Custom software or off-the-shelf? How to choose without regretting it in two years

The question usually comes late: someone on the team says “we can’t keep working like this”, and you are looking at two options that do not compare. One is ready tomorrow but makes you work its way. The other takes months but moulds itself to how you already work. The real difference does not show in year one — it shows in year three.

The short rule: buy what makes you the same, build what makes you different

Accounting, invoicing, electronic signatures, email — you gain nothing by building these yourself. They are identical across the economy, regulated from the outside, and a supplier doing them for ten thousand companies will do them better than you.

On the other hand: how you take an order, how you push it into production, how you decide which crew goes on site, what you check on delivery — those are yours. They are exactly why clients pick you. Force them into software designed for someone else and you will end up adjusting your company to the software, not the other way round.

Four signs off-the-shelf is the right call

• Your process has nothing unusual about it compared to the rest of the market.
• You need a result in weeks, not months.
• You need something working now, and your process can wait for a better fit later.
• The number of users is small and stays small — subscriptions get expensive at scale.

Four signs it is worth building

• People keep a “parallel” spreadsheet alongside the official system. That is the classic sign the software does not match reality.
• The same data gets entered two or three times, in different systems.
• You have a market advantage that comes from how you work, and standard software flattens it.
• You keep paying for modules you never use, but cannot do without the rest.

The calculation few people do

Judge over three years, not one. Off-the-shelf asks for little effort at the start and steady effort afterwards: configuration, extra modules as you grow, adjustments at every process change. Custom software asks for the big effort upfront, but by year three only maintenance remains.

Weigh in something that appears in no presentation: lost time. If five people each lose 20 minutes a day to duplicate entry, that is close to 370 hours a year — nearly a quarter of one person's working year, spent on work that produces nothing.

The third option, the most overlooked one

Often the right answer is not either-or. You keep the accounting software you have, and build the missing layer on top of it — an application that takes orders, assigns them and passes on only what needs invoicing. The two talk to each other through an integration, and you rewrite nothing that already works.

It is the approach we recommend most often, because it preserves the investment you already made and fixes exactly the part that hurts.

Three questions before signing anything

1. What happens to my data if I part ways with the supplier? A vague answer is a real risk.
2. Who pays for changes when the law or my process changes?
3. Can I integrate it with what I already have? Ask to see the API documentation, not just a promise.

If you got this far and still are not sure, that is normal — the decision hangs on details that do not fit in an article. Write to us with your actual situation and we will tell you honestly whether something is worth building, or whether you only need an integration between what you already have.

← Back to the blog