November 2, 2026
Spreadsheets vs custom systems: what the switch actually costs
Spreadsheets look free and systems look expensive — but both carry hidden lines. An honest two-year comparison: what each state of a business really pays, and how to time the switch without over-buying.
Share this article
The short answer: spreadsheets don't cost nothing and systems don't cost just their invoice. Spreadsheets bill you in errors, rework, reconciliation hours, and dependence on individuals; custom systems bill you in build cost, phase discipline, and maintenance. For a small operation the spreadsheet wins; past certain complexity the system wins — and the art is timing the crossover before the spreadsheet's hidden bill exceeds the system's visible one. Here's the honest ledger, line by line.
The spreadsheet's real bill (the invisible lines)
- Error cost. A wrong formula quietly multiplies across weeks; a overwritten row deletes an order from history. Every error has a recovery cost — finding it, redoing it, apologising for it — and they occur at rates no one has measured because the tool never recorded that anything happened.
- Reconciliation time. Someone merges copies of the truth: sales's sheet, stores' count, accounts' version. This isn't one person's salary line; it's paid in the FOUNDER's hours, at the most expensive rate in the company, every week.
- Key-person risk. The one person who understands the workbook — its hidden tabs, its macros, its conventions — is a single point of failure you're not insuring.
- Scale friction. Cell locks, versions, "final_v7". Every extra user adds coordination load non-linearly: two users, mild pain; six, daily fights.
- Evidence weakness. For disputes (customer debt, staff claims, supplier disagreements), a spreadsheet convinces nobody, because anyone could have edited it. Systems with audit trails end arguments; sheets prolong them.
The custom system's real bill (the visible lines)
- Build cost, phased properly. A disciplined version 1 — the money process, decent mobile UX, audit trail, five reports — costs real money but is FINITE. The known scam in this lane is the unbounded phase that never ships; the known misbuy is the over-scoped version 1 you never needed (our outgrowing-spreadsheets guide lists what a v1 must contain).
- Maintenance. Software is pet, not furniture — updates, fixes, questions, small changes. Budget the recurring line; don't leave the build without a named maintainer.
- Change management. People must actually switch. Training, a transition period where both systems run, resistance, and a deadline for the sheet's funeral. Budget weeks, not launch day.
- Data migration. Years of spreadsheet rows, quoted items, half-formed notes — someone cleans and imports them. It's a real cost, usually small but never zero.
The crossover rule of thumb: when the spreadsheet's annual hidden bill (errors + reconciliation + key-person risk) approaches the system's annual maintenance, the switch is already late. Businesses rarely see this moment — which is why they decide by anecdote instead.
Timing the switch sensibly
- Cost your current process honestly. Count reconciliation hours per week × weeks × your own hourly worth; count the last two visible errors' recovery cost; add the cost if the key person quit this Friday. That number is your current system's price — write it down.
- Scope the smallest system that deletes the biggest pain (one workflow, not seven). If a proposal can't be scoped that small, keep shopping or check off-the-shelf tools first.
- Insist on trialability. A demo on YOUR data beats ten slide-deck promises; a phased build with a first deliverable you can abandon beats a 100%-upfront bet.
- Plan the exit. Whatever you build, keep your data exportable — the day the system silences the business, you'll want the door your data needs.
Where custom genuinely pays, and where it doesn't
Custom pays where your process is your moat: the quoting engine, delivery zoning, credit tracking, or production flow nobody else can buy off the shelf because it IS your advantage. Custom doesn't pay for commodity needs (invoicing alone, basic stock lists) — rent those from existing tools and save the custom budget for the money workflow. The full decision path is in the outgrowing guide; once the system is real, our ERP cost guide covers whether to unify everything under one spine.
How we price these at IOI
Our business systems work is scoped phase-by-phase with a published per-phase price and an exit point after each — the structure exists so you're never further than one phase from deciding it was a mistake. Bring your worst week (the one with the duplicate orders and the missing credit note) and we'll scope what digitising that week requires, with the cost of doing nothing spelled out beside it.
Frequently asked questions
What's the biggest hidden cost of spreadsheets?
Key-person dependence: the workbook only one person fully understands. When that person is away or leaves, the business's operational memory leaves with them — an exposure most owners underestimate until they live it.
How much does a custom system cost to maintain yearly?
Real maintenance is proportional to the system's surface area and its integrations — updates, fixes, support questions, small changes. Ask bidders to quote maintenance terms explicitly at proposal time; vendors who promise free maintenance forever are either minimising it or haven't built anything that lives.
Can I phase the migration — half spreadsheet, half system?
Yes, and it's often the right way — run the new system for one workflow while the rest stays on sheets, with a hard deadline for completing the move. What doesn't work is no deadline: the "transition period" quietly becomes the permanent state.
When is custom software the wrong answer?
When your process is standard (buy it as a tool), when the pain is actually coordination/communication (software won't fix culture), or when nobody can state in one sentence what version 1 exists to do. Our outgrowing guide's five-point minimum is a good smell test.
