By Matt Vranicar
May 6, 2010
In today’s rapidly changing marketplace, the need for enterprise-wide business intelligence (BI) has never been greater for companies. BI can lead to better decisions and more effective approaches to all company initiatives - from reaching customers to honing internal processes.
Yet studies show that more than 50 percent of enterprise-wide BI efforts fail, and they fail for a variety of reasons. They fail because they lack executive sponsorship, they do not quickly garner user adoption, they focus on technical approach rather than business value, and more. But, at the heart of all these failings is the premise that BI, to be done properly, must be an “all or nothing” proposition.
An initial foray into BI targeted at finding the solution to a specific question or concern can deliver rapid ROI and set the stage for future success. By “starting small,” the goals of minimizing capital expenditure and achieving business intelligence are no longer at odds. By focusing on high-profile business issues that are solvable with available, company-owned information, organizations of all sizes can take an appropriate first step on the BI journey. They can avoid the 50 percent failure trap that has plagued too many good companies for too long.
Of course, even the simplest of projects can hamper progress if not built with an eye to the future. Success requires a careful balancing act - one that combines a targeted, rapid-ROI implementation with scalability to meet a long-term vision.
The answer is less complicated than you might think. Here are four tips to help you get started:
-¢ Focus on targeted areas of the business that deliver the highest ROI - Look to deliver value to a targeted area of the organization quickly. Focus initially on a high-profile problem within a specific department, but that is also of strategic importance to the company as a whole. An incremental approach allows companies to view BI as a series of projects, each of which fits neatly into a larger strategic vision to improve decision-making across the business.
-¢ Involve users early and often - Focus on meeting the practical business needs of end users from the beginning. While database structure and layout are important elements to overall BI design, implementing technology for technology’s sake is misguided. By involving end users early and often and sharing development progress in a continual loop, you increase their investment in the system and better meet their needs. When users are part of the BI effort from the beginning, you remove psychological barriers to adoption, imaginations are ignited, and requests for additional analysis and reporting capabilities become commonplace. BI growth and expansion to an enterprise-level is a natural progression of involving the end users.
-¢ Scalable technology - BI does not necessarily require an enterprise-level application, but your BI technology should be flexible and capable of scaling as your business needs change, or as you do scale it to an enterprise solution. There are BI platforms available today that can affordably be introduced on a small scale initially, while laying the foundation for expansion to enterprise-level platforms as the need arises. This scalability can take different forms depending on company size and needs. For instance, smaller companies may leverage a virtual data warehouse that can transition to a physical one. Other companies may start by implementing a framework for “pluggable” applications. No matter which approach you decide upon, the progressive element is critical. The last thing you want is to redo all the work previously performed when expansion becomes appropriate.
-¢ A flexible data warehouse - The underlying data architecture, responsible for storing the data displayed in business intelligence reports, should not be a black box created for a specific purpose and left alone. Proper data organization and architecture allows for different types of data and transactions to be added to the warehouse as business needs require. These new subject areas of data will complement one another by creating an ever-increasing view of your data, while allowing detailed analysis in each targeted subject area. Remember, with each targeted effort, you want to further fulfill the enterprise vision.
By ensuring that your first step into BI focuses on these four critical elements, you can transcend the pitfalls that plagued business intelligence implementations of the past, proving that BI can be conducted with minimal impact on an organization’s time and budget. In addition, by recognizing the ROI of a targeted implementation, your organization can more accurately assess the potential return on broader programs. In fact, by selecting the right initial implementation, you have a great shot at generating enough ROI to fund future efforts.
Matt Vranicar is the vice president of Business Intelligence and Data Warehousing at Rolta TUSC (www.tusc.com), which works with large enterprise organizations across the U.S.