Simple Complexity

Building a Quality Business Intelligence Framework

The July issue of DMReview has a great article by Brian Swarbrick, Building a Quality BI Framework Solution Starts with a Quality.  Brian writes that there is no shortcut to building a quality business intelligence framework.  In fact, as the solutions go across the enterprise, picking the best features and capabilities from multiple solutions, quality becomes harder.  It needs to be spread across the entire architecture.

There is significant attention paid to the front end because that is what users and customers can see.  This is frequently to the detriment of the underlying data quality.  In the end, solutions can only be successful if they are serving up high quality data.

The key to building a quality ETL solution is to define and develop standard processes and templates throughout the architecture, design and development phases of the project and ensure that those standards are leveraged correctly. These standards encompass more than just corporate naming standards for database development; they also include process standards, standards for all ETL processes and data movement activities across the entire data architecture. Developing a consistent set of standards ensures conformity across the components, reduces development and testing times, simplifies maintenance and reduces total cost of ownership. Also, because many BI efforts are now developed using a combination of local and contractor resources, including offshore development and support models, it is increasingly important that these solutions are designed and built with solid common standards and processes in place that have been tried and tested.

The cost of low quality data is extremely high.  It is hard to convince organizations to pay adequate attention to their data quality.  It is hard, painful, and time consuming work.  It is however, the foundation that your business intelligence system sits.

Popularity: 6% [?]