Data Warehouse Building Blocks

Collecting information is the easy part. Knowing how to store it, access it, and analyze it makes all the difference.

Jay-Louise Weldon and Alan Joch

To extend your core business, you need comprehensive information. For many companies, that means a robust data warehouse that draws together disparate and unfiltered data and presents it in creative new ways. As tools to capture and explore detailed data mature, so will our ability to find ways to exploit the data we've collected.

In the last 10 years, two factors have combined to help data warehouses proliferate. First, we've recognized the benefits of on-line analytical processing (OLAP) beyond the traditional areas of marketing and finance. Organizations now find that the insights buried in the masse s of data they routinely collect on their customers, product s, operations, and business activities contribute to cutting operating costs and increasing revenues, not to mention making it easier to arrive at strategic decisions.

Second, the growth of client/server computing has spawned server hardware and software that's more powerful and more sophisticated than ever. Today's servers now rival yesterday's mainframes and offer technologically superior memory architectures, high-speed processors, and massive storage capacities. At the same time, modern DBMSes provide more support for complex data structures and promote standardized middleware. From this hardware/software renaissance emerges the multiterabyte data warehouses we're now seeing in client/server environments.

How do you take advantage of these technology advances? In the following pages, we'll describe how to choose the right warehouse for your enterprise. "Warehouse Cornerstones" explains the pros and cons o f centralized and multitiered warehouse architectures and gives advice on how to choose the right servers and DBMSes. "Better Clients, Better Decisions" will help you match the right analysis tool to the cross-section of people who will be using your data warehouse. And "Take Your Data to the Cleaners" discusses the choices you have among home-grown and commercial programs that filter out nagging inaccuracies and inconsistencies in your information.

A data warehouse consists of myriad pieces. If you choose them wisely, you could end up with a repository of invaluable data to inform your strategic decision making. In all likelihood, there is gold buried in the data dispersed across your enterprise. You only need to find it.