An example of an expensive delay in decision could be that of an e-commerce website - When a page takes a few extra seconds to load, the customer moves on to competition. So the question is, how soon can the operations team identify the root cause of customer attrition and take preventive action?
The key is to use all kinds of data and apply the right tools, policies, and models to analyze it and transform it. This, in essence, is the data refinery. It helps extract value through usable visualization and actionable insights. The efficiency and accuracy of data refinery depends on how the data 'pipes' are laid within the 'refinery' and how its components are integrated and configured. The architecture determines the manner in which data is ingested, orchestrated and crunched.
For today's businesses it is important that this happens at scale regardless of input devices, data types and data volumes. And more importantly, the business should be able to quickly re-configure the insights engine, and use the data 'pipes' and analytical models in new combinations and structures, based on the changing business needs.