I feel a lot more pessimistic about analytics. I think almost all analytics, including non-real-time analytics, are bad for business. It has everything to do wi

I feel a lot more pessimistic about analytics. I think almost all analytics, including non-real-time analytics, are bad for business. It has everything to do with the more toxic aspects of business management culture (quite possibly MBA culture, but not necessarily). It manifests itself in the form of a Cockpit Fetish: "Give me all of the data, and I will be the wise and benevolent mastermind that will tweak and turn knobs to tune the business." The problem is that with sufficient data engineering and analysis, you are often only a few steps away from a completely automated and mathematically provable optimality...one that "genius" business leaders can only attempt to replicate. And instead of focusing on the optimization problem, the engineers and analysts instead focus on data presentation and visualization, in order to please the demanding eyes of the genius autocrats. Of course, not all data analytics are like this. Some are legitimately useful as warnings or alarms, some are useful as ad hoc analyses, some are useful as reporting and accountability. But far too many run into the cockpit fetish problem, and end up as a massive and suboptimal waste of time at their very best.