Excel isn't the villain. It's the witness.

The file is called “Control_FINAL_v7_really.xlsx”. It has eleven tabs, formulas calling other formulas and a macro nobody fully understands but nobody dares delete. One person maintains it. When that person goes on holiday, month-end close slips a week.

It would be easy to make Excel the villain of this story. It would also be unfair. Since VisiCalc appeared in 1979 and turned the personal computer into a business tool, the spreadsheet has probably been the most democratic software in business history: it lets anyone model a problem without asking anyone's permission. That's exactly its virtue, and its risk.

How a spreadsheet becomes a system

Nobody sits down and decides bank reconciliation will depend on a macro. It happens gradually. The ERP doesn't have the report you need, someone exports the data, matches it by hand and the result works. Next month it happens again. By month three a tab gets added. A year later, the spreadsheet isn't supporting the process: it is the process.

That spreadsheet acts as a system without a system's controls. In this scene it has no formal owner, change history or approval record. Other companies do govern critical spreadsheets; the risk arises when one becomes essential without anyone taking responsibility for those controls.

The mistake of declaring war on it

The usual reaction is to buy a platform, migrate the spreadsheets and call the problem solved. A year later the mess is the same, in a more expensive tool. Metrics change formula without anyone knowing why, every team keeps its own customer list and exceptions still get approved over WhatsApp.

A new tool on top of rules nobody governs is the same problem, with better design.

In this scene, Excel was the witness to a wider governance failure. Nobody in the company explicitly decides three things about each important rule: who defines it, who can change it and where the change is recorded. That's governance, and it doesn't take committees or two-hundred-page documents. It takes names.

What to do instead of banning it

Banning Excel without replacing what it provides often fails: people go back to it because it solves something the official system doesn't. The useful question is which parts of each critical spreadsheet are already business rules. The matching criteria, the approvals, the definition of each number. Those rules deserve an owner and a system with permissions and a record, where they run exactly as written.

The rest, one-off analysis and exploration, can stay in Excel guilt-free. That's where it shines. The difference isn't that the company stops using it. It's that operations stop depending on it.

The scenes in these episodes are composites of common situations, not customer cases. Any figures in them are illustrative.