Published in DM Review Online in September 2000.|
Printed from DMReview.com
Business Intelligence: Top 10 Warning Signs of a Troubled BI Projectby Jonathan Wu
A successful implementation of a business intelligence (BI) application requires the project team members to develop a solid understanding of the functional requirements and the technology in order to meet or exceed user expectations. Having worked on numerous BI projects in various capacities, I have experienced a number of challenges in implementing these applications. These challenges range from lack of internal support to technical difficulties Ė any of which could have caused a project to fail. From these experiences, I have developed a thought-provoking list of the top 10 warning signs of a troubled BI project. The warning signs displayed in progressive order of concern.
10. Inadequate Sponsorship
The project sponsor is an individual who approves and provides the funding for the project. This individual is committed to the success of the project and is typically the internal champion who promotes and helps resolve any internal disputes or issues that are impacting the project. Without dedicated sponsorship, the project may not receive the recognition or the support it needs to be successful within the organization.
9. Inexperienced Implementation Team
A project team that lacks experience with the BI application, the subject matter being addressed and effective communication skills is at risk of failure. Not that bright and hard working individuals can not get it right the first time, but the experience is an incredibly steep learning curve. Most BI projects do not have the luxury of time, which may be a mitigating factor to an inexperienced staff. Additionally, an inexperienced staff will not necessarily approach the project from a best-practice perspective either. Individuals that know how to implement the BI application are critical to the success of the project.
8. Lack of Resources
The lack of resources can mean a number of things but, in this context, Iím talking about adequate funding, access to the appropriate supporting technology and quality team members. Without adequate funding to support the BI initiative, the scope of the project will need to be realigned appropriately which may not satisfy user expectations. In addition, supporting technology such as middleware for database connectivity, relational database management systems and hardware that are needed for the BI application to work successfully must be addressed in a manner that supports the BI project. Another resource issue is the quality of the project team members. Do they have the experience, knowledge, skill set, aptitude and desire to work collectively as a team and strive for the success of the project?
7. Informal Selection Process of the BI Application
Was there user representation during the software selection of the BI application? Does the BI application meet the usersí functional needs? Understanding the needs of its users as well as the functionality of BI applications that are available is the basis of a successful selection. Using a formal approach that is a business-based methodology can greatly improve the opportunity for selecting the best BI application for the organization. This process builds consensus and provides a means for documenting the decision that was rendered. This is an involved and time-consuming process. However, BI software selections that are approached in this manner have a much higher success rate and are more likely to meet user expectations. By investing the time to plan and select against defined goals, an organization can minimize potential selection risks and maximize the opportunity for selecting the best BI application for the organization. (Please refer to "Which Application is Best for an Organization?" DMReview.com Online Columnists, June ,2000 for further information.)
6. Inadequate Functional Testing
Has the BI application been adequately tested to ensure that the configuration meets the information needs of its users? A properly configured BI application will provide its users with the ability to access the information that they need in order to answer the business questions that they have. Functional testing provides the project team with a means of checking whether or not the BI application was configured according to the needs of the users. Do not shortchange this step. If the BI application does not provide the information the users need, then the project will fail because they will not use it.
5. Performance Problems
From a user perspective, instability or waiting for the BI application to produce the information requested can be annoying. If it crashes too often or takes too long, it becomes extremely frustrating. Users do not want to wait a long time for the information they have requested or work with an unstable application. After a couple of frustrating experiences, most users will abandon the BI application. Performance problems can arise from the configuration of the BI application, the network bandwidth and/or the source database schema. The BI application should be thoroughly tested before it is deployed so users do not encounter these problems.
4. Inadequate User Training
Will the users of the BI application know how to use it, and will they understand the subject matter being accessed? If not, how will they be able to effectively work with the BI application? The objective of training is to provide them with an understanding and practical experience of the features and functionality of the BI application through lecture and lab exercises. In addition, users will need to be trained on their BI configured environment, the associated business rules and definitions as well as the data that is available to them. The training needs to put the learning experience into the context of the data that the users will be accessing for decision-making purposes. (Please refer to "Why Provide User Training?" DMReview.com Online Columnists, May 2000 for further information.)
3. No Formal Project Plan
The old adage of "a failure to plan is a plan for failure" is applicable to BI implementation projects. The project sponsor, the implementation team and the user community need to know who will be working on the project, what are the roles and responsibilities of each team member, what will be delivered and when, the critical success factors for the project, and the user acceptance criteria. This information must be documented and communicated to all involved and interested parties so there is no misunderstanding and everyone is working toward a successful implementation and deployment of the BI project.
2. Project Scope Not Formally Defined
What will be delivered? The expectations of the project sponsor, the project team and the user community will all differ unless the scope of the project is clearly documented and communicated. A project can never be successful unless everyone involved understands the project scope. Projects fail because the deliverables do not meet the expectations of the user community. Expectations can not be managed unless they are first established, agreed to and communicated to all that are involved in the project.
1. Inadequate User Involvement
How can the project team configure the BI application to meet the needs of the users without their involvement? The naming conventions, business rules and data context that is needed by the user community must be incorporated into the configuration of the BI application. Since the BI application is to be used by them, they should have a voice in the implementation. If the project does not have adequate user involvement, then the project has a high likelihood of failure because the BI application will not satisfy the expectations of the users.
Any one or combination of these warning signs can cause a BI project to fail. The intent of the top 10 warning signs is to provide you with an idea of what could lead to a failure so you can address the problem before it dooms your project. Good luck!
Jonathan Wu is a senior principal with Knightsbridge Solutions. He has extensive experience designing, developing and implementing information solutions for reporting, analysis and decision-making purposes. Serving Fortune 500 organizations, Knightsbridge delivers actionable and measurable business results that inform decision making, optimize IT efficiency and improve business performance. Focusing exclusively on the information management disciplines of data warehousing, data integration, information quality and business intelligence, Knightsbridge delivers practical solutions that reduce time, reduce cost and reduce risk. Wu may be reached at firstname.lastname@example.org.
Copyright 2006, SourceMedia and DM Review.