If documents are not already available digitally, but rather in paper form or as microfilm, they must first be digitized. This is usually performed with scanners, which create an electronic image of a document, and are thereby already capable of performing a large number of processing steps. Tasks such as empty page removal, straightening, noise reduction, sharpness optimization and barcode recognition are performed during the scanning.

Automatically generated documents that reach the company in a computer-readable form ("Coded Information"), e.g. from the own or external ERP systems, can be transferred and indexed by automated COLD procedures. The term COLD is historically derived from "Computer Output on Laser Disk", but is still used today for modern electronic formats such as XML. This process automatically extracts the desired metadata from the structured electronic documents and uses it for the indexing.

Information can be extracted from scanned documents whose digital images are only available as image formats (e.g. TIFF/JPEG), and then used to index the documents through the use of modern OCR/ICR methods. OCR is the automatic recognition of printed characters ("Optical Character Recognition"), whereas ICR ("Intelligent Character Recognition") is typically used as a term to recognize handwriting and is often KI-supported in a modern form using artificial intelligence.
Such acquisition processes are very often performed by specialized external service providers who also perform the necessary data extraction as well as the indexing, and then make the indexed documents available for import in a specified format together with the associated index data (often in XML format).

Very often such data entry processes are carried out by specialized external service providers who also perform the necessary data extraction and indexing and then make the indexed documents available for import in a defined format together with the associated index data (often in XML format).

As a hybrid of document image and associated metadata, PDF/A-3 allows the fully automated capture and indexing while ensuring a representation similar to that of a paper document. The PDF/A-3 document format, as used in the ZUGFeRD standard, offers an option to use such mixed forms.

Even without a content extraction, the further processing in the company can be ensured: for example, by applying barcodes and then linking the barcode, and classifying the document in the processing systems.