Buy, rent, or build? Choosing software for your business

Sooner or later every business needs software to do something — manage stock, handle bookings, track customers, run accounts. And every time, there is a choice underneath that owners rarely make consciously: do you buy a ready-made product, rent a service that runs it for you, or build something custom for your specific needs? Most businesses stumble into one of these by accident — usually whichever a salesperson or a developer happened to push — and live with the consequences for years.

It is worth making the choice deliberately, because it is one of the most expensive recurring decisions a business makes, and the three options have very different costs, risks, and lock-in. Here is a framework for choosing well, in plain terms.

The three options, honestly described

Rent (subscribe to a ready-made service). You pay a monthly or yearly fee for software someone else built and runs — the booking tool, the accounting service, the point-of-sale app. You get it instantly, someone else maintains it, and you carry no technical burden. In exchange, you fit your business to how the tool works, you pay forever, and you do not control its future. This is renting a finished, furnished space.

Buy (a one-time product you then run). Less common now, but real: software you pay for once and operate yourself. Lower long-term cost in theory, but you take on running and maintaining it, and "buy once" software increasingly comes with hidden ongoing costs anyway. This is buying a building you must then maintain.

Build (commission something custom). You pay developers to make software shaped exactly to your business. You get something that fits perfectly and that you own, with no monthly fee per user to a vendor. In exchange, it is expensive up front, it takes time, and you are now responsible for maintaining it forever — because custom software, like a custom building, needs upkeep nobody else will do. This is building your own premises from scratch.

The default should be rent

Let me state my bias plainly, because it will be right for most readers most of the time: default to renting a ready-made service, and only deviate when you have a clear reason.

The reason is that building is far more expensive and risky than it looks, and the maintenance burden is permanent and underestimated. A ready-made tool has been built and debugged by a company serving thousands of businesses, who spread that cost across all of them. You could never justify that investment for yourself alone. For the common needs — accounting, scheduling, basic point-of-sale, customer records — renting a proven tool is almost always the right call. The tool already exists, it works, and someone else carries it. Building your own version of a solved problem is a way to spend a great deal to end up slightly worse off.

So the question is not "should I build?" The question is "is there a specific reason renting will not work here?" — and only a yes moves you off the default.

When building is actually justified

Building custom software earns its cost in a narrow set of situations. Reach for it when:

• Your need is genuinely unusual. No ready-made tool fits because the way your business works is specific and unlike others. Note: most businesses believe they are unusual and are not. Be sceptical of your own specialness here. But sometimes it is true — the Smart School Management System we build exists because the off-the-shelf tools were designed for a context that did not match African schools at all.
• The thing is your core competitive advantage. If the software is your edge — the thing customers come to you for — you may need to own and control it rather than rent it from someone who also rents it to your competitors. You do not outsource the heart of your business.
• The long-term rental cost is genuinely larger than building and maintaining. At sufficient scale, per-user subscription fees can exceed the cost of owning something. This is rarer than vendors fear and founders hope, so do the actual maths before believing it.

If none of these is clearly true, build is probably the wrong answer, however appealing "our own custom system" sounds.

The questions that decide it

Faced with a real decision, work through these:

1. Does a proven ready-made tool already do most of what I need? If yes, the burden of proof is heavily on not renting it. Adapt your process to a good tool before you commission a custom one.
2. Is this need core to my competitive advantage, or is it just plumbing? Plumbing — the necessary-but-generic functions every business has — should almost always be rented. Save building for your edge.
3. Can I actually carry the maintenance? Building means owning the upkeep forever. If you do not have, and will not have, the capacity to maintain custom software, renting (where maintenance is someone else's problem) is the honest choice regardless of other factors.
4. What does each option cost over three years, not just at the start? Compare the full picture: rental fees over three years versus build cost plus three years of maintenance. The cheap-looking option at month one is often not the cheap option at year three, in either direction.

The takeaway

Buy, rent, or build is a real decision with real money attached, and it deserves to be made on purpose rather than stumbled into. Default to renting proven tools for everything that is common and solved — which is most things — because someone else has already paid to build and maintain them better than you could alone. Build only where your need is genuinely unusual, central to your advantage, or where the long-run economics truly favour ownership, and even then, build only that piece and rent the rest.

Get this right and your software budget goes to the few things that actually differentiate your business, while the rest quietly runs on tools you never have to think about. Get it wrong — build what you should have rented, or rent what you should have owned — and you pay for it monthly, for years. It is worth the hour of clear thinking up front.