If you feel like you are drowning in data and starving for information, you are a candidate for Business Intelligence (BI). The promise of BI is to turn raw data into useful information. This has become the tagline, in one form or another, for most BI solution providers.
Deciphering pages of columns and rows on traditional reports can be insightful, but often the process is more gut-feel than evidence-based decision support. Questions typically lead to more questions in an effort to understand root-cause business issues. This often leads to an overwhelming volume of information – sometimes referred to as “techno babble” − usually generated by IT, which perpetuates itself. The process may help owners and managers make intelligent business decisions, but it is not Business Intelligence, nor is it the best way to empower knowledge workers.
The true power of BI
In a BI environment, knowledge workers are presented with dashboard-like gauges and indicators to advise them on the “health” of performance metrics within their organizations. When indicators are within acceptable tolerances, the knowledge worker can move on to address only issues that require attention – just like the red lights on a dashboard. Here the journey begins. The knowledge worker can then “drill” into the problem area indicated by the red light indicator.
The true power of BI becomes apparent when the knowledge worker finds himself discovering answers to questions that he could not have even formulated at the outset of his own analysis.
Data warehouse - essential for BI
At this point, my technology senses kick in. Like most other solutions, there is no magic to BI. Between the volumes of raw data and the dashboard is “the plumbing”, as we used to say in my data warehouse teams. A data warehouse is just as it sounds: a place to store data to efficiently and effectively serve up BI. Think of it this way. A truck delivering raw goods into the physical warehouse for subsequent assembly and shipment to end consumers is like a transaction-based system (or ERP) delivering data into a virtual warehouse for subsequent delivery to information consumers.
This raises the question: “Do I need a data warehouse to ‘do’ BI?” The purists will say yes. If you forgo a data warehouse in the beginning and position BI directly on top of your transactional files (I would hope they are at least summarized transaction files.), just be aware that you will most likely need to revisit this decision sometime in the future.
Deploying BI directly on top of an ERP can eventually grind your computing environment to a crawl as data volumes grow and applications become more complex. Consider these factors in your cost/benefit analysis when designing your enterprise information strategy and roadmap.
If you have questions or comments on this blog, please contact Dale Fisher at firstname.lastname@example.org.