Spreadsheet Risk

The Real Cost of Spreadsheet Errors in a Business

Spreadsheet errors cost money in four places: rework, wrong decisions, money that moves incorrectly, and the hours spent checking for errors in the first place. The last one is the largest and the least visible, because it never appears as an error — it appears as normal work.

7 min readUpdated 2026-09-15

When the checking hours exceed what the fix would cost, Datanova helps businesses replace the spreadsheet with a system that validates at entry.

Why spreadsheet errors are normal, not negligent

Academic audits of operational spreadsheets have repeatedly found errors in a large majority of the files examined, and the people who built those files were competent. That result is a property of the medium, not of the people.

Nothing prevents a wrong value being entered — any cell accepts anything
A formula dragged one row short produces a plausible wrong total, silently
A paste can overwrite a formula with a value and nothing announces it
No test suite runs when the logic changes, because there is no test suite
The person checking the file is usually the person who built it

Where the money actually goes

Cost	What it looks like	How visible is it?
Checking	Hours every cycle spent verifying figures before they go out	Invisible — it is counted as normal work
Rework	Redoing a report or a batch of invoices after an error surfaces	Partly visible, rarely attributed to the spreadsheet
Wrong money movement	Mispriced invoices, credit notes, overpayments, margin leakage	Visible, and usually blamed on a person
Wrong decisions	Stock, pricing or capacity decisions made on a bad number	Almost never traced back
Customer trust	Disputes you cannot settle quickly because there is no history	Visible late, as churn

The first row is usually the biggest number, which surprises people. Checking is not error cost in the intuitive sense — nothing went wrong — but it is time spent solely because the medium offers no guarantees.

Putting a number on it

You do not need a study. You need four figures from your own process, and estimates are fine — the goal is an order of magnitude, not precision:

Checking: hours per cycle spent verifying figures, times cycles per year, times a loaded hourly cost.
Rework: how many times last year something had to be redone because a number was wrong, times the hours each time.
Money moved wrongly: total credit notes, price adjustments and write-offs traceable to a data error. Your finance team can usually produce this.
Add them. Compare the total against the cost of automating the validation, or of replacing the file.

Do this before asking anyone for a quote. Most teams have never added these up, and the total is frequently larger than the project they were hesitating over.

Reducing it without replacing anything

Several of these are worth doing this week regardless of any larger plan:

Protect formula ranges so they cannot be pasted over
Use data validation on input columns — dropdowns instead of free text for anything with fixed options
Add cross-check formulas that compare two independent routes to the same total and flag a mismatch
Separate input, calculation and output onto different sheets so pasting cannot reach the logic
Convert ranges to tables so formulas extend automatically instead of falling one row short
Automate the recurring checks, so verification stops being a person reading numbers

These reduce the error rate meaningfully. What they cannot do is make bad data impossible or tell you who changed a figure — those are structural, and they are what a system provides.

Frequently asked questions

How common are errors in business spreadsheets?

Studies auditing real operational workbooks have consistently found errors in the large majority of files examined. Treat the precise percentages with care — sample sizes and definitions of error vary — but the direction is consistent enough to plan around: assume any complex spreadsheet that has never been audited contains errors.

Can we prevent spreadsheet errors with better training?

Training reduces them and cannot eliminate them. The medium accepts any value in any cell and gives no feedback when logic breaks. Careful people using a tool with no guardrails still produce errors, which is why the fix is structural rather than behavioural.

What is the cheapest way to reduce spreadsheet errors?

Data validation on input columns and protected formula ranges, together, take about a day and prevent a large share of the common failures. Add cross-check formulas that compare two independent calculations of the same total and flag mismatches.

Does moving to a web application eliminate errors?

No, and be sceptical of anyone who says it does. It changes which errors are possible: validation at entry stops whole categories of bad data, and an audit trail makes the rest findable and attributable. People can still enter a wrong-but-valid number.

How much is checking costing you?

In a free consultation we look at the verification your team does every cycle and tell you what it would take to automate it — or whether the process needs something more than validation.

Book a Free Consultation