Advertisement

You will be redirected to the page you want to view in  seconds.

Sharing the data wealth

Feb. 4, 2014 - 03:45AM   |  
By JEFF SORENSON   |   Comments
  • Filed Under

In the last three posts, this series has covered how commercial companies are realizing big data’s promise of limitless intelligence, the importance of building a smart data haystack, and ways to mine big data into actionable, predictive insights. This final post on big data addresses the last piece of the big data puzzle: data governance.

Previous Posts:

» Big Data: How to fulfill the promise of limitless intelligence
» Building the data haystack is key to Big Data
» How to turn too much data into just enough information

Leading companies use their data governance models to leverage data, technology, and analytic assets across the enterprise in order to share their data wealth throughout the enterprise in an effective and efficient manner. These best practice models offer the potential for doing the same for the DoD to help change the way it accomplishes its mission.

For the most part, leading companies begin building their data governance model through a framework that supports the enterprise, ensuring that data integrity is maintained and the data stays clean. It typically has the following key components:

Vision and Mission. Data governance’s vision and mission are aligned with the organization’s strategic objectives. In addition, all functions throughout the organization are aligned with this vision and mission so competing priorities of different departments can be minimized by using a common set of data.

Organization. The data governance organization covers IT and all business units. In addition, all roles and responsibilities are clearly defined and assigned to specific owners. Clear guidelines are established so everyone knows who has the final authority over writing, enhancing, controlling, and reading the data.

(Page 2 of 4)

Data Ownership. Data governance is determined along the whole value chain, and a data governance policy is implemented.

Process. The process of managing data quality and related metadata is clearly defined with IT and the business units involved in determining this definition.

Metrics. Data governance metrics are established. They also are measured regularly and used to drive data improvement efforts.

Technology and Tools. The technology and tools used for data governance are implemented for the data governance model and assessed regularly.

This framework has led most companies to select one of three data-governance models: decentralized services, embedded shared-services, or standalone shared-services. Each model will work, but companies may choose one versus the other depending on some specific set of criteria, which might include, for example, the enterprise’s maturity along the data spectrum, the need to embed the model within a particular business unit, or the fact that analytics is a core competency.

Decentralized Services

In this model, each business and function has its own analytics group. This encourages rapid decision-making and ensures that team members are close to the business, issues, and customers and that analysts can provide immediate tactical input. The model also offers an increased focus on the business or function, but limits the enterprise view, perhaps, undermining opportunity. Without a dedicated role for strategic planning or best-practice sharing, it can lead to duplicate resources and infrastructure.

Many consumer packaged-goods companies find this model works best. Often these companies have, for example, a separate consumer insights group in the marketing department, a separate analytics team in the supply chain, as well as a separate group of financial analysts within finance. This decentralization helps the individual functions focus on external trends as well as competitor, technical and manufacturing intelligence to provide better insights into performance.

(Page 3 of 4)

Embedded Shared-Services

This model is centralized under an existing function or business unit but serves the entire organization. Standardized processes and methodologies are developed under one roof, providing better coordination among departments. Functional expertise resides in the existing shared services. The model speeds execution and decision making, and its structure, support processes, and standards increase efficiency and IT expertise. However, the structure could lead to a less transparent allocation of analytic resources and could result in a focus on the priorities and agenda of the hosting business unit or function.

A top tier tech firm that delivers hardware and solution services relies on a hybrid of this model. Its offshore unit provides support to centralized and dispersed business-unit, function, and geographic teams. The company considers analytics a core capability to support its aggressive price proposition and to focus on direct marketing channels. Thus, its analytics team is mandated with relevant top- and bottom-line improvement objectives.

Standalone Shared-Services

This model is similar to the embedded-services model but exists outside any business or function. It reports directly to the executive level, ensuring faster, better organizational adoption. It elevates analytics to a vital core competency rather than restricting it to an enabling capability. Standardized processes and methodologies are developed through a cross-enterprise mandate. Independent viewpoints, best practices and insights are shared organization-wide. Some businesses or functions might feel the standalone group lacks functional expertise, or they might push back about the possible reallocation of resources.

A leading international hospitality company uses this model. The company’s centralized data and analytics unit provides coordinated information to improve decision making across the organization. This team includes a customer knowledge group responsible for analytics, data warehousing, and automation, and for developing predictive models to forecast customer loyalty and property utilization.

(Page 4 of 4)

Regardless of the model selected, leading companies have found that having a comprehensive data governance model enables the enterprise to accomplish a number of key actions that strengthen the sustainability of the big data program. First, it shares the data throughout the enterprise by making sure everybody who needs the data has the ability to read it. Second, it sets strict guidelines for who will have “write” privileges to enhance, augment, or update the data, thereby establishing the data’s integrity and underscoring the trust and confidence everybody has in the data. Third, it drives the enterprise’s adoption of big data and drives the change in the decision-making from a gut-driven culture to one that makes informed decisions through the use of big data and analytics.

Using its own set of criteria and proven commercial best practices , the DoD will be able to implement an optimal data governance model that supports the entire enterprise—a model that is singularly effective for the department and for each individual service’s combat and business operations. Perhaps it could be a hybrid of the decentralized-services and embedded shared-services models. The former would ensure that the Army, Navy, and Air Force each has direct and immediate control of its own operations, while the later would provide the Office of the Secretary of Defense’s guidance to ensure that all key functional elements are centralized.

More In C4ISR & Networks

Start your day with a roundup of top defense news.

Subscribe!

Subscribe!

Login to This Week's Digital Edition

Subscribe for Print or Digital delivery today!

Exclusive Events Coverage

In-depth news and multimedia coverage of industry trade shows and conferences.

TRADE SHOWS:

CONFERENCES:

Defensenews TV

  • Sign-up to receive weekly email updates about Vago's guests and the topics they will discuss.