Market Analysis Data Mart Design

How to rapidly convert the data into knowledge for business and how to assist department managers to make their decisions, optimize their process of management, improve their abilities of fast response to the market changes, are all of the urgent problems which need to be resolved by marketing department. The established data mart merges both the internal and external data of the business, effectively reflects the timely changes of sales market to the decision-makers.Firstly, the necessity of creating a sales data mart is discussed and then the processing procedure of ETL is analyzed on the basis of the discussion of methods on integration data, including the issues of how to process noise data, the consistency of data, and data quality. The data migration tools are used to unit the data sources and to establish multi-dimensional data mart models for meeting requirements of users' ad-hoc queries. The snowflake schema and galaxy schema are applied to the design of the data mart. Accordingly, the discussion of dimension's granularity and strategy of data partitioning are made. In the physical model, it is achieved based on the physical table structure and index strategy. Finally, it's given the prospect of sales data mart in sales analysis.