OCR
While researching features pertaining to coding image only PDFs, it became apparent that what people actually needed was an OCR capability within the product
First iteration
The first iteration gave people the ability to review the detected text side by side with the original document. If they highlighted text on the PDF, the detected text would be highlighted on the right. So confident were we with the design, that we almost took it to completion.
However, on testing, people assumed that the saved output would be the plain text on the right pane, and not something identical to the original.
Second iteration
So we removed the split pane all together, and let users highlight sections of the PDF to see what the detected text was. This eliminated the initial confusion on what the output would be
Delivered version
