Using Two Different Governance Models in the Same Organization
Content Summary
In my report, I explored two distinct governance models: centralized governance and local governance, focusing on their unique characteristics, strengths, and weaknesses. Centralized governance, dealing with global master data, ensures high data quality due to adherence to global standards, but incurs high maintenance costs due to its extensive use across various industries. In contrast, local governance, which manages application data specific to individual businesses, offers greater adaptability and innovation opportunities within those businesses. However, this model faces challenges with data quality, including redundancy and inaccuracy. This distinction forms the basis for understanding the application and management of these models in data governance.
I demonstrated the practical application of these governance models using the example of an online retail store. Centralized governance is suitable for managing master data like user addresses and phone numbers, ensuring consistency across different industries. Local governance, on the other hand, is apt for application data management, handling data like a user's browsing history which is unique to a business and not shared externally. This dual-model approach in a single organization allows for the effective management of different data types, balancing the need for standardization and innovation.
However, implementing both governance models in one organization presents challenges. The primary challenge is managing two different systems simultaneously, which can lead to increased labor costs and complexities in determining which data should be governed by which model. Despite these challenges, the combined use of centralized and local governance models offers a balanced approach to data management, catering to the diverse needs of an organization's data landscape. This strategy allows for efficient management of various data types, each with its own change frequency and governance requirements.