Q: How to measure the impact of indirect data products?

A: Even indirect data products contribute to the overall business value chain. It's essential to establish a system that traces and attributes value throughout the data product tree, ensuring even indirect contributions are recognized.

Q: How to validate assumptions made when defining connections in data value chains?

A: Start by defining clear criteria for assessing the value and feasibility of opportunities. Use a scoring system to rank opportunities and ensure transparency in decision-making. Additionally, actively engage with business stakeholders to validate assumptions and ensure alignment.

Q: How to decide between functional, product-based, and matrix organizational structures?

A: Consider factors like the frequency of product landscape changes, skill and resource needs, willingness to reorganize, the importance of cross-disciplinary teams, and the ability to manage a matrix effectively.

Q: When to consider a product organization?

A: A product organization is generally recommended unless you are solely a shared service center or a platform provider. However, even in these cases, a product-centric mindset can be beneficial.

Q: How to know when it's time to move away from a functional structure?

A: Organizational theory suggests a span of control limit of 70-80 people for a functional structure. In practice, challenges like misaligned priorities and communication issues may indicate the need for a change.

Q: Who is responsible for communicating changes and ensuring data quality in a data contract framework?

A: The team that owns the data product is ultimately responsible. However, automation and data lineage tools can help facilitate communication and proactively notify downstream consumers about potential impacts.

Q: How to address data quality concerns in legacy systems like SAP and Salesforce?

A: Leverage the exposed CI/CD workflows in these systems to insert circuit breakers and enforce data contracts, preventing incompatible changes and ensuring data quality.

Q: Is there a generally recommended design option?

There is no one-size-fits-all solution. Organizations should carefully evaluate their needs, prioritize business value, and adapt their strategies and structures to support the evolving data landscape.