This document discusses spreadsheet errors and studies that have found a high prevalence of errors in real-world spreadsheets. Specifically:
- Studies show that up to 90% of spreadsheets contain some type of error. Even studies of financially significant spreadsheets found errors.
- When people review spreadsheets to find errors, they are only moderately successful at detecting errors. Not all errors are caught.
- A case study example showed how a simple input error in a financial model caused a 20% difference in results. The error was only caught when the client reviewed the interim model.
- Reasons for errors include that spreadsheets are complex but safeguards and testing are limited. Developers may also be overconfident