When I hear of a nonprofit interested in a "data analytics" project, my mind jumps to the usual technical stuff: torturing data; making pretty pictures out of the confessions; maybe something a bit over-engineered (a regression tree where the leaves are Bayesian deep learning neural networks?). A couple of Pro Bono Analytics consultations, though, have focused on what you might think of as precursor topics: what data does/should the organization have; where is it/should it be parked; who gets access to it.
A recent article in Analytics Magazine
, "The rise of self-service analytics
", raises related questions in the context of corporate analytics. (A slightly modified version of the same article appears in OR/MS Today
.) If any of us end up working on one of these "data strategy" consultations (for lack of a better term), I think this article might provide some ideas for how to frame the discussion.