In the first part of this series (January 2000), we reviewed two of the most important structures of building the data warehouse (DW) in order to support business intelligence: project management and the business case development. In this column, we will continue with a discussion about how to collect business requirements to support the business case utilizing the business question assessment, as well as how to establish the data warehouse technology strategy employing the architecture design and tool selection technique. These sections are highlighted in red in Figure 1.

Business Question Assessment

The business question assessment (BQA) provides an approach to understand the business users' needs, establish the data warehouse subject areas and scope the individual project iterations. This is accomplished by further defining and prioritizing the business requirements that were set forth in the business case development phase.

The first step is to identify the individuals who understand the detailed business questions and have a vested interest in the project. Focus sessions and interviews should be used to identify key business questions the data warehouse must satisfy and their associated data sources. Items produced as a result of these sessions include detailed information requirements, data elements and source systems, and performance requirements for the data warehouse.

Development of a high-level logical data model is critical. The logical data model is an overview and should represent the integration points between subject areas that will ultimately be incorporated into the overall warehouse. As part of this effort, data elements are combined around central subject areas to produce a model that will later be turned into the facts and dimensions of the warehouse.

Next, implementation alternatives for populating the data warehouse should be developed. A plan should be put into place that allows an iterative delivery of the subject areas within the warehouse. A good way to do this is to develop a feasibility/value matrix. This gives insight into those projects that provide the highest value and are most feasible.


Figure 1: The Data Warehouse Development Approach

Finally, you should estimate the cost based on the selected population project. The investment cost should be presented to the audience responsible for making the decision to move forward. If the stakeholders have been informed throughout the process, this final step should be a simple formality. Once you receive approval, you should be prepared to assemble the development team and begin work.

Architecture Design

The architecture design sets the overall process and technology framework for all iterations of the project and includes data architecture, application architecture, technical architecture and the support architecture.

The data architecture defines the quality and management standards for data and meta data and helps ensure that data loaded into the warehouse is useful, timely and accurate. Areas that should be addressed are source data tracking, data population projects and a data ownership program.

The application architecture is the software framework that allows for the overall implementation of business functionality within the warehouse environment. It controls the movement of data from source to user, including extraction, cleansing, transformation, loading, refreshing and accessing (reporting, querying).

The technical architecture provides the underlying computing infrastructure that enables the data and application architectures. It includes components such as end-user workstation hardware/software, database platform, security software, platform server, network, etc.

The support architecture includes those areas necessary to manage the technology investment effectively and the DW software components: tools and structures for backup/recovery, disaster recovery, performance monitoring, reliability/stability compliance reporting, data archiving and version control/configuration management. The purpose of the support architecture is to ensure the DW operates smoothly over time and serves its users well.

Tool Selection

In the development of the data warehouse, many tools will be required to do the job. The development of the technical architecture defines the tools needed to successfully complete the data warehouse.

The most important aspect is to choose the right evaluation criteria. First, look at high-level criteria such as market reputation, analysts' ratings, compatibility with existing tools, scalability, availability and price. Eliminate the tools that will obviously not work for the environment. It is recommended that the vendor list be reduced to a short list of no more than two to three finalists per tool category for the actual evaluations.

The next step is to develop a matrix that lists the critical evaluation criteria for each tool. Arrange for the short list of vendors to demonstrate their tools and test them on the specific requirements of the data warehouse that will be built. Test cases should be created and a test environment made available which mirrors the production environment's technical infrastructure as closely as possible. Results of the tool selection should be documented and presented to management. It is recommended that the same team members be used for each vendor test to ensure consistency. Apply scoring to the evaluation matrix to determine the best tool for the job.

Summary

You have now created the foundation for your data warehouse by understanding the overall business needs, selecting the technical architecture and tools. The next step is to develop the actual data warehouse and business intelligence solutions for your business users. Next month, I will provide information on how to deliver business intelligence solutions by implementing iteration project planning, detailed design, development and transition to production.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access