Intelligent Enterprise

Better Insight for Business Decisions

RSS
Webcasts
Whitepapers
Subscribe
Home

November 15, 2002

Designing the Financial Data Warehouse

Now that the finance function has knocked on your door, what's next?

by Dan Hughes
Edited by Ralph Kimball

Continued from Page 1

Many of the more difficult questions about enterprise performance can't be answered efficiently because of the separation between the GL and subledgers. By integrating the GL with the subledgers, all the summary and underlying data are in a single repository, thus providing all consumers of financial data access to transaction-level details.

A Proven Approach

As you think through the design of a financial data warehouse, always start with the GL and gain a firm understanding of the COA. Whenever a requirement to analyze subledger data comes to you, look at how the financial measures roll up to the GL. Here are a few key questions to ask:

Where's the information represented on the financial statements?
Are there users of the financial statements who could benefit from having access to the detail information contained in the subledger?
What are the data keys (specifically, COA segments and system identification links) that tie the two systems together?

This approach lets you evaluate the business requirement of providing the user access to data contained in the subledger while thinking about the organization from a perspective of global financial-statement analysis. Once you integrate the first subledger into the data warehouse, future subledgers can be added incrementally.

Over time, the repository becomes more valuable to the enterprise and eventually the single platform for financial decision support. This approach gives the subledger users a tool to analyze transaction data, while the financial statement users have greater insight into the numbers at an enterprise level. The requirements of both groups of users can be satisfied.

Rate This Article

Comments:

Optional e-mail address:

Dimensional Modeler's Dilemma

Creating a data warehouse that integrates many diverse sources of data always requires the crucial step of imposing a global discipline on the textual labels and numeric measures found in these data sources. The dimensional data modeler calls this step "conforming the dimensions and the facts."

Experience convinces me that building an integrated financial reporting system, spanning the high-level GL and the subledgers that feed this GL, is a successful approach. Conforming the dimensions and facts requires you to adjust your data collection and reporting systems to reflect the revised labels and measures. So what do you change: the GL, the subledgers, or both? Before you can answer that question, you need to create the first design deliverable — a dimensional model that captures the information recorded in the GL. In a future column, I'll describe a dimensional model that supports basic balance sheet and income statement analysis. With this information in hand, you'll be ready to tackle the subledgers.

Guest Columnist Dan Hughes [dhughes@clariteegroup.com] is a partner with Claritee Group LLC in Chalfont, Penn.

RESOURCES

See all past Data Warehouse Designer columns at www.intelligententerprise.com/ports/search_webhouse.jhtml.

Go to the Data Warehouse Info Center for more articles on the subject.