The real cost of running your business on spreadsheets

Nobody chose to run their business on spreadsheets. It happened the way most things happen: a sheet solved a problem on a Tuesday, then became the place that information lived, then became load-bearing. By the time anyone questions it, four people depend on it and one of them built it.

The argument for keeping it is that it is free. That is true of the licence and untrue of almost everything else. Here is how to work out the actual figure, using numbers you already have.

Count the re-entry, not the typing

The first cost is entering the same information more than once. A customer's details go into the sheet, then into the invoicing tool, then into the scheduler. Each copy is a few minutes, and each is invisible because it is spread thinly across the week.

To measure it, pick one recent customer or job and physically trace every place their details were typed. Count the systems, not the keystrokes. Then multiply by how many new customers or jobs you handle in a month. Most businesses are surprised by the multiplier rather than the per-instance cost.

Count the reconciliation

When two systems each hold a version of the truth, somebody has to make them agree. This is the nightly export, the month-end tie-out, the hour before the management meeting spent making the numbers match.

This one is easier to price because it is usually a named task done by a named person on a known rhythm. Take the hours per month and multiply by that person's fully-loaded cost. If the person doing it is you, use what an hour of your time is worth when spent on winning work instead.

Count what the errors cost

Spreadsheet errors are not random. They cluster in predictable places: a formula that stopped covering the last row, a copy-paste that carried the wrong customer name into a quote, a filter left applied when someone read a total.

You do not need an error rate to price this. Instead, list the mistakes that actually happened in the last twelve months and what each cost to fix, including the credibility cost where a customer saw it. Businesses usually find between two and five incidents they had already half-forgotten.

Count the single point of failure

Somebody understands the sheet. They know which tab is current, why that column is hidden, and which figures not to trust. This is the cost that never appears on a report until the week they are on leave or the month they resign.

Price it as a question rather than a number: if that person left tomorrow, how many days would it take someone else to run this reliably, and what would go wrong in the meantime? That is not a rounding error, it is a business risk you are currently carrying for free.

Count the ceiling

This is the largest cost and the hardest to see, because it is the growth that did not happen. Manual admin scales linearly with volume. Doubling your customers doubles the entry, the chasing and the reconciliation, which means the next stage of growth requires hiring for admin rather than for the work you actually sell.

A useful test: imagine your enquiry volume tripled next quarter. Would that be good news, or would it be a staffing problem? If the honest answer is the second, your tools are already limiting the business, whatever the licence costs.

Then compare it against the right thing

Add the five figures up and you have an annual number. Compare it against the cost of fixing it, but be careful to compare against the right fix. The instinct is to compare against a full custom system. Often the correct comparison is much smaller:

• Connecting two tools you already pay for, so the re-entry stops
• Automating one chase sequence, so the follow-ups happen on time
• Moving one process out of the sheet, leaving the rest where it is

Most businesses do not need to replace everything. They need to remove the two or three worst joints. The value of doing the arithmetic above is that it tells you which joints those are, and it converts a vague feeling that things are inefficient into a number you can act on.