What is true automation for analytics and is it desirable?

The word automation is often music to the ears of business leaders. Automation brings with itself the promise that things will run on their own, as cheaply as possible [at least over a longer term], and as consistently as they should.

Given this, it is hardly surprising that clients are looking for automation for analytics platforms. Before we go further, let us define what true automation for analytics means.

If we think of analytics value chain as a series of steps from “Data Generation” through to “Data Driven Business Decision”, then true automation should enable all these steps to be carried out without human intervention. It is important to emphasize that in most cases, the first two steps are anyway automated [almost by necessity].

Automation for Analytics


Sometimes when I speak with my clients and prospects, I find them confused about whether they need to invest in building a piece of software [i.e. a machine based capability] or a human capability. In other words, they are not clear how much should they [or can they] automate?

Businesses spend enormous amounts of time and money seeking the right solution for their business problems when they haven’t even framed the business problem correctly.

I am hoping to provide a simple framework that enables decision makers understand and determine what they need in order to achieve their business objectives.

[Re]Framing the Analytics Problem

Formulating the right approach for analytics in this context is a function of understanding and evaluating business requirements along two dimensions:

  1. Frequency: How often is the data generated?
  2. Value: What is the potential value of insight generated as a result of analytics?

The value can be positive (upside from the “right decision”) as well as negative (downside from a “wrong decision”). It is important to emphasize that it is the value of each individual decision or transaction that matters, the collective value at stake may be large due to sheer volumes involved.

The interplay of the two dimensions, and the proposed approach that you should take to your analytics solution is shown below:

Analytics Strategy Framework

Of course, there are other factors also that you need to consider (e.g. do you have the scale to justify investments into building a system?) but arguably they are easier to answer once you have answered the most fundamental questions.

Analyzing your analytics requirements through the lens of value and frequency can also help you adopt a more pragmatic execution strategy. For more details on this, please see my previous post: 5 Things to Remember When Formulating Your Analytics Strategy.