BACKGROUND

The IT Governance process worked very well for individual projects or initiatives, but it did not focus on making sure particular programs, such as Business Intelligence, were successful across the Enterprise. In order to close this gap, leading-edge companies decided to take the Governance concept down a level and focus entirely on the programs, thus giving born to the BI Governance concept. Some of these organizations defined BI Governance as the process they followed to prioritize BI requests along different criteria such as: Project ROI, Organizational Budget, Expertise of the team, People Availability, Infrastructure capacity and Organizational Politics.

This article will focus on defining BI governance, and detailing a good BI Governance process must go beyond the basics of approving and prioritizing initiatives.

DEFINING BI GOVERNANCE

a) As a resource rationalization exercise. This is the traditional definition of BI Governance, a prioritization mechanism by which projects can be approved, rejected and sequenced based on specific criteria. Many companies today have some kind of process to prioritize BI requests; however most of them still rely on subjective factors to determine how a project should move forward.

1 >> 2 >> 3 >> 4 >> 5 >> 6 >> 7

b) As a series of guidelines/rules /recommendations. This is a relatively new way of defining BI Governance. Traditionally IT has been solely responsible for defining the Architecture, Standards and Best Practices to follow in Business Intelligence, however as business starts realizing the significant impact that these decisions may have in BI Projects there has been a trend to make these topics part of the BI Governance process.

c) As the definition of roles and responsibilities for both IT and Business stakeholders. As Business Intelligence projects are by nature highly complex, establishing the proper interaction and outlining areas of responsibility between and IT and the Business becomes critical for a project to succeed.

Figure 1 - The three dimensions to BI Governance

UNDERSTANDING THE NEED FOR BI GOVERNANCE

While it is true that a good management team should have already a general understanding of the priorities of the organization, and how the company strategy(ies) translate into a series of BI initiatives, establishing BI Governance provides a framework by which the proper resources can be aligned to the Business Priorities. In order for the initiatives to be successful they need to be executable. Only by aligning people and resources to a particular project and documenting a project plan is that initiatives become ready to be executed.

Furthermore, a good, solid BI Governance process will establish proper Change Management (CM) and Training policies that facilitate the user adoption process and promote the overall use of BI minimizing the amount of fear and resistance to new technologies from the business users.

The proper BI Governance process will also drive the infrastructure and technology decisions. As both perspectives, IT and Business, are considered when making these decisions the likelihood of choosing the right vendor/platform significantly increase; thus reducing the risk that one of the parties chooses a platform without considering the implications that it will have over the other party (Figure 2).

Figure 2 - Business and IT collaborate to align the BI resources to the organization priorities

The fist step to establish an organizational structure of the member of the BI Governance is identifying the stakeholders within the organization. A stakeholder can be defined as someone who has some “skin the game”. Stakeholders usually come three areas:

a) Business: Every business area/department should nominate who their representative is going to be for the BI Governance committee. This will enable equal representation from business across the organization, and thus provide the right forum to make decisions that might impact the revenue and thus budget for a particular business unit. The business area representatives will be responsible for sponsoring particular projects for their home units, explaining to the committee the benefits of the initiatives. At the same they outline a high level impact and change management plan that allows the committee to understand the project being proposed from many possible angles. Part of their responsibilities will include identifying people in their functional areas to support the Data Modeling, Data Integration and BI-Front End efforts from a business perspective.

1 >> 2 >> 3 >> 4 >> 5 >> 6 >> 7

b) IT: IT provides the back office support for the committee. It is its responsibility to provide a Project Manager, Data Modeler, Data Integration Lead, BI Front-End Lead, and Enterprise Architect and Trainer Lead. The project manager will be the facilitator for the committee, making sure the meetings take place at the right place, right time and with the right people. The Data Modeler will be in charge of creating a data model that supports the business needs. It will be the Data Integration Lead’s responsibility to implement the appropriate data integration framework to populate the data model. The BI Front-End Lead will interact heavily with the business users to define the report and standards to follow. The Enterprise Architect will keep the solution in check with Enterprise Standards. The Trainer Lead will be responsible for creating the material that will be used to train the trainers, and thus create an infrastructure to train the user universe in a quick and effective manner.

c) CFO Office: As BI projects are considered strategic in nature and depending on the implementation cost can easily escalate to millions of dollars, it is recommended to have a direct link to the CFO in the BI Governance committee. This will facilitate the prioritization of projects, based on alignment to the corporate strategy (ies) and Return On Investment (ROI), at the same time that provides an opportunity to the team to justify the infrastructure and development costs directly to the ultimate approver.

As outline above, every group of stakeholders needs to play a role and commit to some responsibilities in order to establish BI Governance. The picture below outlines how the different groups interact through the life cycle of a BI initiative (Figure 3).

