Rewired, 2nd Edition

CHAPTER 27Organize to get the most from your data

Like many things we’ve covered in this book, successfully enabling your data is less a technical challenge and more a people one. Or more specifically, an organizational one.

The questions are myriad and complicated. Who should own the data? Should we have a data organization? How should it link with the business? And with IT? Who should own data engineering? Should data privacy and compliance be separated?

These questions are often hard to answer because it comes down to people agreeing on a common vision for how data should work and who has responsibility for it. Further complicating the issue is that data is everywhere, it’s a source of power in an organization, and so many stakeholders need to get aligned to make the model work in practice.

A centralized or federated data org model?

This is an age-old question, with the pendulum swinging to and fro on it. Some companies have primarily centralized models, with teams of data staff that serve the whole enterprise. Others have a primarily decentralized model, with each business unit and function developing capabilities to meet its own needs.

Both models can work on a limited basis but are not generally responsive enough or able to scale to meet broader business needs.

Leading companies deploy a federated model with a strong central data management function that sets policies and standards, and provides support and oversight. Business units and functions then perform data ...