Managing Data Integrity as Part of Master Data Management

Abstract

The present thesis concentrates on data cleaning actions due to a Master Data Management(MDM) Program in a case company. The goal of the MDM project is to streamline the data in order to offer high quality master data which will support business processes.

The purpose of this research was to provide guidelines and best practices on how to achieve and maintain high level integrity of customer data. In this research, first the theories around MDM, data quality management and data cleaning were studied, followed by carrying out a current state analysis, and then providing practical guidelines based on the existing literature and experiences with data cleaning. This research answers to the question how to ensure data integrity in the future. In addition, it provides understanding on how data cleaning can be executed manually. The research results can be used for future data cleaning projects inside and outside the case company. This research was conducted with using qualitative approach as the prime emphasis was on gaining understanding, and the data was collected with action research.

According to the research, good data quality has many benefits. It makes the data more trustworthy and decreases the time that the user has to spend searching, checking and correcting the data. Quality problems in the core system are most likely also transmitted to the business target systems. Good data integrity results from valid, accurate and consistent data. When data integrity is high, data follows business rules and it is timely, and satisfies the business needs. Common business rules and data quality rules ensure that the data is entered similarly throughout the organization in an accurate way. Both business rules and data quality rules were studied and reported for the case company during the project and they are introduced in this Thesis. Defining and communicating the rules is usually not enough, and data controlling is necessary. In this research means for data controlling are introduced.

Finally, guidelines for data cleaning project are provided in the thesis. With this research it is proven that the data cleaning project can be executed manually without external cleaning services, but the most beneficial combination of manual and system cleaning would be an interesting topic for future studies.