The Four Components of BI Governance

If you’re convinced that business intelligence is a good idea but you don’t know quite where to start, it’s perfectly understandable. No doubt about it—business intelligence (BI) is complicated. The objective of BI seems relatively straightforward—using data to advance your business. It sounds easy, yet very few companies reach their own BI maturity goal.

What makes BI implementation so challenging?

The technical aspect of BI presents some perplexity. In the 21st century, we have many tool choices, and decisions about which tool to use for which purpose can be a bit daunting. The good news is  that tools are maturing rapidly and their proponents are many. Whatever tool you choose, it can be tailored to your environment and can be expected to grow with you. Network bandwidth, server capacity, and other technical hurdles are important but not the driving force for creating a successful BI strategy.

The Key Is Governance

It’s all about BI governance—defining and implementing an infrastructure that will support enterprise goals. That infrastructure includes the hardware, software, staffing, and strategy needed to glean intelligence from data.

To understand the BI governance process, you need to understand the progression of business intelligence. Your company may be among many that approach BI in a problem/solution framework. When a problem is identified, the solution may be a tool, a data mart, or a data warehouse. You would then purchase the appropriate supplies, and voilà, you would have business intelligence. Unfortunately, that’s not how it works.

When you use the problem/solution framework, you omit a high-level view of all the available resources; you focus on only one a small area. Although you may have the best possible solution for one problem, that solution may generate more problems or impact previous solutions. At the very least, a total business picture, essential for good decision making, will not be possible. You’ll end up managing multiple tool sets, supporting diverse end users, increasing inconsistency, encouraging redundancy, and making poor business decisions.

In addition, dependencies on source systems can complicate manners. Training and end-user support also come into play, of course, and a framework for supporting BI efforts will provide an approach for managing BI projects and enhancements.

The benefits of a consistent methodology are many. If you can create a repeatable process, you can increase the success of a project and support continuous improvement while reducing the time to market.

BI governance builds in flexibility by creating robust processes capable of scaling to any size and scope, and all aspects related to the approach of BI efforts are clearly defined. Along with a mechanism to manage your BI strategy, BI governance provides measurements for gauging success. One non-tangible benefit: improved morale as the staff involved builds their knowledge base and shortens their learning curve.

We’ll explore several BI governance challenges.

Challenge #1: How Do You Get Your Arms Around Your Current BI Environment?

Ask yourself a few questions and you may make revealing discoveries. If you’ve implemented BI in a piecemeal fashion, you’ll probably have an environment that has multiple tools, duplicated data sets, differing business definitions to describe the data, and a wide variety of other challenges.

For example, one department counts widgets sold by market with a result of 5,001. Another department does the same count and the result is 6,052. Which is correct? What business rules were applied in reaching the total count—were returns included or excluded? Were different tools used—or different versions of the same tool? Did the data come from the same source? Who will reconcile the differences?

What if your department is looking for data that’s not available in your BI environment? It won’t matter if everyone is using the same tool—there’s no underlying data for the tool to produce. How can you find the data? Is it available in another department’s data mart? Can you access a portion of it, but not enough to be effective? Who will help you get to the data you need?

Challenge #2: How Do You Get Your Arms Around Your Current BI Strategy?

First, do you have a BI strategy? If so, is your strategy based on your department’s needs or the goals of the enterprise? If you’ve pinpointed multiple BI environments, you probably have multiple sponsors, and teams that make decisions based on local priorities. All the teams need to sublimate their local strategy for the overall enterprise goals. Even if this is your initial BI experience, a global strategy is vital to a successful implementation.

Each department must focus on the effort to develop a common BI strategy that will benefit everyone. You must integrate your methods. How will you handle prioritization? What will the BI roadmap look like? Is alignment with business goals automatically maintained within the plan? Who will focus the effort to support the enterprise goals?

Challenge #3: How Do You Get Your Arms Around BI Value?

You’re not alone if you have trouble measuring the value or success of your BI strategy. Without being able to definitively measure your success, you won’t be able to discover and act upon any unknown opportunities. A non-existent or poor BI strategy makes using data mining techniques and uncovering hidden potential nearly impossible.

Some of the measurement possibilities include establishing a return on investment, shortening the project lifecycle, reducing project costs, and validating marketing strategies. Are some of these already baselined? Has anyone collected all of the supply costs in one location? Can you measure intangible benefits like user satisfaction and morale?

