Reestablishing Unambiguous Data

August 9, 2004 by admin

The second week of March 2004 marked a turning point for the Erfurt Municipal Utility. More than one data record had erroneously been created for more than 45,000 of its more than 210,000 standard customers. Within 36 hours, the duplicate records were tracked down and each of the corresponding contracts was consolidated into a single customer record. The utility extensively planned for and tested the data cleansing. At the end of the process, Mike Wahlicht, project director at the utility said, “The consolidation on the production system took 36 hours. It went faster than planned and did not encounter any problems at all.” In most cases, the data could be cleansed automatically. Only about 70 duplicates required manual intervention because of errors in the SAP customer-change data. The utility originally planned to use Saturday afternoon and Sunday as safety buffers for such cases and to coordinate current accounts, but work was complete by noon on Sunday.
The consolidation of mass data in a productive system is a demanding task. In addition to software specialists and system administrators, user departments must be closely involved so that they can make decisions about content-oriented consolidation and the quality of the data. The cleansing of data became necessary because the utility planned to implement mySAP Customer Relationship Management (mySAP CRM). “Even the best CRM solution does not produce the expected success when the quality of the data does not enable a comprehensive view of customers,” says Manfred Wollrab, director of IT. “Cleaning our business master data created the best preconditions for the upcoming implementation of mySAP CRM. And because the call center at the utility uses cleansed data, it can identify customers and process their requests more quickly, so it was clear that all the efforts would be amortized quickly.” It was important for the utility to work with a coherent product that could perform all the required tasks – from identifying duplicates to coordinating accounts – and could be operated simply.

Cleansing in preparation for a long time

On January 1, 2001, the Erfurt Municipal Utility migrated from SAP R/2 RIVA to the current accounting solution of the SAP industry solution for utilities, SAP IS-U/CCS (today called SAP for Utilities). During the migration, data from several legacy systems was transferred into one accounting system – the primary reason for the existence of duplicates. Duplicate customers cost real money because of increased processing costs in the call center or more correspondence from customers. And the duplicates also stood in the way of implementing mySAP CRM and SAP Business Intelligence (SAP BI).
The municipal utility has worked on its data cleansing projects since the summer of 2003. After design and planning, the first tests began on a mirrored system in December 2003. After optimizing the test procedures, the utility began three cycles of mass tests before consolidating the production system.

Consolidation with the customer change transaction

To consolidate the data, the utility chose the P305 business partner consolidation solution of PCO GmbH. “When selecting the software, we focused on a solution that performed database updates with SAP function modules,” says Wahlicht. The PCO solution is based upon an SAP dialog transaction called Customer change. The transaction uses a transfer contract to migrate a contract from one customer account to another. This approach means that customer consolidation can occur risk free and can be cancelled. No final invoice must be created for a customer, so the actual affects of the process for customers are minimal.
Before the consolidation, the software automatically checked the upcoming Customer change transaction for factual restrictions. That process disqualified about 1,500 partners because of periodic billing or because accounts had gone into collection. The check ensued that the automatic consolidation would not terminate abnormally. During the consolidation of approved business partners, the PCO product saved the account balances before changing the customer. The open claims, payments, rates, and budget billing plans were transferred into one master data record along with the contracts. The account balances were then saved again and underwent an automatic comparison.
“With more than 28,500 customer changes, the process was an impressive performance,” says Mike Wahlicht with some pride. The saved account balances are still available for informational and auditing purposes. SAP tables, processing logs, and the application log recorded all changes, which are also stored in change documents.

Online checking for duplicates is now standard

During business partner consolidation, recognition of duplicates is realized by distance control logic (DICOL), which has been used successfully for many years. DICOL examines the business partners in SAP for Utilities for duplicates by using individually defined criteria and similarity requirements. The procedure reliably recognizes and displays typical input errors for names and addresses – Green or Greene, for example. Phonetic checks, such as “ph” and “f” can be activated easily. To estimate the scope of the duplicates present, the identification runs can be repeated at various levels of sensitivity and simple statistics derived from the runs.
The Erfurt Municipal Utility also uses PCO’s P324 solution to check duplicate business partners to protect it from new duplicates in the future. When new business partners are created online, it checks for the presence of similar records in the system and prevents the creation of new duplicates as much as possible. Nevertheless, the consolidation of duplicates is generally not a one-time task. The data should be examined for duplicates at regular intervals. “During peak hours, the call center does not have enough time check for existing business partners when creating customers. That’s why a user-friendly solution is so important to us,” says Ute Hartwig, director or IT applications.

Bernd Knobel

Tags: ,

Leave a Reply