Garbage In, Garbage Out: Priming Organizations for Success at the Dawn of Automation

“How Accurate are the Scanners We’re Using to Digitize Documents?”

Successful automation of document-intensive workflows begins with the answer to this question. To many buyers, it seems like this could be found by just looking at the DPI (dots-per-inch) specification for the scanner. DPI is important for image clarity, but data accuracy goes far beyond that. OCR (Optical Character Recognition) rate is the primary measure by which scanned documents are judged by the software responsible for converting the words and numbers within a captured image into usable, structured, readable data required for downstream automation opportunities.

Check reviews from independent testing labs like BLI (Buyer’s Lab International) that analyze scanners’ OCR and color dropout performance. The ratings you find are percentages of characters read accurately by software recognizing and extracting data from the scanned document. A rating of 75 means that for every 100 characters in a scanned document, 75 are read correctly. If those 25 incorrect characters span 10 of 20 critical form fields, the data entering the process you’re trying to automate has been reduced to 50% accuracy. Garbage in, garbage out.

To ensure quality input to your business process, look for scanners that produce images with a minimum OCR rate of 90 for black text and at least 85 for red, blue, and green colors. The best software can often compensate for the remaining gap and bring your OCR accuracy close to 100%.

This translates to fewer potential errors in the automated processes that require accurate data, fewer worker-hours spent correcting those errors, and fewer dollars lost to billing discrepancies.

In the age of automation, capturing accurate data from paper documents with a high-quality document scanner can deliver residual benefits for many years to come.