Databases Reference
In-Depth Information
variety of problems. It could be that the appropriate people were not polled,
which may indicate communication problems. A lack of business-oriented
responses may also indicate that business results have not been captured and
tracked. With further exploration, the true value of the data warehouse may
be discovered. Don't be shocked if no concrete value can be identified from
the data warehousing environment. Your organization is not alone. Many data
warehouses are less than successful, and some are a complete failure.
Discovering the Data You Already Have
Of course, another critical part of these systems is the data itself. You don't
need a list of specific data elements, which is at too fine-grained a level of
detail. More useful is an overview of the type of data that is stored and used in
this reporting environment. Questions to help you understand each reporting
environment include the following:
What type of data is included?
What level of detail is included? Individual transactions or summaries?
At what summary level?
How much history is provided?
How often is the data updated?
From which source systems is data pulled?
What is an estimated size of the data (terabytes or megabytes)?
What is the confidence level in the data (high, medium, or low)?
Are there any plans to make significant changes to the data?
Is a specific data warehouse architecture or methodology being used for
this system (e.g., Inmon or a Kimball approach)?
How is the data organized? Is it in flat files? Is the data structured
dimensionally?
Document what you learn about the data. Table 3-1 shows an example of a
data summary. Complete what you can, but don't take months to track these
details down. Simply capture what you can in one or two meetings. When this
is collected across all the different data warehouse-like systems, it is easier to
highlight areas of overlap and possible redundancy. This alone cannot be used
as justification to eliminate one system versus another. There are often many
valid differences that require additional research and analysis to determine
what should or could be done. This is simply a mechanism to understand what
you already have in place.
