AI-supported optimisation of master data at ERDINGER Weissbräu
25th April 2024
ERDINGER Weissbräu is the largest family-owned wheat beer brewery in Germany. For more than 130 years, ERDINGER Weissbräu has stood for sophisticated Bavarian wheat beer culture and maximum enjoyment – with total sales of 1.5 million hectolitres of beer in 2020, and the trend is rising.
Initially only sold in Bavaria and then in Germany, the private brewery now exports its 10 specialities to 106 countries on five continents and – with ERDINGER Alkoholfrei as an isotonic thirst quencher for athletes – becomes the market leader in the non-alcoholic beer sector in just five years. The 11th beer in the ERDINGER product range, the "Festweisse", is not sold on the free market and is reserved exclusively for the Munich Oktoberfest.
In September 2023, ERDINGER entrusted consenso with analysing its current business partner and material master data. The objective: to significantly increase their quality and identify potential for sustainable quality improvement.
Based on the consulting solution Data Mining & Profiling (DMP) developed and provided by consenso, ERDINGER's customer-specific requirements regarding DMP, duplicate checks and rule mining were piloted on the current S/4HANA system within four weeks.
The further procedure was based on the "CRISP-DM" (Cross-Industry Standard Process for Data Mining) methodology, a standardised approach for the implementation of data analysis projects. Accordingly, the project was divided into the following activities:
- Creation of business rules in cooperation with business departments and IT
- Optimisation of business rules through the use of artificial intelligence (rule mining)
- Analysis of identified errors
- Matching and consolidation
- Implementation of error corrections
- Continuous monitoring
With the support of consenso, ERDINGER defined a total of eleven Data Mining & Profiling rules in the business partner area. These rules are used, e.g., to check the syntax of a tax number or to ensure that the postcode field is filled in in addition to the city, depending on whether a street name exists. In addition, several duplicate checks were conducted and further weak points in the context of the business partner master data creation were discussed and resolved.
The result is a graphical overview of the errors and a detailed list of the faulty objects for each rule – with individually defined messages. This list forms the basis for subsequent data cleansing.
A similar procedure was followed for the material master: ERDINGER and consenso jointly defined 34 rules in the DMP. For the purpose of subsequent cleansing, these rules check, e.g. whether the material name contains special characters or whether an EAN has been maintained for the material, including the base unit of measure. Master data in the material master was also checked for incorrect duplicates. And supported by a consenso Master Data Consultant, the master data creation process was optimised.
Following the pilot phase, the defined checks are now being carried out continuously during regular operation of the application, the results are forwarded to the business department and corrections are implemented by the master data team.
consenso has enabled the employees at ERDINGER to independently create additional rules and thus successively expand the checks, so that ERDINGER's data quality continuously improves and the relevant business processes are optimally supported.
consenso continues to support this process - with updates as well as optimisations and enhancements to the tools provided. Plans include an automated cleansing function as well as the individual creation and processing of specific objects using data governance structures.
The defined KPIs for measuring master data quality have already improved significantly. Project successfully completed.
We would like to thank all those involved in the ERDINGER project for their open and cooperative partnership and look forward to further joint activities.
Photo: Private brewery ERDINGER Weissbräu Werner Brombach GmbH, Erding
From our Competence Centre Digital Transformation
Categories: Digitalisation | Data Mining & Profiling | Artificial Intelligence | Master Data