Show Me The Data

The computer is becoming an effective tool for managerial and high-level decision making. This analytical processing environment, which is very different from the transaction processing environment, looks at and identifies trends and patterns vital to the vision of the management directing the organization. This environment empowers the business user to look for causative factors exploring "Why" and "What if" versus the traditional "What".

The success of your data warehouse implementation depends on the success of your data warehouse front end tool. The end user access tool is the only part of your data warehouse implementation that the user sees. Although experts tend to agree with this statement, with the exception of consultants assisting in gathering requirements, technology integrators have done very little to aid businesses in the selection of an end user access tool. To add to the complexity and confusion of this environment, new tool vendors are popping up at a astonishing rate almost all of whom claim adherence to the industry standard for tool vendors, E.F. Codd's original 12 Rules of OLAP (which in itself is controversial).

Because of DM&A’s expertise in the field of data warehousing, we frequently advise clients on all aspects of data warehouse implementations. This article is a quick synopsis of what we’ve learned through hands-on data warehouse implementations. We’ve focused on the end user access tool because of its significance to the success of a data warehouse implementation and because most articles, papers and reports have little to say on the topic that is immediately applicable.

This article focuses on how to select an end user access tool based on your business end user requirements.

For space and time reasons, it’s assumed that you’ve already made your decision concerning whether to build an enterprise data warehouse or a data mart or both. It’s also assumed that you’ve decided on whether you’re going to use a RDBMS or a multi-dimensional database for your datamart, and how you’re going to do your extraction and scrubbing routines to populate your data mart.

(If you haven’t made these decisions or would like more information about any of these areas, please refer to the following which are also available from DM&A:)

• The Benefits of Building a Data Warehouse

• The Data Warehouse/Datamart Debate

• The Multi-Dimensional Database/Relational Database Management System Debate

• Database Population--Extraction and Scrubbing

This article introduces a methodology for selecting an end user tool. Which includes:

1. Gathering and weighting end user requirements
2. Using the end user requirements to write a Request For Information (which should include your feature weighting)
3. Determining which vendors to submit your Request For Information
4. Scoring RFI responses and evaluating vendor demonstrations

To help you determine which vendors have tools that meet your end user requirements, several diagrams and/or charts are included in this article. The following diagram categorizes each vendors’ tool by query & reporting, OLAP or IT development.

(Briefly, Query & Reporting tools enable users to access enterprise data for reporting and graphing without having to know SQL. Query tools enable users to answer questions in lists. For example, a user could ask (and expect an answer) something similar to: "Retrieve all products costing between $50 and $125 purchased from ABC Supply Company on May 1, 1996." OLAP tools, on the other hand, are multi-dimensional. OLAP tools empower the business user to look for causative factors exploring "Why" and "What if" versus the traditional "What". For example, "Compared to the previous year, how have the last 12 months of increased advertising expenditures impacted my product sales in NYC compared to San Francisco?" IT development tools provide screen painters with 4GLs to provide query, reporting, graphing, and multi-dimensional analysis without programming, but they result in less flexible black-box IT applications.)

             4) Desktop-ROLAP: Cognos’ PowerPlay*, Brio*, Business Objects*, Pablo 3.1, IQ/Vision, Platinum Forest and Trees

             5) Desktop-MDD: Cognos’ PowerPlay*, Brio*, Business Objects*, Excel, Lotus 1-2-3, Oracle PC Express, Pilot Designer

*Indicates integration of Query & Reporting and OLAP characteristics in one tool

After determining which category of tool best meets your end user requirements, you can then review feature/comparison charts (similar to the following) to chose the most appropriate vendor to whom to submit an RFI. (Note: Feature comparison charts researched and updated quarterly are also available from DM&A.)

Query & Reporting Tools Feature Comparison Chart

Excellent Very Good Good Ö Satisfactory - Unacceptable