Databases Reference
In-Depth Information
Why does this matter? Why pay such attention to the purpose of a data
warehouse? Without a valuable yet feasible purpose that simultaneously
expresses the value of a data warehouse, that data warehouse is vulner-
able. The gravitational pull of data may cause a data warehouse to fail to
achieve its purpose and goals. Admit it. Every data warehouse achieves
value and ROI, which is neither published nor measured, and may be the
“real” reason for the existence of a data warehouse. To continue to deliver
its real, yet unpublished, purpose a data warehouse must be able to resist
the influence of the gravitational pull of data.
GrAvItAtIoNAl Pull of A DAtA WArehouse
A successful data warehouse that achieves its purpose and adds value and
ROI can become its own problem. Well-organized, high-quality data with
complete metadata in a data warehouse can attract a problem to itself—the
gravitational pull of data. Oddly enough, if left unchecked, the gravita-
tional pull of data will increase the probability that it will destroy the data
warehouse as the data warehouse increases its size and success.
Gravity: Two objects, if they are close enough to each other, will be
drawn toward each other. When one object is larger than the other, the
small object (i.e., the apple) moves while the large object (i.e., the earth)
remains stationary. The same is true for a d ata warehouse. Within an
enterprise, a d ata warehouse can be the large object attracting other
objects to it. In an enterprise this gravity occurs by the economies of
scale presented by a data warehouse. For example, a data warehouse may
have Sales, Product, and Customer data. That would mean that a d ata
warehouse has spent and expensed the investment necessary to analyze
the Sales, Product, and Customer data in operational source systems.
That would mean that a d ata warehouse has spent and expensed the
investment necessary to design and develop the ETL applications that
capture and retain that Sales, Product, and Customer data. Finally, that
also means that a data warehouse has spent and will continue to spend,
expensed and will continue to expense, the overhead incurred by captur-
ing and retaining Sales, Product, and Customer data. A new application
needing Sales, Product, or Customer data can either replicate all those
expenditures, expenses, and investments, which will expand the budget
and reduce the ROI of the business unit sponsoring the new application,
