Why you should federate your company's data

Article

Published August 2015

Download

DownloadArticle, 539 KB, Microsoft Word file

Data is typically managed by an organization in one of three ways: centralized, distributed, or federated. What does it mean for a business to have its data managed under each scenario, and why would you want to pick one over the other? This article explains how each data management model works and how the federated model combines the benefits of other models.

Centralized model: pros and cons

If an organization's assets and data are managed by one entity that sets standards for data uses and processes across the organization, the data management model is centralized. Traditionally, these responsibilities fall to the IT department, which is then held accountable for any breaches or data problems.

The main advantage of the centralized model is efficiency. Data can be aligned very easily, and very little time is spent reconciling disparities between different departments.

The two main disadvantages to this approach are that individual business units within the larger organization can feel disempowered and may complain about a lack of agility to solve problems for their particular business needs. According to Rajesh Nagpal, Director of Business Operations and Programs at Microsoft, these disadvantages are particularly apparent if the organization is growing or changing a lot, and if individual business units must be highly responsive to new environments.

Distributed model: pros and cons

At the other end of the spectrum, a distributed data management model means that each business unit sets its own standards for data uses and processes, which are adhered to within that group but not necessarily by any other group.

This model offers a business unit outstanding potential for empowerment and agility. For example, a start-up company that must prioritize sales targets or product launches may adopt a distributed model to reach short-term goals with the utmost speed.

But ultimately, this model will show its inefficiency, Rajesh explains, especially in large, global organizations. Why? Because integrating data from disparate groups makes basic enterprise tasks—like reporting earnings—very challenging. "What if the country (or region) name used by the subsidiary doesn't match the central repository? What if a single product is called by several different names across the company?" This can lead to revenues not aligning, and extra work to reconcile the differences.

Federated model: the best of both worlds

A federated data management model provides a good balance between the centralized and distributed approaches. An organization can specify exactly what needs to be centralized and can leave the rest to the discretion of business units.

At Microsoft, for example, data pertaining to geographies, revenues, products, employees, titles, job levels, and other enterprise-critical standards are centralized. But business units are empowered to make decisions appropriate to their own business needs in less-critical areas that are more local and less applicable across the whole enterprise. Examples include tax rules specific to regions, product discounts that they want to apply, and pay scale differences depending on the standard of living.

Another reason the federated approach has been successful at Microsoft is that it allows for customization and flexibility, even when there is a standard in place. Xbox, for example, may be sold as individual components (console, controller, Kinect), or in a bundle if a particular subsidiary has conducted marketing surveys indicating that their customers prefer it that way. But the product elements and pricing remain standardized across the company.

Another key benefit of moving to a federated model is that organizations can use it to build consensus among representatives of all the subsidiaries about exactly how the line should be drawn between centralized and distributed data management. Rajesh notes that this will produce a good, healthy debate, and that the organization can move forward based on the results of that debate.

"You don't decide where the line is drawn based on what some book says," said Rajesh. "It's the reality of the company that decides."

Conclusion

While some organizations will benefit most from the efficiency of centralized data management, and other organizations will opt for the agility of distributed data management, more and more organizations are now moving toward federated data management.

The flexibility and capability to negotiate where data decision-making authority resides, based upon the data and what makes the most sense for the organization as a whole, is a great solution for Microsoft. Microsoft continues to develop flexible solutions to meet the diverse data management needs of their customers.

Resources

Video: Top Reasons to Federate Your Company's Data

https://channel9.msdn.com/Shows/TechNet+Radio/Top-Reasons-to-Federate-Your-Companys-Data

For more information

For more information about Microsoft products or services, call the Microsoft Sales Information Center at (800) 426-9400. In Canada, call the Microsoft Canada Order Centre at (800) 933-4750. Outside the 50 United States and Canada, please contact your local Microsoft subsidiary. To access information via the web, go to:

http://www.microsoft.com

Microsoft IT Showcase

http://microsoft.com/ITShowcase

© 2015 Microsoft Corporation. All rights reserved. Microsoft and Windows are either registered trademarks or trademarks of Microsoft Corporation in the United States and/or other countries. The names of actual companies and products mentioned herein may be the trademarks of their respective owners. This document is for informational purposes only. MICROSOFT MAKES NO WARRANTIES, EXPRESS OR IMPLIED, IN THIS SUMMARY.