This article originally appeared on the BeyeNETWORK.
Intelligent Learning Organization
As we move further into the Information Age, or the “Knowledge Worker Age,” as Stephen Covey describes it in his new book, The Eighth Habit, we information professionals must increase our enterprise’s ability to exploit its information and knowledge resources. Covey describes how we move from effectiveness to greatness by applying the eighth habit—which is not a new habit—but a discipline of finding our voice and helping others find theirs.
Effective learning organizations manage their information as a resource and ensure quality of the information products by applying sound quality management principles to improve, error-proof and control the processes to meet knowledge worker expectations. They capture lessons learned in knowledge bases that enable knowledge workers to learn from the experiences of others. The operational data stores (ODS), data warehouses (DW) and data marts are true “strategic databases” that support the strategic and decisional processes effectively and efficiently.
The irony is that we are at about the midpoint of the Information Age, and most organizations have the technology of the Information Age, but they have built and delivered poorly designed applications on top of disparately defined databases. Rather than enlightening knowledge workers, we have condemned them to a life of “information ‘hunters and gatherers,” having to hunt and chase for their information, and then verify and fix it before they can use it. Many times the very data warehouses and data marts we build to aid those knowledge workers, fail to meet their most basic needs.
Increasing the Intelligence Quotient of Business Intelligence
Some definitions are in order. I define business intelligence as “the ability of an enterprise to act effectively through the exploitation of its information resources.” The business intelligence environment is, “Quality information in stable, flexible databases, coupled with business-friendly software tools that provide knowledge workers timely access to, effective analysis of and intuitive presentation of the right information, enabling them to take the right actions or make the right decisions.”
Steps to increase information quality in our business intelligence environment include:
- Acknowledge the product of our business intelligence environment is information.
- Understand the information consumer’s information quality requirements. This requires meeting with them to understand their needs and the consequences of nonquality.
- Design the strategic database models with an “enterprise-strength” design. Do not replicate the defects in the legacy systems in the base ODS. Rather, design the ODSas a true enterprise database. Facilitate the definition of the data on a subject by subject basis, involving business subject matter experts who understand and can represent ALL information views of the enterprise from that subject. (1.) Operational Data Base (ODB): If designed properly, this ODB can become the go-forward record-of-reference database. Use it to re-engineer legacy systems and their disparately defined databases when feasible. You can control this re-engineering in incremental, manageable projects. (2.) Operational Data Stores (ODS): Design the ODS with time-stamps to reflect the historical view of your enterprise to support data mining. (3.) Data Warehouse: Design the DW as a HUB from which data marts can be generated to meet specific sets of information views.
Assess information quality in the source databases early. Do NOT wait till the last minute to check the quality of the source data. Any required corrective maintenance can delay the implementation data—or worse—you implement defective information on time and alienate your information “customers.” If you lose their trust, you have a much more difficult time trying to gain back your credibility.
- Assess accuracy, not just validity. Contrary to popular belief, you cannot measure accuracy electronically. This requires physical comparison of the data to the characteristics of the real-world object or event the data represents. Even if you take a small sample of 300-500 records, measure accuracy. Data that meets all business rules and is reasonable but inaccurate can have devastating consequences on business decisions.
- Correct dataat the source if the data is still used there. The only exception is if there are legal requirements prohibiting this, such as information in legal documents, or if the data is not ever used from the source; it is only used as a data collection database, such as in hand-held computers. Not correcting data at the source creates negative side effects: (1.) Sub-optimizes the data correction costs. Processes using data at the source continue to fail. (2.) Reports that should balance from source to target will not balance. (3.) Defective data at the source can comeback to corrupt the DW data, if subsequently propagated later.
- Improve the processes to prevent recurrence of the defects discovered. No data correction initiative can correct all errors. Some data, such as event data, may never be discoverable after the fact. Other data will be prohibitively expensive to correct.
- Observe knowledge workers when they first start accessing the new business intelligence databases as they are introduced. You want to observe if there are problems. If you do not see this, the knowledge workers will go away with “complaints” about the information quality and may not come back. Be there so you discover any problems and address them quickly to retain your information customers.
- Measure customer satisfaction at the outset of your information delivery, and periodically to assure you are consistently meeting customer expectations.
By providing quality information in your business intelligence databases you increase business effectiveness and enable the Intelligent Learning Organization.
Please share your ideas for the Intelligent Learning Organization atLarry.English@infoimpact.com. For additional Information on Information Quality please click on my latest white paper: Ten Mistakes to Avoid if Your Data Warehouse is to Deliver Information Quality!