I will defend spreadsheets to anyone. They are the most successful end-user programming tool in history, and every operations platform we have ever built started life as somebody’s spreadsheet. The spreadsheet is not the mistake. Staying on it two years too long is.
The trouble is that the failure is gradual. No single Tuesday feels like the day the spreadsheet stopped coping. So here are the seven signs we look for when a client asks whether it is time, ranked roughly by how loudly they predict trouble.
The seven signs
- 1One file has a full-time babysitter. Someone spends hours each week merging versions, fixing broken formulas, and chasing people to close the file. That salary line is your first software budget.
- 2Version chaos is normal. If "final_v7_USE_THIS_ONE.xlsx" is a real filename in your business, decisions are being made on stale data somewhere, weekly.
- 3Multiple people need to edit at once. Concurrent editing on complex operational sheets ends in silent overwrites. The worst data loss is the kind nobody notices for a month.
- 4The spreadsheet has become a workflow. Colour-coded rows meaning "sent to client", columns tracking who approved what, a tab called "DO NOT TOUCH". You have built an application. It just has no permissions, no history, and no validation.
- 5Customer data lives in it. Personal information in an emailable file is a POPIA incident waiting for a mis-typed address. Access control by hope is not access control.
- 6Reporting means rebuilding. If every month-end someone spends two days copy-pasting into a management pack, you are paying analyst rates for clerical work, monthly, forever.
- 7Errors have reached customers. A wrong price quoted, a duplicate invoice sent, a delivery missed because a row was sorted wrong. Once spreadsheet errors are visible from outside, the cost has moved from annoying to reputational.
What it is costing you right now
Put numbers on it before deciding anything. A logistics client of ours, EvolveVision, ran their fleet allocations on eleven linked workbooks. When we costed the babysitting time, the month-end rebuild, and the error rate, the spreadsheets were quietly costing around R38 000 a month, before counting the two customers lost to missed deliveries that year.
11
linked workbooks running one fleet operation
R38k
monthly cost of spreadsheet upkeep and rework
14 weeks
to replace the core workflow with a real system
What graduation actually looks like
Moving off spreadsheets does not mean a two-year enterprise project. The right first step is usually small: take the single most painful workflow, the one with the babysitter, and move only that into a proper system with logins, validation, history, and one screen per job. Everything else can stay in spreadsheets until it earns its migration.
Sometimes the answer is off-the-shelf software, and we say so. Our own product exists because construction teams were running document control and site operations across exactly this kind of spreadsheet sprawl, and an industry tool like fluxems fit the job better than a custom build. For workflows that genuinely are your secret sauce, a focused custom tool wins. We wrote a full framework for that call in build vs buy.
Keep the spreadsheet where it wins
After the move, spreadsheets still have a job: ad-hoc analysis, one-off models, exploring an idea before it deserves software. The goal is not a spreadsheet-free business. The goal is that nothing irreplaceable lives in a file that one wrong sort can corrupt. Prototype in the spreadsheet, run the business in the system.
If you counted three or more signs while reading, we do a short assessment that maps your spreadsheet estate and prices the two or three options honestly, including the option of staying put for another year. Start with a consultation.