Over the last decade, businesses have seen a fundamental shift in the way decisions are made. Winning organisations have moved from an environment, in which personal experience and intuition is king, to a data driven world where decisions have support from quantifiable historic evidence. We now sit on the precipice of the next transformation, the movement from simple to advanced analytics. This change, like any other, requires thought provoking questions to be asked on both a macro and micro level. A key question transformation leaders should be asking is, “Does the movement from simple to advanced analytics make sense, or is there another route we should consider?”

With the proliferation of online courses, highly publicised successes and general media hype, advanced analytics (marketed in varying forms) has taken centre stage as business’s next evolutionary step. The hype has positives, as businesses that saw little value in their data before are beginning to place trust in it, and are reaping the benefits. However, the other side of the coin is not as shiny. The hype is also creating expectations that are leaving businesses frustrated and many employees demotivated.

In the attempt to increase sales and trust, businesses have jumped on the bandwagon of ‘being on the cutting edge of AI and analytics’_. _Whether this strategy is successful for boosting sales is unclear, however it does create expectations for new, analytically gifted resources joining the business. For the sake of simplicity I will refer to these resources as ‘Lukes’. Many Lukes join AI driven businesses wanting to “stretch their legs” and build the latest and greatest predictive models (which is understandable as these models are technically interesting and increase your street cred). Unfortunately, Lukes are unaware of the training that needs to be done and the seeming impracticability of their core technical skills in business. This causes frustration and demotivation within the Luke faction. Frustrations are exacerbated by the decrease in technical understanding with the increase in business seniority.

Regardless how businesses market their analytical capabilities, there is one common truth amongst them. Businesses want to get actionable and believable insights from their data. Little value is placed on the methodologies or techniques used. Instead, true value is gained from analytical capabilities when the insights extracted from data allows for superior solutions (especially in competitive markets). When considering businesses operating in a successful data-driven environment, you will find several business leaders who are particularly good at extracting these insights at minimal levels of complexity. I shall refer to these leaders as ‘Yodas’ from here on out.

In order to curb frustrations, Lukes are provided significant problem solving freedom, with business leaders expecting limited value from their solutions. This approach does reduce frustration in the short-term, however, results in a large portion of analytics work no longer being ‘effective’ and usable by Yodas or business.

Effective Analytics

In order to balance the tradeoff between the usefulness of analytics work for Yodas and frustration among Lukes, the conceptual framework of ‘Effective Analytics’ is proposed.

To introduce the idea of Effective Analytics imagine that for every problem there are an infinite number of possible analytical solutions. Further, this plain of solutions ranges from simple to advanced/complex.

Figure 1. Plain where all analytical solutions exist.

The solution of every problem will fall somewhere on this plain, with the exact point depending on the relationship between the Yoda and the Luke. Yoda naturally prefers simpler, more understandable solutions, whereas Luke favours more complex, advanced solutions. The precise location of where the solution lands will depend on the ‘strength of force’ exhibited by each party. The ‘strength of force’ for each party will depend on their respective credibility, as well as, the business importance of the problem.

#analytics #data-driven #business #leadership #data-science #data analysis