Figure 3 - Stakeholder’s major roles and responsibilities through the life cycle of a BI initiative

1 >> 2 >> 3 >> 4 >> 5 >> 6 >> 7

BUSINESS ROLES AND RESPONSIBILITIES IN BI GOVERNANCE

In a traditional BI Governance, the job of a business area representative is “complete” when one of their initiatives is approved by the committee. However this is rapidly changing to accommodate other responsibilities like participating in data modeling, data integration and the users’ BI experience.

Business user participation during the data modeling exercise is highly requested as they are the ones who know the operation of the business; they can help the data modeler understand the business processes that are established and identify current and future business scenarios.

The business users’ involvement in data integration is critical to the success of the initiatives. Their knowledge of the data and the business processes makes them ideal candidates to become or designate the data stewards. A data steward is the person responsible to define the data loading exception scenarios and approve the recovery strategies proposed by IT. Furthermore as the business area representatives they have ownership of the Data Quality criteria and policies for the initiative.

Furthermore, as the business area representatives speak for everyone in their areas, it is their responsibility to take ownership of the user experience. The definition of naming standards, folder structures and the report organization in general impacts this user experience. As subject matter experts, they also own the metadata and object descriptions. The establishment of guidelines to govern the creation of additional end user objects (e.g. reports, metrics, etc) also belongs to them as it significantly contributes to define the user experience.

1 >> 2 >> 3 >> 4 >> 5 >> 6 >> 7

IT ROLES AND RESPONSIBILITIES IN BI GOVERNANCE

Providing the right leadership for project execution is still critical, IT and the Project Management Office (PMO) are in a unique position in the organization to provide thought leadership and a strong methodology to run the projects. The business users heavily rely on IT to structure work streams along timelines and resources to achieve a committed dateline.

The BI Front-end work has seen a significant change over the last couple of years, before IT used to set the direction and the business was expected to follow. Given that this approach did not take into consideration many of the business needs, and it did not encompass a seamlessly defined user experience across initiatives, we are seeing a trend where the business users representatives are stepping up to the plate to participate in the decision making. While it is now a more collaborative effort that it used to, IT is still on the hook for critical decisions such as: selecting the proper tools, defining the BI architecture, establishing the proper infrastructure, implementing the standards as defined by the BI Governance Committee, Training the users and last but not least, coordinating support and maintenance across the business areas.

Data Modeling, which traditionally was done behind the scenes is now being introduced as part of the BI Governance model to ensure the business users will participate in the data modeling sessions to capture the right elements and accurately reflect business processes.

Data Integration, also known as data plumbing, was not something the business user had a lot of visibility into it before. Imagine the surprise of a business user when he was contacted by IT to be designated as a data steward; with no context or understanding of the activities that they were going to commit to. No wonder IT did not get much traction from the business users and had to resort to escalation processes to solve particular data challenges. In the new model IT partners with business to ensure success of the initiatives. In addition to providing the proper infrastructure, IT owns the Tool selection process, the data Integration Architecture and the implementation of data loading and data quality policies. Not to mention that IT has the responsibility to interface with the Data Stewards to resolve data situations and correct data issues as soon as possible.

1 >> 2 >> 3 >> 4 >> 5 >> 6 >> 7

An important element that was not present in BI Governance before was the Enterprise Architecture representative; it is the responsibility of the Enterprise Architect to define how the BI architecture will fit within the Enterprise. These decisions might materially impact the cost and performance of BI initiatives; furthermore if the Enterprise Architect is not aligned he/she can delay the implementation of a project to the point of affecting the business value of the project.

Training is another key role that traditionally has been bounced between IT and the Business, in the past nobody wanted to take ownership of training as it implied a significant invest in time to prepare the material and probably even more to teach it. As the organization evolves, developing the training material and training the trainer is starting to become IT’s responsibility, while the business areas provide end trainers who then train the end users.

IMPLEMENTING BI GOVERNANCE, WHAT ARE THE FIRST STEPS?

Once the stakeholders are identified and executive sponsorship is present it will be the time to bring the people together. A mission statement and specific strategies/goals will have to be crafted and approved during the first session. During this meeting sub-committees might be defined to discuss specific topics that do not require getting the whole group together. If possible, it is also suggested to schedule at least the next four meetings of the group so everyone knows the dates and commits to them.

REFERENCES

1. The Data Warehouse Lifecycle Toolkit: Expert Methods for Designing, Developing, and Deploying Data, Ralph Kimball, John Wiley & Sons, 1998.
2. Building the Data Warehouse, W. H. Inmon, John Wiley & Sons, 2002

About the Author