Methodology is very important to the success of a data warehousing project. Methodology is, in essence, the map that will help lead the data warehousing project team from point A to point B during the course of a project. When starting a data warehousing project, most project teams don't even think of using a methodology to help them. Most folks either go out and purchase books such as Building the Data Warehouse by Bill Inmon or The Data Warehouse Toolkit by Ralph Kimball and try to follow those guidelines to help them get started. While both books offer sound approaches to data warehousing, they don't give project managers a specific road map to avoid the pitfalls that inevitably occur during the course of a data warehousing effort.

A methodology is a collection of best practices and repeatable processes designed to guide a project team through a data warehousing effort. It is not just for the project manager to aid in planning and budgeting the project, but rather it is for the entire project team to help them in executing the project and to provide samples on how to develop key data warehousing products.

A good data warehousing methodology will provide the project team with:

Step-by-step detailed activities to complete the data warehousing effort. This is a checklist of the things a project team needs to consider when implementing the data warehouse. With the checklist, there should be a description of what is required in each activity.

Resource guidance on staffing the project team. The methodology should aid the project manager in identifying the number and types of people that should be involved in the data warehousing effort. The resource guide should document the desired skill sets of the individuals necessary to staff the project.

Predecessor/successor relationships. The activities themselves are not enough to successfully complete a data warehousing effort. The project manager needs to know what activities must be completed before moving on. A good methodology must take into consideration the dependencies between different activities.

Key outputs from each of the activities. Items, or outputs, are produced as part of an activity in the data warehouse project. These outputs may take the form of documentation, models or an actual product. The project manager and project team must know what outputs to develop during the course of a project.

Templates and samples for each of those outputs. The outputs alone are not enough for a project team to be successful. There should be example documents and templates for each of the outputs to help the project team in developing them.

Solid metrics for warehouse development. The whole point of a methodology is that it is a collection of best practices and repeatable processes. These are time-tested processes used in actual data warehousing projects. A methodology should have metrics to indicate how long particular activities take to help the project manager in the planning and budgeting of specific activities.

In addition to including the components previously mentioned, a strong methodology should take into consideration the following items:

Be iteratively driven. The days of building the monolithic data warehouse are over. Everyone wants to approach the data warehouse a "chunk" at a time while following an overall blueprint. Any methodology should follow an iterative development strategy and show a project team how to be successful in iterative development. This includes activities on how to identify the overall scope of the warehouse, how to identify business imperatives that will satisfy the end-user needs, how to scope individual iterations based on business value and feasibility and how to execute each individual iteration.

Be driven by a business case. Too many warehouses are unsuccessful because they are IT-driven activities. The voice of "if you build it, they will come" works only in Hollywood, not in data warehousing. A good methodology should drive the business benefits of the data warehouse. These benefits will help justify the warehouse concept to upper-level management and gain support for it throughout the organization.

Include roles and tasks for business user participation. A data warehousing project is not an IT project, but rather it is the enactment of business decision support capabilities. Due to the integral involvement of the business community, any set of data warehousing activities should consist of business-user involvement. This integrated activity list will help identify the key areas where IT and the business must work together.

Be flexible and customizable. Every organization is unique and needs to approach data warehousing differently due to its internal processes. A good methodology will be easily customizable and flexible. It is this ownership and improvement of an out-of-the-box methodology that will help an organization move to a higher level of data warehouse maturity.

Overall, a strong methodology will help your project team succeed in implementing a data warehouse. It is not a panacea for all the problems of data warehousing. You still need strong sponsorship, skilled resources and organizational buy in, but a methodology is a key component that will help identify what you need. In this era of rampant data warehousing, the use of a data warehouse methodology will help project teams avoid a collision with failure.

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