The Secret Behind Failed Business Analytics: The Missing Link

Business analytics projects tend to make big promises, such as revealing actionable insights and facilitating more predictive and prescriptive decision-making. Regrettably, many of these projects end up failing to fulfill their goal. They analyze data and highlight insights.  But all too often, they don’t manage to go beyond the surface level of the obvious, which is rather disappointing. Why is it that only a handful of business analytics solutions manage to go deeper than the tip of the iceberg, and pinpoint hidden, even surprising insights?  What is the secret behind those solutions?

Implementing a business analytics solution is not an easy journey. Projects have to overcome the hurdles of deployment and integration, as well as skepticism and political pressures, before they see the light of day. Many articles have been written on the numerous obstacles to bring business analytics projects to successful completion. The question is: why, after overcoming all these difficulties, are analysis results so poor? Can we do better than that?

It is common to blame the GIGO effect – Garbage In = Garbage Out –pointing at the data as flawed, old, missing, etc. However, in most cases it is the analytics model that is the one to blame, and not the data. Available business data is abundant, even overwhelming. Many analytical models narrow down the scope of their analytics and make a priori assumptions in an attempt to arrive at quick conclusions.  It works well for the common cases.  For example, did my marketing campaign influence my sales? Comparing Marketing and Sales figures will provide some pretty good answers. But is that all there is to it? Maybe a competitor changed its pricing? Maybe some other product cannibalized the promoted one? Maybe a new regulation created an unexpected hurdle in certain geographies? If we don’t look into all possibilities we’ll never know if there were any other influences, and are likely to arrive at greatly over simplified conclusions.

To avoid flawed insights that lead to futile decisions, we have to ensure that every bit of available business data is taken into account. Don’t analyze in a silo; don’t exclude data that doesn’t seem relevant. Rather, put it all together and only then, use a powerful, automatic analysis tool that will find correlations and interdependencies, between expected as well as unexpected phenomena.

With today’s powerful computing technology, we can automate the process and take business analytics to the next level, to find concealed insights that will give us an edge in managing our business. But even the most sophisticated solution cannot analyze data it doesn’t see. To ensure the most insightful analysis, start your business analytics with integration of all influences into one big picture. Don’t make any a priori assumptions that leave data out. There are thousands of possible correlations to be examined. Results may just surprise you.