BI Maturity

Now that you know where you stand, it’s time to take BI to the next level. Be honest, is your BI strategy focused on monitoring the state of the business? If so, it’s a good place to start. You know how many widgets you sold. What if you could gain vital pieces of information that would help you increase that sales number? That’s the next stage— using BI to increase the efficiency of your business.

In the next stage, you’ll utilize BI to your advantage. Wal-Mart became the leader in supply chain methodology by starting with a BI strategy that recognized opportunities with suppliers. Harrah’s Entertainment has become a leader in gaming due to their customer loyalty program— a program identified thanks to BI. As with any maturation process, BI maturity comes from knowing where you are and clearly defining where you want to go.

BI Governance Components

You can start to evolve your BI strategy by reviewing the challenges and the maturity level of your BI environment. While the technological aspects of BI certainly play a part in the value of BI, the business process side is equally important and often ignored. This is where BI governance comes in.

What is BI governance? It’s the process of defining and implementing an infrastructure that will support enterprise goals. Jointly owned by IT and its business partners, the process evolves the direction and the value of BI as a strategy.

There are four components to BI governance:

  • A BI governance committee
  • A framework for the BI lifecycle
  • An end-user support structure
  • A process to review BI programs

Governance Component #1: The BI Governance Committee

Remember asking, “Who will reconcile the differences?” That’s the purpose of the governance committee. It’s composed of people from all of the involved departments. You’ll want to include several as-needed resources from IT to act as subject-matter experts. Be sure to involve all levels of your business, from senior management to end users. The most important thing to remember when forming your governance committee is that it must accurately reflect the enterprise.

The role of the governance committee is vital to the success of your BI strategy. The governance committee keeps you aligned with business goals. Each representative is able to bring their team’s perspective and priorities to the table. Sharing those priorities is the first step in determining which project will add the most value. Redundancies and opportunities to combine efforts will become self-evident.

In the beginning, the governance committee’s focus is on setting enterprise-wide BI priorities, identifying BI impacts and opportunities, securing funding, and providing mentoring resources for new efforts. Ideally, you already have a steering committee or some other group that can leverage the existing processes of review, approval, and project initiation. This is ideal because it’s much easier to build on an existing process, avoiding additional bureaucracy. An existing team can show quick results by implementing small changes.

The main point of a BI governance committee is to provide greater visibility to all of the enterprise initiatives —aligning with overall business goals and providing value to each participating entity.

Governance Component #2: The BI Lifecycle

When you wondered, “Who will help you get to the data you need?” you might have been thinking about a specific person. With a BI lifecycle, you can get the data you need by developing an infrastructure to support BI needs. We’re back to processes—as your environment changes, your projects can vary in scope and size from enhancements to a major corporate data integration.

What is a BI lifecycle? It’s a framework that includes all of the processes needed to implement a project of any size—data integration; new analytical structures for the data warehouse; new views, universes, cubes within the BI tool suite; an entirely new BI tool; data education for end users; tool training for end users; change management for existing data structures; and others. You may find that the BI lifecycle plays a small part in much larger enterprise project that includes cross-system impacts.

What does the BI lifecycle look like? The BI lifecycle is a consistent framework that supports BI efforts end to end and is agreed to by IT and the business users. Different sections of the framework can be used separately or in combination, depending on the scope of the project. The intention is to have a framework handle the different aspects of BI regardless of order or dependency. The framework has six sections:

  • Tool selection
  • Data integration
  • Analytics
  • Custom BI tool work
  • User acceptance
  • Training

Tool Selection

End users often view BI as the end-user tool. This tool is the part that end users see and touch; it is their interface. With that in mind, choosing the correct BI tool for your project is a key success factor. Obviously the end-user requirements are going to drive selection, but other factors should also be considered. The end-user requirements should define how many users need to be supported, what types of users (ad hoc analysis, freehand SQL, data reviewer), data presentation needs (drill down, OLAP, MOLAP), etc. You may already have an appropriate tool or you may need to find a new one. Where possible, select a tool within your existing tool suite—it will save time and money.

Key: Determine the BI tool that is right for the job.

Data Integration

If you have chosen a BI tool, great! Now you need to get access to the data. Initially, you’ll need to understand the source of the data. Your BI effort can stagnate if the data is not easily identified, captured, and integrated for easy access.

Can your BI tool access the data in its current form? If yes, then focus on the condition of the data. Is the data of high quality, available at the right frequency, and complete? What if the data is not available in a form that can be accessed by your BI tool?

A data acquisition project may be necessary to obtain the data. As part of that project (or in a separate project), you’ll want to integrate the data with whatever other data is currently available. If you haven’t yet created a central repository for the data, you may need to invest in the creation of a data mart or data warehouse.

