Automated invoice processing at BASF

Approximately 2,000 to 3,000 invoices arrive at the Accounts Payable department (AP) at BASF in Ludwigshafen (Germany) every day. Many of these allow significant savings through discounts if processed fast enough.

The Challenge

BASF identified the following objectives and criteria to opt for the automated invoice data capture solution by Paradatec and IBM:

• reduction of processing time (1-2 days versus 4 days),
• improvement of data quality,
• higher degree of transparency in the entire process as well as an accelerated flow of information,
• automatic deadline monitoring for workflows & automatic assignment of tasks by SAP workflow,
• improved systems integration and
• reduction of processing costs (personnel).

The Bottom Line

Oliver Bieder, manager of the BASF project team, described the advantages of this automatic invoice processing as follows: “We shifted most process steps from the desk to the screen. Via early scanning we are able to avoid errors which occur for example when typing in figures, and of course the electronic way is much faster than the manual data entry.” (Cit.: Ulla Biernat, “Früh gescannt – schnell geprüft“, published in: BASF information July 2004)
Qualitative improvements are achieved as typical human mistakes arising from manual data entry are now avoided. Most of the data is transferred automatically, accurately and at very high speed into the workflow. Since the start of production the system has been successfully running and achieves an 85.7% automation rate across all relevant fields – including complete line items extraction.

“Early Scanning” allows same day processing of a large proportion of invoices (reduction of processing time: 3 days). An automatic assignment of tasks, as provided by the SAP workflow, is now possible and the particularly labor and cost-intensive transport of original documents between the different stations is no longer necessary.

Further Information

You can download the complete case study here: BASF Case Study (PDF, 300kB)