Data StrategyBusiness IntelligenceSmall Business

Your business is leaking money — and your spreadsheets are hiding it

By Infra IT Consulting · June 1, 2026

Every business has a number it doesn’t see. Not on the P&L, not in the bank balance — but real, and recurring. It’s the cost of running decisions on data you can’t fully trust, assembled by hand, in tools never built for the job.

Here’s the uncomfortable part: the better your team is at managing the mess, the more invisible the cost becomes. A skilled analyst who spends three days a month stitching reports together doesn’t show up as a problem. They show up as dependable. The leak gets reclassified as competence.

The size of the leak

Gartner has estimated that poor data quality costs the average organization around $12.9 million a year. For a small or mid-size business, that figure isn’t literal — but the percentage behind it is. A widely cited Harvard Business Review analysis found that only 3% of company data records met basic quality standards, and 47% of newly created records had at least one critical error. Most organizations never measure this, which is precisely why it persists.

The cost rarely arrives as a single dramatic event. It’s the gradual tax of small things: a number that’s a week stale, two departments reporting different figures for the same month, a decision made on a report nobody had time to check.

When small errors become large ones

Occasionally the tax does come due all at once — and the examples are instructive because none of them happened to careless companies.

In 2003, the Canadian power generator TransAlta lost about US$24 million — roughly a tenth of its annual profit — when a cut-and-paste error in an Excel bid spreadsheet went undetected before submission. The CEO described it publicly and plainly: a copy-paste error in a spreadsheet, caught too late.

In 1994, an accountant at Fidelity’s Magellan Fund omitted a single minus sign while transcribing a $1.3 billion net loss into a dividend spreadsheet. The loss was recorded as a gain, throwing the estimate off by $2.6 billion and forcing the fund to cancel a promised shareholder distribution.

In 2012, JPMorgan’s “London Whale” loss — $6.2 billion — was traced in part by the bank’s own internal report to a risk model that ran through spreadsheets updated by manually copying and pasting between them. One formula divided by a sum where it should have divided by an average, and understated risk by roughly half.

And in 2020, Public Health England lost 15,841 COVID-19 test results from its daily totals because results were loaded into an old Excel file format with a row limit far smaller than the data required. The records didn’t error out — they silently dropped, and contact tracing was delayed during a pandemic.

Different industries, different decades, the same root cause: critical decisions running on manual tools with no checks, no version control, and no alarm when something goes wrong.

This is normal, not negligent

If that feels familiar, it should. Decades of research by spreadsheet specialist Raymond Panko found that the large majority of operational spreadsheets audited in the field contained errors. A separate study by researchers at Delft University of Technology examined more than 15,000 real spreadsheets from a corporate email archive and found that about a quarter of those containing formulas had at least one formula error.

The point isn’t that spreadsheets are bad. They’re brilliant — for what they were designed to do. The problem is the quiet promotion: the workbook that started as one person’s calculator becomes the company’s reporting system, its month-end process, its single source of truth. Nobody decided that. It just happened, one tab at a time.

The leak you can actually feel: time

Even when the numbers are right, the cost shows up as hours. Surveys of finance teams consistently find that a large share of the work is data gathering and cleanup rather than analysis — one study of data professionals found that nearly half their time goes to preparing data before any insight begins. A survey of CFOs found many spend around ten hours a week personally working in spreadsheets. Recent benchmarking found the vast majority of finance teams still rely on Excel to close their books, and many point to it as a direct reason the close runs slow.

Put a dollar figure on it. One analyst, two days a month rebuilding the same reports, on a loaded salary — that’s a measurable line item you’re paying every month and calling something else. Multiply across a team and the “invisible” cost becomes very visible.

The fix is not “be more careful”

Telling people to double-check won’t close the leak — the research shows reviewers miss a meaningful share of errors even when they’re looking. The fix is structural: move recurring reporting and reconciliation off manual tools and onto infrastructure built for it. A single trusted source of data. Reports that refresh themselves. A clear definition, agreed once, of what each number means.

This doesn’t have to be a large or frightening project. Independent and commissioned studies of businesses that modernize their reporting consistently report strong returns and faster closes — often cutting month-end cycle time by a third or more and freeing analysts from manual assembly. The first step isn’t software. It’s measurement: find out how many hours your team loses, and how much of your data you’d actually stake a decision on.

If any of this sounds familiar, that’s the leak. It’s fixable — and the first step is simply seeing it.

If this resonates, I offer a free 30-minute data health check — a straightforward conversation about where your reporting is costing you time and trust, with no obligation.

Sources

Need help with your data infrastructure?

Book a free 30-minute consultation to discuss your data engineering and analytics needs.

Talk to our team →