Traditional business intelligence (BI) places a strong emphasis on data control and the delivery of static dashboards and reports. The underlying data may be refreshed, maybe even in real time, but the metrics, KPIs, and configuration of the visualizations and reports remain the same unless someone, most probably an IT developer, updates or changes them.

The process for working with the customers of BI to build new reports or find new analytic insights is usually too slow. The customers tend to make ill-formed and indiscriminate requests because they are not quite sure what they want, are often precluded from truly agile development in which they can quickly change or drop things that have been built, and insure themselves in a way by keeping a long list of all possible requirements in the queue. Meanwhile, the IT professionals, perhaps overwhelmed with requirements, perhaps coping with limited resources, take too long to analyze the requirements...