What is OCR?
Optical character recognition (OCR) converts images of text, such as scans and photos, into machine-readable characters, without understanding what the text means.
In more detail
OCR returns text and positions. It does not know which number is the invoice total. That is why OCR on its own needed templates ('the total is in this box'), and why a template breaks whenever a supplier changes layout. Accuracy depends on scan quality, fonts, skew, stamps over text and handwriting. Clean digital PDFs often carry a text layer already, so OCR is unnecessary for them. For mixed inputs, OCR is the first step, and a classification and extraction layer (IDP) does the understanding. Measure the character error rate on your real scans before you choose a tool.
Example
A finance team scans 500 paper delivery notes a month. OCR produces raw text, then an extraction step picks out the PO number and quantities. Skipping the second step leaves staff reading text dumps.