Skip to content

Should You Build It or Buy It?

This puts honest numbers on a decision most people make on the sticker price alone: should you build a lean version of a system or rent one off the shelf. You put in what the subscription costs, how many seats, how long you’ll realistically use it, and how much of what you need is genuinely your own workflow versus a commodity. It returns the real five-year picture — what renting actually costs over time against what owning a lean version would — so you’re comparing the whole bill, not this month’s invoice against a build quote.

How to read your result

The number that matters isn’t the monthly subscription. It’s the total over the time you’ll actually run this, and the calculator’s whole job is to surface the costs that don’t show up until later: the per-seat creep as you grow, the tier you’ll be pushed onto for the one feature you need, the integration work you’ll do anyway, and the fact that the day you stop paying, you have nothing. Read the result as the true cost of renting, lined up against the cost of owning — because that’s the comparison nobody runs and the one that changes the answer.

When the result leans toward build, it’s usually telling you that what you need is small, specific, and yours — a workflow only your business runs — and that a few weeks of focused work buys you something you’ll never pay rent on again. When it leans toward buy, take that seriously too: the calculator is not rigged to always say build. It’ll point you at buying when the thing is a commodity — payroll, the accounting ledger, card processing — where the off-the-shelf option is cheap relative to the build, battle-tested, and getting it wrong has legal consequences. Don’t build your own payroll engine to save a few pounds a month. That’s not leverage, it’s risk you didn’t need.

The result that should make you stop is the close call that still says buy. A close call usually means you’re about to rent something you could own — and ownership wins ties, because the costs of renting keep arriving forever while the build is paid once. If it’s near the line, lean build.

What I’d do about it

If it says build, don’t build the whole thing — that’s the mistake that makes building look expensive. Build the 20% that does your actual job, and build it fast. The work one off-the-shelf tool genuinely does for you is usually small and well understood; a lean version is a few weeks, not a moonshot, and the year-one subscription money you’d have spent often covers the build outright. Then the real win lands in year two: with SaaS you start paying again from zero, while with a system you own you’ve banked that money and you compound — you extend the MVP into the next thing and connect it to the system next door. The way I’d actually approach that lean first version is here.

The reason this matters more than the spreadsheet is what you’re really buying when you rent: limited access to your own data. When your operation lives inside someone else’s tool, your data sits in their schema, behind their API limits, exportable only in the shape they allow, and gone the day you leave. Own the system and the data is yours to connect, automate, and build on — which is where all the leverage actually is. That’s the worldview underneath this whole calculator, and the full argument, including the cases where I’d happily tell you to buy, is in the build or buy playbook.

If it says buy a commodity, buy it and move on — don’t let a builder’s instinct turn a solved problem into a project.

Talk it through

If the result surprised you — especially if it said build when you’d assumed you’d buy — tell me about it. Twenty minutes on your specific case usually makes the call obvious, and it’s the kind of decision where getting it right saves a five-figure mistake either direction.

If you want the full reasoning rather than the number, the build or buy playbook is the long version. Two related diagnostics from my toolset: the tech debt calculator shows what the rented, glued-together tools are already costing you in friction, and the developer day-rate benchmark helps you sanity-check a build quote once you’ve decided to build. And if the build the number points to is large enough to need a team and a process behind it, that’s exactly what my agency, Digital Royalty, is for.

Build vs buy for a payment system

Card processing itself is the clearest buy on the list. The processor carries the compliance and the fraud tooling, and a percentage of each transaction is cheap next to building any of that yourself. The build decision is the layer around it, the invoicing, the subscription logic, the reconciliation, the customer records and the reporting, which are your workflow, and renting those from a billing platform is where the per-seat and per-transaction fees start adding up. So in the calculator put the processor’s fee in as a cost you’d pay either way and compare the rest, the billing platform’s subscription against a lean layer you own that talks to the processor’s API. How big that layer is depends on how much of that list you need, invoicing on its own is small, subscriptions and reconciliation together are not.