bois construction

Design and Construction of a Data Warehouse
Managing Customer Data in an Enterprise-wide Warehouse
Several fundamental characteristics guide the development of a data warehouse, a corporate system which will enhance an organization's management of customer data.
   
Four significant aspects need to be considered in the design and development of a data warehouse:

The major function: to analyze historical data, to reveal trends and correlations, and predict future outcomes
Potential business insight : valuable and unpredictable insights often result from analytical expeditions, based on information derived from the system, where starting points, routes taken, and end points were not initially known
Granularity of data: the capability to drill down into the detailed data
Assembling data form diversified sources- to ensure consistency of interpretation throughout the organization
Limitations of a Data Mart
A data mart containing summary data does not meet the criteria of the basic data warehouse concepts defined above, a drawback that needs to be recognized by the project team if the data mart approach is to be taken. A data mart based on summary data cannot evolve into a a fully functional data warehouse and requires extensive rework to meet enterprise-wide objectives.

As well, summary data may obstruct the view of the business, preventing an end user from drilling down to locate essential transaction data required to resolve an issue. The most effective methodology in designing a data storage facility which will provide the greatest benefit to the corporation, is to acknowledge the fundamental requirement for the most extensive detailed data to be gathered and stored, for access by business users.

Data Concepts
The concept of data and its categorization in a data warehouse leads to a second set of fundamentals regarding this data.

The first one is that the detection of customer trends is related to the amount of historical data stored- the more history, the better. Most companies retain 15 or 20 months of historical data, depending on the business sector and the potential longevity of customer relationships.

The second is that the most significant factor in delivering a successful data warehouse application is an understanding of the business purpose driving the need for information.

Restrictions of Summary Data
Summary data can restrict the view of a business, if it does not have the parameters that allow it to synchronize with other data inputs. For example, in the retail sector summary data based on aggregated sales limits the ability of the retailer to identify the one-day impact on sales created by an ad which runs twice weekly for a period of time.

In the consumer goods sector, summarized data that does not separate product quantities by distribution channel, will not enable the organization to determine the volume of product, number of returns, or profitability by channel.

A full-function data warehouse which stores and manages detailed data can provide considerably more relevant, complete, and timely information than a data mart.