One Typo in a Spreadsheet Cost $92 Million — Here's Why

Two researchers, Simon Thorne and Angela Collins, went through the EuSpRIG archive — a long-running collection of spreadsheet horror stories — and cleaned it up into 121 documented incidents stretching from 1995 to 2026. They sorted every case into twelve failure categories and asked three questions: what went wrong, in what setting, and what did it cost? The answer is unsettling in its ordinariness.

The first pattern is plain human slips. Formula errors, mistyped numbers and mishandled data account for 71 of the 121 cases, and they never stopped happening. In 1995, Fidelity lost money to a missing minus sign. In 2024, a single mistyped date cost Norway's sovereign wealth fund roughly $92 million. Three decades apart, same basic mistake.

The second pattern is newer and rising fastest: information that sits inside a file but is invisible to the person receiving it. Hidden rows, hidden sheets, leftover pivot-table caches and embedded objects don't make any calculation wrong — but they can still cause serious harm. The worst example: a 2022 UK Ministry of Defence spreadsheet that exposed the personal details of about 18,700 Afghan applicants, simply because rows were hidden rather than deleted.

The paper's conclusion is blunt. Spreadsheet risk is no longer only a question of wrong numbers; it is also about visibility, disclosure, governance and trust. Most organisations treat spreadsheets as casual office tools while quietly running critical operations on them — payroll, budgets, applicant lists, safety records. Because that authority has grown, the same old mistakes now carry far bigger consequences. The practical takeaway is small but valuable: before you email a spreadsheet, open it, unhide everything, and check what you're actually sending.