In the early years of multidimensional database development, the purpose of the technology was two-fold:
- To eliminate the massive amounts of manual effort required to support financial and management reporting.
- To provide intelligent tools for navigating through business information.
Summarizing transactional data can be a daunting task, requiring specific technical expertise and massive amounts of computing power and storage. Yet the ROI was compelling, given the reductions in time and resources needed to compile business statistics. Automating monthly reporting produced specific savings in reducing the number of production hours, the number of employees assigned to the projects and the turnaround time for reporting packages.
However, we are in a new era, where the customer demands an instant response containing all aspects of the "corporate" knowledge, including analytics. This presents infrastructure teams with a new priority requirement: rapid response. If your current data warehouse contains gigabytes or even terabytes of data, response time becomes the primary issue with customer service.
As with every other problem that arises out of technology paradigm shifts, this one can be solved. However, the resolution is not simply technology. It is a fundamental change in the way we think about
Requires Free Membership to View
When you register, you'll begin receiving targeted emails from my team of award-winning editorial writers on the latest customer relationship management (CRM)and call center technology issues today. Our goal is to keep you informed on the hottest issues facing this fast-changing industry.
Hannah Smalltree, Editorial DirectorSince the shift in the Web-world puts the customer in the information driver's seat, the data warehouse and its database engines have to be tuned to respond to this new driver. The tuning process begins with an understanding of the types of requests that the customer will submit. Based on these requests, Knowledge Marts are created that contain knowledge "slices" consisting of the most appropriate data for the request. These Knowledge Marts become a key tool to enhancing the customer experience, as well as supporting the evolution of employees into knowledge workers. Within the warehouse strategy, this new component of "on-the-fly" knowledge slices; consisting of views of legacy data, relational table rows and multi-dimensional retrievals coupled with pertinent business rules and documents; has to be defined, developed and deployed.
Businesses need to use analytics to quickly and accurately retrieve the data part of knowledge. Large data warehouses come with inherent overhead that extends the retrieval times beyond acceptable customer wait times. The result is lost customers. Knowledge marts consist of indexed views that facilitate retrievals and benefit organizations by:
- Improving accuracy, timeliness and consistency of answers.
- Making business processes more efficient.
- Contributing to the knowledge base.
It is imperative for organizations to understand, strategize and implement new business processes that build customer loyalty, improve knowledge assets, reduce administrative overhead and encourage participatory communications. Brand recognition grows out of the results of successful customer interactions. It is further strengthened by effective teaming, personnel growth and knowledge exchange.
About the Author
For the past 11 years, George has been involved with OLAP technologies as a visionary, project manager, developer and supporter. Most recently, he has an interview included in the IBM DB2 OLAP Redbook.
For More Information
- The Best Data Analysis and Business Intelligence Links: tips, tutorials, scripts, and more.
- Ask your technical data analysis questions--or help out your peers by answering them--in our live discussion forums.
- Have a data warehousing or BI tip to offer your fellow developers and administrators? The best tips submitted will receive a cool prize--submit your tip today!
This was first published in May 2001
Join the conversationComment
Share
Comments
Results
Contribute to the conversation