Deciding (or planning) based upon acquired insight

In this step, the data scientist hopes to obtain value from their efforts in the form of an insight. The insight is gained by performing the preceding described phases, aimed at gaining an understanding of a particular situation or phenomena. The idea is that this insight can then be used as input to make better decisions.

A fun example that illustrates a creative use of insights mined from data is the (as of this writing, experimental) Roztayger personality match process powered by IBM Watson. Using either your Facebook or Twitter feeds (or you can enter a short bio), Watson will, on-the-fly, perform an analysis of your personality. The results are interesting and pretty spot on, and these insights are then used to suggest designer labels that may best suit you and your personal style.

Once the (real-time) data science analysis is complete, the aforementioned website not only provides its recommendations but also shares the data behind its insights, showing an easy-to-understand, well-organized tabular view of the results, and an eye-catching visualization as well, as shown in the following figure:

This illustrates another key aspect of this phase of the data science progression, that is, once the data scientist identifies an insight, he must clearly present and communicate those data insights/findings.