Operational business intelligence (BI), active data warehousing, business activity monitoring—10 years ago these concepts were visionary ideas of what might be; today they are a reality. Maturing technology and increasingly sophisticated applications are prompting CXOs to view their data warehouses as instrumental tools that can be used to chart the future, gain competitive advantage and protect their company's position in an uncertain economy.

Despite the turbulence in world markets today, organizations trying to capitalize on data warehouse potential continue to spend significant amounts on their data warehouses. Research referenced in the January 2007 report, "Magic Quadrant for Business Intelligence Platforms, 1Q07" by Kurt Schlegel, Bill Hostmann and Andreas Bitterer, indicates that BI software revenue will grow at a rate of 9.5% through 2010 and that deployments will expand beyond just query and reporting to include scorecards, dashboards and predictive models.

Data warehousing applications that reach beyond the traditional touch all parts of an organization, including the front-line staff. Visibility is heightened, analysis results are woven into the operating processes of the organization, and the data warehouse is elevated to a mission-critical status. As the stakes go up for the data warehouse, the IT implementation group experiences a proportional increase in pressure. Understanding business processes, keeping pace with increasing demand and responding to issues in real time are just some of the added hurdles they may face. Tuning the data warehouse team to meet these challenges is paramount if an organization is to realize the potential of today's data warehouse. But leading organizations are moving beyond a simple tuneup; instead, the teams are transforming into BI Center of Excellence (COE) groups whose reach extends far beyond the traditional.

BI COE: the key to value
Intelligent Solutions Inc. defines the BI COE as "the set of people, processes and technologies for promoting collaboration and the application of best practices; including the development of templates, documentation of best practice and the development of education."

The COE consists of one to four formal cross-functional teams that span business and technology. Its processes facilitate the development of data warehouse priorities and assist in applying insights to the business. With a COE in place, the organization will leverage its investment in the data warehouse in several capacities and will be equipped to treat data as a strategic asset by establishing quality metrics, standards and data stewardship. The COE assists in reducing total cost of ownership (TCO) by promoting reuse of technology, data models and best-practice templates. It also assures that data warehouse projects are aligned with strategic business objectives so that the data warehouse delivers ongoing business value.

Figure 1, below, illustrates the areas that can be included in a COE, with each one sharing a governance function. Getting Data In (GDI) and Getting Information Out (GIO) are the two primary teams in a typical COE group. They are formal bodies made up of various IT and business users and are responsible for establishing, enhancing and maintaining the data warehouse. Some organizations also split Quality Management into one team and Application and Use into another team. Other organizations incorporate Quality Management into the GDI team and Application and Use into the GIO team.

The number of teams in a COE group is not crucial; it is more important that the group has clearly defined roles and responsibilities, that it has the authority to carry out its charter and enforce compliance of its policies, and that critical tasks are covered.

Getting Data In
GDI is a technical team in the COE group that focuses on data integration and data quality. Generally a centrally located team (see left side of figure 2 below), the GDI reports to the IT department and includes a program manager, a BI architect, an extract, transform and load (ETL) architect, ETL developers, business analysts, data analysts and occasionally a data warehouse database administrator.

GDI members are responsible for getting data from the operational environment into the data warehouse and performing all of the data acquisition activities. Besides creating and maintaining the data warehouse, staging area databases and the technical metadata, the team ensures that all data in the data warehouse has common definitions and uses across the organization. Finally, it monitors the data warehouse so it consistently meets or exceeds the service level agreements developed with the business community.

The GDI team incorporates best-practice standards that:

	Determine a standard set of data integration and data quality tools
	Establish a technical metadata repository
	Communicate to the GIO team what data is available
	Coordinate formalized and efficient ways and sources of exchanging information, including points of contact and go-to personnel
	Develop processes to handle accelerated requirements
	Leverage reuse of components, including data mappings, ETL code and enterprise data models
	Implement data quality processes and metrics; define responses to quality problems
	Work with the business community to identify data stewards

enlarge The COE can cover a broad range of functions, including getting data into the business intelligence (BI) environment, getting information out of the BI environment, quality management, and the application and use of BI. Companies that want their COE to cover all of these activities can build one team or split the activities into several teams as indicated above, but they should still consider a single governance mechanism that is shared across all teams.

Getting Information Out
GIO is a business-oriented group that acts as a bridge between the business community using the data warehouse and the IT developers creating it. The right side of figure 2 (below) illustrates a typical GIO structure. GIO team members focus on understanding business needs and translating these needs into specifications for reports, dashboards and other data delivery mechanisms. As a team, they ensure that new data requirements and data quality problems are communicated to the GDI team and the business community.

This group can be decentralized to report to individual business lines as long as a central component is responsible for developing templates and standards, for coordinating the development and use of end-user applications, and for identifying overlapping applications for combination or retirement.

The decentralized group typically includes business analysts, delivery developers and tool configuration specialists. The centralized component can be incorporated into the GDI team and includes the program manager and a GIO specialist with the skills to develop best-practice templates and recognize overlapping applications. Like the GDI team, the GIO follows a set of best-practice procedures that:

	Establish expertise in how the business interfaces with the data warehouse environment
	Maintain expertise about the business's usage patterns and procedures
	Determine a standard set of data access technologies, such as reporting and querying, online analytical processing, mining, exploration and predictive analytics, and appropriately match users to these technologies
	Establish the business metadata repository, which may be in the same repository as GDI technical metadata
	Provide analysis expertise and training to promote application and use of BI across the business community
	Develop the training and education programs to further the business community's understanding of data warehousing concepts and of the data available in the data warehouse

Governance: the superglue
It is essential to establish governance when building a COE. Without governance the COE can become aimless and ineffective. A cross-functional body with executive-level membership such as a governance team is required to set and promote the COE scope, provide funding and resources, and visibly bestow authority on the COE to carry out its responsibilities. Governance should:

	Establish data warehouse priorities and align projects with corporate priorities and objectives
	Sign off on business rules and business policy changes
	Communicate business rules and business policy changes to the organization
	Arbitrate conflicts across business units
	Provide guidance on future business directions and initiatives
	Obtain funding and resources for data warehouse efforts
	Facilitate compliance with standards, policies and process changes

enlarge The above flow charts illustrate possible organization structures for the Getting Data In (GDI) and Getting Information Out (GIO) teams. The GDI team is frequently centralized and typically includes the resources displayed in the section outlined in red. The GIO team can be decentralized with resources reporting into various business units, but it should matrix (as illustrated by the red outline) into a single program management director to ensure coordination.

With strong governance and authority established, the COE group can efficiently and effectively carry out its mandate, provide templates and promote best practices, and ensure a collaborative environment. It is the governance that will allow the COE to help the organization reap the benefits from its data asset, minimize the TCO of the data warehouse and derive business value from the data warehouse.

Achieving excellence
As data warehouse applications become closely linked to the operations of the organization, data warehousing moves from touching an isolated few users to touching the masses. The potential benefits increase exponentially. Scorecards and dashboards provide a real-time view into the organization's performance.

Executives, managers and front-line staff have a vast amount of information at their fingertips to aid in decision making, and predictive models help chart the future and shape strategy. Reaping these benefits requires a new type of BI team: one that understands the business, can meet rapidly increasing demands, and can ensure continuing high-quality data. The BI COE is that team, and leading organizations are transforming their traditional data warehouse teams into COE groups that can transition the data warehouse into the future. T

Lisa Loftis is a senior vice president for Intelligent Solutions and a customer relationship management and BI expert. She is also a speaker and co-author of the book "Building the Customer-Centric Enterprise."

Photo illustration by Randall Nelson

Teradata Magazine-March 2008