This way you will push through SISO-principle
Poor data quality in IT systems cost money and trust. To break through the vicious circle of „SISO“ („shit in – shit out“) you require a systematic and iterative approach as well as collaboration between business data experts and technical data scientists from IT.
With our approach of a continuous data quality circle with engaging measures we are able to achieve short-term effects to repair acute data inconsistencies as well as a sustainable and high quality level of critical business data:
- During Analysis phase we get an overview or the state of master and operational data (usually with already existing analysis tools). For that purpose data from different source systems will be extracted and validated. We will identify causes for poor data quality, e.g. deficits during data entry and processing, inconsistencies in data attributes, violation of rules for data entry, interface errors, etc.).
- During Planning phase we elaborate „fire-fighting measures“ commonly with the business lines involved – whenever risk is prevalent that false data or inconsistencies might cause consecutive faults or bad public image for the company. Furthermore, we start developing a concept with mid-term and long-term organizational and technical measures to improve data quality level sustainably. This will be continuously optimized in the course of the following phases.
- During Cleansing phase we coordinate execution of planned and agreed measures which shall remove deficits and optimize data quality.
- During Monitoring phase we establish processes and „quality gates“ which ensure continuous monitoring and documentation of data quality corporate-wide. Scorecards and cockpits help recognizing and assessing quality of data base at a glance.
Results of Monitoring phase are the base for further analysis of data quality and close the circle.
Project Reference: Cleansing of complex data constructs in live system