Contextual Analytics

Too often our analytical work is done without knowing the necessary context of what our results are used for. I like to call it the Perpetual Circular Analysis (PCA). No one knows what the analysis is suppose to do. So, if one has produced and shared the data driven results, then there will be more questions and doubts, so one just restarts the analysis with slighlty different approach. This can go on forever until a senior executive is so fed up that she or he just ignores all the analytical work and makes a hasty and ungrounded decision. Then the circle is closed and can start again with a new questions. Doesn't this happen so often in our professional work?

But one ignored factor of our analytical work is that we create our own context through the applied methodology and our communicated results. We are shifting and changing the context of where the team was prior to the analytical work, prior to any data related insights. And this is a good thing. I suggest focusing more on this creation of a changed context than fixing every single PCA challenge. But we have to be more deliberate and conscious of how we change the team's context and for which purpose.

Borrowing and expanding a thought from Adrian Holovaty (founder of http://www.chicagocrime.org/) analytical work should achieve three contextual elements:
- Creating a context in which other people can think
- Creating a context in which other people can sharehttp://www.chicagocrime.org/
- Creating a context in which other people can create
This often neglected purpose of analytics is well demonstrated by the above mentioned project of analyzing and sharing crime data in Chicago or by the “We feel fine” project which collects, analyzes, and visualizes Online mentioning of “Feelings” (http://www.wefeelfine.org/ ). These analytics create contexts that enable thinking, sharing, and creating.

Therefore the best analytical work creates a new context along the Open Source philosophy of making Data and Methodology available, enabling understanding through transparency, and encouraging derivative and correlated work. Maybe this approach will help us to experience less analytical work along the Perpetual Circular Analysis pattern but more stimulating and interesting team discussions, inspired and driven by Analytics.