Business Intelligence Strategies: Keys to Success
With business intelligence strategies getting so much attention, why is there a problem with delivering on the promise of BI projects?
Most of these “strategies” developed are really more tactical than anything and are very prescriptive. Companies typically embark on such “strategic” exercises because of a central event, such as:
•    Changes in leadership,
•    Enterprise cost reduction initiatives,
•    M&A activity, or
•    Implementation of a new enterprise resource planning system.

This list is not exhaustive and was not intended to be. In any case, the driving event usually ends up being the central theme in any subsequent strategy that’s developed. The environment in which businesses operate is so dynamic that any strategy driven by a single perspective does not stand the test of time. Even when organizations have the best of intentions, all the executive support they need and commit the necessary resources to these endeavors, they often end up with misalignments with projects and overall business strategies.

So, if anything, the central theme of any strategic framework has to be flexibility. A strategic framework guided by core principles as opposed to a tactical response to a situation or an event is called for. It is essential to properly identify the core principles to craft your systems around.

As business strategies are formed, operations and IT teams need to be able to execute to the best of their abilities. A systematic approach requires ways to deliver on business capabilities desired, the technology and processes that need to be adapted for execution, the products or services that need to be delivered (in our case, information) and ways to maintain strategic alignment.

Some fundamental dimensions to develop capabilities around are:
•    Demand management and delivery teams,
•    Technology fit,
•    Information-centricity, and
•    Strategic alignment.


In this first article of a four-part series, I will focus on the demand management and delivery teams.

Delivery Teams and Demand Management

How do you staff your BI teams? What kinds of skill sets do you hire and develop in-house? What kind of skill sets do you augment?

When you first develop a BI organization, typically the instinct is to look at standard organizational charts for BI teams and staff accordingly. If BI and information management are looked at from a service-oriented perspective, four areas emerge that could serve a majority of the business needs of any organization. The four services are:
  • Ad hoc service: Supporting the day-to-day information needs of the business, whether they’re operational or strategic in nature.
  • Standard report service: Producing, maintaining and enhancing canned reports for management and various functions within the company. This includes deploying standard reports as well as continuously validating information produced and supporting queries from business users.
  • BI infrastructure development: Developing an enterprise data warehouse, operational data stores, application data marts. This team would be driven by business requirements but would continuously focus on developing the right architectures.
  • Decision support services or advanced analytics services: Most advanced analyses that require piecing together information from multiple reports, or those analyses that do not lend themselves well to the traditional BI slice and dice but are more algorithmic in nature.


Managing the Demand for Information

Once you have this operating model, how do you manage demand? BI managers in any industry will tell you how the appetite for information has grown exponentially over the years. When BI teams aren’t able to scale according to business needs, the impulse is to think of an IT fix. Once again, what we have at play here is the classic supply side versus demand side fix. Actions that can be taken on the supply side are:
  • Improve productivity,
  • Better project management,
  • Adopt best practices,
  • Implement better technologies,
  • Improve processes, and
  • Hire more people.
All of the above - while necessary - do not necessarily tame demand. Actions that can be taken on the demand side include:
  • Develop self-service capabilities,
  • Train the end users to focus on analyses more than reports,
  • Meet periodically with business leads to better anticipate demand,
  • Institute a governance council to prioritize projects and tasks for BI teams (the priorities are all business priorities), and
  • Increase transparency on the supply side to show capacity.
With a more comprehensive look at both supply and demand, one can aspire to manage delivery better.

Skill Set

The next question is staffing the BI operating model. The cornerstone of any BI/DW implementation is a strong data foundation, including data models (logical/physical), an enterprise data warehouse, metadata management and business requirements management.

Core skills to hire/develop/retain are architecture, data modeling, BI requirements management and data warehousing project management. The implementation requires certain key additional skill sets like extract, transform and load development and reports development. Once the high level and detailed architectures are developed, this area can be handed down to DW/BI programmers who can execute. Nevertheless, most companies hire or procure talent in this second tier, which is often easier to find because it is more toolset-based. (Not to trivialize the nuances of any tools, but to do most simple/medium complexity tasks in any tool, ETL or reporting is not too difficult to pick up. Given that ETL development and reports development form the bulk of the implementation, most companies focus on hiring for these activities.)

However, in a changing business environment, priorities get shifted, business units get consolidated and projects get shelved. How, then, are you supposed to rationalize and retain talent based on varying demand? Even in a scenario where the business is thriving and demand for BI continues to grow, a more rational approach would be to organize and plan for staffing at two tiers. One tier is fixed and is core to the organization. The other is hired on demand to support initiatives. This tier can be trimmed if project priorities change or if projects get shelved.

Major investments are made in tier one to upgrade and retain talent. Since tier one sets the strategy and architecture, this is where more seasoned resources are required. These resources could move between projects, lending architecture and design, project management and stakeholders management services.

It is also important to maintain good vendor relations and select a couple of strategic partners who understand your company’s strategy and are able to provide skilled resources in a timely fashion.



The highlighted area in the figure represents staff augmentation. This is where you can bring in external resources with specific tool/technology skills and have them work from detailed specifications and instructions. Tier one represents your internal staff where significant investments are made to develop industry and company knowledge as well as capabilities in core areas like architecture and modeling.

This is the most important dimension in our strategic framework. Remember, to execute on a strategy, a service-oriented organization model combined with a flexible staffing plan is the cornerstone to maintaining flexibility in the face of change.

In part two of this series, I will discuss technology fit and how flexibility can drive your technology decisions and eventually support your business strategy.


Article taken from:  InfoManagement Direct, May 13, 2010
About the author:
Kary Kalapanda is director of business intelligence at Beacon Health. Kary has over 15 years of data warehousing experience.