Key: Ensure data is available, complete, and of value.

Analytics

After you have selected a tool and your data is accessible, you may need to marry the two to accomplish true BI. If a requirement exists for drill-down capabilities with certain dimensions and metrics, a star schema may be required. If this approach does not support the requirements, denormalization of the data could be needed. Analytical structures to support performance or for ease of use within the BI tool could also be the answer. Aggregates and the joins of multiple data sources may be required. Views to correctly present the data may also be
needed.

The essence of analytics revolves around ensuring data presentation and easy access, thereby increasing the value of the project. Your BI strategy loses credibility when the data is accessible but not in a workable format.

Key: Ensure data presentation and access.

Custom BI Tool Work

Now you have the basics—tool, data, and analytics—you’ll find that BI tools often need customization. A view of the data specifically developed for the tool will allow your end
users to easily access the data that is meaningful to them. Other custom work might include cube structures for OLAP capabilities or a suite of canned reports/queries that an end user could easily
reproduce on demand. End-user communities have different skill levels and needs which will determine the level of customization needed.

Key: Ensure the BI tool provides the correct functionality and capabilities.

User Acceptance

End-user acceptance of the finished BI product is an iterative process. You’ll want to engage the end users early in the process, gathering requirements and assessing skill levels. Involve the end users in each step of the process, explaining challenges and negotiating solutions. Without end-user agreement to, and sign-off of, both data integration and presentation, success is elusive. Make the enduser involvement as integral as the data and the tool. When the project is ready for a final review, the end users are effective testers who understand what to expect from the data and from the tool functionality.

Key: Ensure the end users test and approve the final BI product.

Training

There are multiple levels of training within a solid BI lifecycle. Ongoing training for the governance committee SMEs is essential to increasing the speed and effectiveness of each new effort. Even more importantly, the end users will need training on both the data and the tool to provide meaningful results.

Approaches vary depending on your environment—just remember to consider each individual’s expertise and intent. A training session that incorporates the why behind the skill is more beneficial than rote learning. Merging tool functionality with data basics is a minimum; increase the level of data training for your end-user analysts who produce reports for others. Tailor the training to the audience.

Key: Ensure that the end user is able to utilize the BI tool to provide value.

Governance Component #3: End-User Support Structure

You might be thinking, “Who will focus the effort to support the enterprise goals?” End-user support focuses on providing assistance and education from a technical, functional, and data standpoint, always keeping the enterprise goals in mind. Part of the challenge in providing valuable end-user support relates to the combination of technical expertise with business acumen. Here is another opportunity for IT to partner with business SMEs. If the end users don’t understand the data, they cannot gain business intelligence.

You’ll derive many benefits from an end-user support structure, including a better understanding of the business goals by both end users and their IT partners. Supporting end users effectively means that you can address the different levels of expertise while continuing to increase value. End-user feedback can help you continuously improve the processes, making them more easily repeatable and much more cost effective. Don’t underestimate the end-user community—they will become empowered to drive the evolution of your BI strategy.

 What does good end-user support look like? It’s composed of three key elements supported individually and collectively: data, tools, and business expertise. Having an end-user support structure that focuses on these three areas guarantees a knowledgeable and empowered BI community.

Data

Supporting data education is necessary to help your end users derive value from the BI environment. As data is introduced into the BI environment, training sessions, data forums, and metadata need to be made available to the BI community. Provide assistance to make use of the new data. Be sure to offer good metadata (a data dictionary, including definitions and data sources), business logic (tie the data back to the original business goal), key dimensions and metrics (including how to use them for best results), and data availability (where and when new data can be found).

Tools

BI tool support should focus on the functionality of the selected tool as it relates to the project. This is one of the easiest areas in which to quickly yield positive results. When the BI governance committee recommends a tool for a specific project, one of the factors in making that recommendation may have been how knowledgeable and comfortable the end users are with the tool. Tool support will require introductory training, mentors or SMEs, and ongoing education. Your BI lifecycle needs to include some additional tool support activities—help the end user choose the right tool; monitor data availability and quality; segment the user community to support each group appropriately; establish service level agreements for response times and issue resolution; and communicate a clear escalation path and contacts.

Business Expertise

An effective BI strategy utilizes business expertise about the data, the tools, and the business goals. If an end user executes a query successfully but the result set is not meaningful, no value is realized. An end user who understands the data but is unable to present it in a form that will enable complete analysis also adds no value.

