have".