The real cost of running a factory on spreadsheets
Nobody budgets for spreadsheets, which is how they become the most expensive software in the factory. The licenses are free. The costs arrive in other clothes.
The reconciliation payroll
Count the people whose morning starts by copying numbers from one sheet to another, or asking a department what its sheet says, and multiply by half a working day. In mid-sized factories this quietly staffs a small department whose entire output is establishing what already happened. That is the first invoice: paid monthly, booked nowhere.
The version tax
Every spreadsheet that matters exists three times: the planner's copy, the merchandiser's copy, and the one attached to last Tuesday's email. Decisions get made on whichever copy is open. The cost is not the occasional big blunder; it is the daily friction of nobody being quite sure, and the meetings that exist only to synchronize versions. A production meeting run this way spends its first twenty minutes on archaeology, which is why we wrote about what a meeting looks like when every number drills to its documents.
Discovery lag, priced
A spreadsheet is updated when someone gets to it. The gap between an event and its entry is discovery lag, and each day of it converts recoverable problems into expensive ones: the short fabric found at cutting instead of at booking, the behind-plan line found at day end instead of hour four, the overbooked week found when the buyer calls. Air freight, overtime, and discount letters are largely discovery-lag charges.
The knowledge mortgage
The formulas, the lookup sheets, the one workbook that computes costing correctly: they live in a person. When that person resigns, the factory discovers that its process documentation was an employee. Training the successor takes months and reconstructs perhaps ninety percent, and nobody knows which ten percent is missing until it bites.
What spreadsheets are actually for
None of this is contempt for the tool. Spreadsheets are superb for modeling, for one-off analysis, for thinking. The failure mode is specific: a spreadsheet as the system of record for facts that many people update and depend on: stock, plans, orders, quality results. Records need one home, an audit trail, and simultaneous visibility, which are precisely the properties a shared workbook cannot provide.
A one-week measurement
Pick one week and tally three numbers: hours spent transcribing or reconciling between sheets, decisions delayed waiting for someone's file, and problems whose late discovery had a price you can name. Most owners who run this exercise stop asking whether real systems are worth the money and start asking why the spreadsheet era lasted as long as it did.