Customer Story
How KPMG's tax practice automates SAP data with Alteryx and the DVW Connector for SAP
KPMG's tax practice runs state income tax apportionment for its clients, work that leans heavily on SAP data. For one major client, getting that data out of SAP and into shape was taking thousands of hours a year.
The challenge
Apportionment meant pulling SAP extracts and BW reports into more than 50 revenue workpapers, plus fixed assets, property and rent, and because state tax rules differ by jurisdiction the team kept 700 separate workpapers for 700 state filings. Thousands of contracts blew past the report limits. Their workaround was a custom RPA bot that logged into SAP to pull each report, and it timed out, crashed, returned incomplete data, and no one in-house could maintain it. Any upstream change meant rerunning everything, so the team put off starting until the last minute.
The solution
KPMG replaced the RPA entirely with the DVW Alteryx Connector for SAP. SAP sales and fixed-asset data is pulled straight into an Alteryx batch macro in seconds and combined with payroll and other sources into one consolidated data model. The state tax logic lives in a separate, maintainable file outside the workflow, so changing a rule is a quick edit and a rerun. The connection works at the application layer, signs in with each user's own SAP credentials so it only reaches what they are authorized to see, and puts no strain on SAP.
The outcome
Thousands of hours of manual apportionment now run as a workflow that regenerates at the push of a button, with full data lineage and reviewer checks built in. It scales, rerun for the whole year or reuse next year in a click, and the data stays clean because nothing is rekeyed by hand.
"We completely automated the extraction from SAP and no longer needed RPA. In seconds, we retrieved all of the SAP data needed for the process." KPMG Tax practice, as featured in the Alteryx and KPMG webinar
Source: Alteryx and KPMG on-demand webinar, Turn Insight into Opportunity · dvwanalytics.com