In this area of end-user support, the focus is on BI mentorship. BI mentorship involves super users—financial, marketing, or operational analysts who understand the data and the BI tool and have been successful in using BI to provide value. Data stewards and architects could also assist here. Your mentors will provide leadership when it comes to tool selection. Mentors can help end users understand key dimensions and metrics.

BI mentors are business partners with a unique contribution to the BI strategy. Their contribution includes (but is not limited to) helping new users get started, identifying and developing data mining opportunities, pinpointing data trends, and staying focused on the business goals.

Governance Component #4: BI Program Review

Now that your BI strategy and lifecycle are in place, how will you measure success? One of the most challenging parts of BI implementation is the review—measuring success and understanding value. BI is not always quantifiable and its role as an enabler does not easily fit into an ROI calculation. Even when you do have quantifiable metrics, your next step may not be clear. The BI review can help you to determine your BI maturity level, identify meaningful metrics, and capture opportunities.

If you go back to the initial process stage, assessing your BI maturity level is the best place to start. Can you identify the ROI of BI initiatives? Do you know the capabilities of the BI environment? Is there a clear focus on BI enterprisewide? What has BI enabled—a better business understanding, a clearer direction, optimized relationships, or something else?

The best BI reviews periodically revisit the initial assessment. Use them as your baseline to determine if your BI environment is maturing and headed in the right direction.

BI measurements or metrics can be established before and after project implementations. One important project metric is ROI. The projected ROI establishes whether the project is worth initiating, the actual ROI determines if real value was derived. ROI on BI efforts is ambiguous if not specifically weighted in the projected ROI calculation.

An example of an ROI metric might work like this: Through BI, it is discovered that customer representatives are crediting customers too much, seemingly without guidelines. These excessive credits cause millions in losses every month. In order to bring this under control, a project is proposed to enhance a BI environment to provide customer care supervisors the ability to determine how much a customer could be credited based on the customer’s value to the company. The project is successful and reduces the erroneous credits and saves $1 million per month. When the actual ROI is calculated, part of it can be directly linked to the BI segment of the project. In this example, the investment segment of the project includes all of the BI enhancements (hardware, software, IT labor), process improvements, and care labor.

Other measurements are important to your BI evaluation. One centers around the new opportunities gained from data mining. Data mining may help you find opportunities to avoid excessive costs, increase revenue growth, lower customer churn, or increase your customer base.

Do your projects have a shorter lifecycle? With the BI governance committee’s assistance, have you noticed that decisions are made more quickly, requirements are met or exceeded, and capabilities are better understood?

Sometimes your end users are a key review measure. Has the demand for data access increased? You may be receiving multiple training requests or enhancement requests. Perhaps the skill level of your user base has increased, making project implementation faster and more effective. Your various departments may notice increased productivity. Key business questions are answered in a timely fashion. New opportunities are identified through end user analysis.

Comparing budget forecasts to actual spending is a vital measurement. As the BI governance committee develops, the budget forecasts are much closer to costs. Budget forecasts decrease as BI maturity increases, saving time and reducing expenses—thanks to employing repeatable processes, enhanced end user capabilities, and use of a stable tool suite.

Lastly, a BI review metric of great importance is alignment with business goals. Is the BI governance committee working effectively? If the right priorities were determined, you’ve probably avoided some unpleasant surprises. You can reconcile the perceived value of the entire BI strategy with the actual success of the entire enterprise.

Risks and Challenges

BI governance is about process and alignment. Process can be cumbersome, inflexible, and perceived as additional bureaucracy. Alignment can be difficult to achieve due to organizational structure or politics. A framework is a guideline that is comprehensive enough to cover an area end to end, but sufficiently flexible to be applied where it makes sense.

For example, consider one section of the BI governance framework (such as the BI lifecycle) and use it as a starting point. You may be able to generate enthusiasm to begin with because a process for supporting IT projects is widely accepted as being valuable. Remember that getting started will require a specific focus. Change in any environment is required but not always easy. You can improve your chance of success by gaining consensus from all functional groups. Each business unit needs to share its authority and compromise on priorities.

Conclusion

BI governance is a framework that helps identify, deliver, and maintain the BI strategy. It includes:

  • BI governance committee (for project direction, alignment, and prioritization)
  • BI lifecycle (to ensure consistent project delivery and meet end-user expectations)
  • End-user support (ongoing improvements to user effectiveness and empowerment)
  • BI review (evaluating maturity levels and tracking metrics)

As you evaluate your environment, focus on where you can add value quickly and easily. Then, as you make progress and gain small successes, your BI strategy will begin to mature.

Share this post

scroll to top