In every survey of business leaders you read, exploiting data for more effective decision making is rated as one of the TOP priorities.

This is reflected in investments in data architectures and technologies.

But the same business leaders report that the effectiveness of data driven decision making in their organizations remains low.

Why is that?

There are various hypotheses for this but the one that fascinates me, the “elephant in the room”, is the fact that irrespective of the best architectures and tools, the business questions that need to be asked are nuanced and crafting the data insights that support effective decision making is tricky.

There is enormous value in data insights, channelled to the right people at the right time, all the time, to support continuous day-to-day process improvement in every business value chain.

These data insights need to be focused on “defects” not performance outcomes. It is the root cause analysis of process defects that drives meaningful process improvement. These defect oriented strategies also touch on “Shift Left” thinking, which I describe in many of my musings.

All the evidence (“aggregation of marginal gains” et al) demonstrates that the value of continuous improvement is far higher than that from “big bang” transformational changes. We need the big changes sometimes too, but we shouldn’t let that obscure reality.

When we are quick to offer oceans of “data” for ad-hoc query by our customers and stakeholders, we often make the situation worse.

The business questions are nuanced (as is the data), and correlation of data insights does not necessarily lead to good root cause analysis.

My favourite example of this (and there are many in our businesses also) is the data bias observed by Abraham Wald at Columbia University that was leading planners and strategists to the opposite conclusion to that required.

The exact opposite. A completely unproductive conclusion.

It happens every day.

It is a very brief read here . . . . Just 60 seconds, but you will still be thinking about it over the next few days.

This is why we need better business AND technology strategies for data driven decision making, than just providing the complete enterprise data set and smart tools for visualisation and correlation.

Thanks for reading . . .