Off-the-shelf software wins on day one: lower price, instant deployment, proven features. Custom software wins on day one thousand: no per-seat licensing, no workaround tax, no ceiling on fit. The break-even point arrives when subscription costs plus workaround labor plus lost capability exceed a custom build amortized over its life — for most growing businesses, somewhere between years two and four. Here's how to compute your own tipping point instead of trusting anyone's sales pitch, including ours.
A confession from a custom software company
We build custom software for a living, so you'd expect this article to conclude that everyone needs it. It won't. We have told prospects — more than once, at the cost of six-figure projects — that a ready-made tool would serve them better. One of those conversations is the reason a client later trusted us with a much larger build.
The truth is that this decision has a real answer for every business, and it's arithmetic, not philosophy. What follows is the arithmetic.
What off-the-shelf really costs (it's not the subscription)
The subscription is the visible cost, and it's honest enough: per user, per month, forever, rising with your headcount. The invisible costs are where budgets quietly die:
The workaround tax. Generic software fits your process approximately, and “approximately” has a payroll. The export-to-Excel ritual. The second tool glued on to cover a gap. The person who spends every Friday reconciling what the software almost does. Measure it once: count the hours your team spends working around the tool rather than in it, multiply by loaded cost, annualize. In our audits this number regularly embarrasses the subscription fee.
The ceiling. Somewhere in year two you'll want a feature the vendor's roadmap doesn't include. You can vote for it in their feature forum — alongside ten thousand other customers whose processes differ from yours. Your competitive edge waits in a queue.
The data hostage clause. Leaving means migration, and vendors know it. Exports are grudging, formats are awkward, and switching costs compound the longer you stay. Pricing tends to reflect this confidence.
What custom really costs (it's more than the quote)
Symmetry demands honesty in this direction too. A custom build costs its development price — our documented projects run $50,000 to $250,000, with focused tools well under that range — plus maintenance (a sensible planning figure is 15–20% of the build cost per year), plus the patience to wait weeks for version one instead of deploying this afternoon.
And custom done badly is the worst of all worlds: an unmaintained bespoke system whose original developer vanished is more expensive than any subscription. If you engage a builder, the questions that matter are: who owns the code (you should, 100%), what does the documentation look like, and who maintains it in year three. Get all three in writing.
When off-the-shelf honestly wins
Solved problems. Accounting, email, payroll for standard structures, CRM for a conventional sales motion — these are mature categories where vendors have invested thousands of engineer-years. Rebuilding them is vanity spending.
Undiscovered processes. If your business is young and its workflow still changing monthly, you don't yet know what to encode. Rent software while you learn; build when the process stabilizes.
This logic, incidentally, is why we sell ready products ourselves. Manufacturing operations, cleaning-business logistics, campus management — solved problems don't deserve custom prices, which is exactly what Rexo ERP, CleanPlan, and Education ERP exist for. The honest vendor answer is sometimes “neither custom nor generic — configurable.”
When custom pays for itself
Your process is your edge. If the way you quote, schedule, price, or serve customers is why customers choose you, forcing it into a template erases the very thing worth protecting. A century-old manufacturer came to us because no vendor sold software for how their divisions, locations, and machines actually worked; the custom platform we built cut their costs 40%. That result was only available custom — the process being encoded was theirs alone.
The integration knot. When you're running five tools connected by exports and prayers, a custom system that makes them one often costs less than the coordination it replaces.
Scale economics. Per-seat pricing punishes growth. At 20 users a subscription is cheap; at 200 users, do the ten-year multiplication and the custom quote starts looking like a bargain with better fit.
Compute your own tipping point
Take four numbers: (1) annual subscription cost at your projected headcount, not today's; (2) annual workaround tax, measured as above; (3) a revenue estimate for the capability you can't have — the feature stuck in the vendor's forum; (4) a real custom quote, divided by five years, plus 20% annual maintenance.
If (1) + (2) + (3) exceeds (4), custom pays for itself — and every year after break-even, the gap widens in your favor. If it doesn't, keep the subscription with a clear conscience and re-run the numbers in a year. Businesses that do this math annually make the switch at the right time; businesses that don't tend to make it three expensive years late.
The decision, minus the sales pressure
Ready-made for solved problems and unstable processes. Configurable products for solved problems with your industry's shape. Custom for the process that makes you money and fits nothing on the market. Most companies eventually run all three — the skill is knowing which category each problem belongs to.
If you want the arithmetic run against your real numbers, that's what a scoping call is for: you'll leave with a fixed, itemized quote — or with our honest recommendation of a tool we don't sell. Both outcomes have happened. Both are free.
Explore custom software development