Buy or build: when custom software beats off-the-shelf

Most businesses should buy their software, not build it. That may seem a strange thing for a company that builds custom platforms to say, but it is true. Off-the-shelf products are cheaper to start, faster to roll out, and someone else carries the cost of keeping them current. For accounting, email, payroll, and most of what a business runs on, buying is the right call and nobody should talk you out of it.

The question is not whether custom software is better. It is whether your operation has reached the point where the off-the-shelf option costs you more than it saves. That point is real, and it is usually easy to recognise once you know what to look for.

Start with the honest default

When we run a discovery phase for a prospective platform, about one in five ends with us recommending that the client does not build at all. Sometimes an existing product fits once it is configured properly. Sometimes the problem is a process issue that software will not fix. We would rather lose the quote than build the wrong thing, and any partner worth hiring should be willing to say the same.

So begin with the assumption that you should buy. Then test that assumption against the signals below.

Four signals that building wins

1. The workflow does not fit, and your team is working around it. Every off-the-shelf product encodes someone else's idea of how the work should flow. When that matches your operation, it is a gift. When it does not, your people end up keeping side spreadsheets, re-keying data between systems, and following unofficial workarounds that only two staff members understand. That hidden labour is a cost, even though it never appears on an invoice.

2. The subscription bill keeps climbing. SaaS pricing is built to grow with you: per user, per record, per module. At small scale it is a bargain. At larger scale, several tools stacked together can cost more each year than a platform built to your specification, and you still do not own any of it.

3. The integrations you need do not exist. Your operation may depend on systems that no product vendor has any reason to support: legacy databases, specialist hardware, industry-specific equipment. When the integration is the hard part, a generic product rarely solves it.

4. You need to own the data, the rules, and the roadmap. Some businesses cannot accept a vendor changing features, raising prices, or discontinuing a product on short notice. Others have audit and compliance obligations that need the system to behave in a specific, provable way. Owning the platform puts those decisions back in your hands.

One signal on its own rarely justifies a build. Two or three together usually do.

What custom software actually costs

Our custom platform engagements typically run R150,000 to R750,000, over three to six months. That is a real range, not a teaser, and where a project lands in it depends on the number of integrations, user roles, and workflows the platform has to carry, plus any compliance work.

The fair comparison is not the build cost against one month of subscription fees. It is the build cost plus ongoing maintenance, set against the total cost of the off-the-shelf alternative over several years: the licences, the add-ons, the workaround labour, and the price rises. Run that comparison honestly and the answer is often clear. Sometimes it clearly says buy.

Why this matters

The wrong software choice rarely fails loudly. It fails slowly. Staff lose hours each week to tools that do not fit. Data lives in three systems that disagree with each other, so nobody fully trusts the reports. Errors creep in at the hand-over points between systems, and some of those errors reach your customers as late deliveries, wrong invoices, or missed follow-ups.

That last part is the expensive one. Customers do not see your software; they see the mistakes it lets through. A business that looks disorganised from the outside loses trust it spent years building, and that damage is far harder to repair than a line in the IT budget.

The opposite mistake costs just as much: commissioning a custom build for a problem an existing product would have solved, then carrying the upkeep of a platform you did not need.

What it looks like in practice

Every custom platform we have built started with the catalogue failing to fit.

• Palabora Mining Company needed real-time dispatch with block-cave draw control. That was not in any vendor's catalogue, so we built BCD Dispatch with Graybeard Solutions. It coordinates LHD (load-haul-dump) fleets across drawpoints in real time and runs 24/7. Read the Palabora case study.
• DM3, our own monitoring platform, replaced off-the-shelf agency dashboards that would have cost roughly R10,000 to R50,000 per month at the scale we run, and none of which combined the sources we needed. Read the DM3 case study.
• BusinessLink needed multi-company digital business cards with per-company branding and batch onboarding for hundreds of staff at once. No digital card product we could find did that. Read the BusinessLink case study.

None of these were built because custom software was exciting. They were built because the operation had outgrown what the market offered.

Where to go from here

If two or more of the signals above sound familiar, it is worth a conversation. Have a look at how we approach custom platforms, internal tooling, and SaaS development, or get in touch and describe the problem. If the honest answer is to buy, we will tell you.