THE DATA DRIVEN DECISION MAKING - The Alternative Approach

There is no denying that data science has increasingly gained a prominent place in several companies around the world and the reasons could not be more obvious. Companies nowadays, especially the bigger ones, generate huge amounts of data at a speed never seen before. And for that reason the demand for qualified professionals to analyze this data and generate valuable insights is growing at the same pace. Nowadays it´s common than ever, to see posted job adverts with titles such as “data scientist”, “business analyst”, “insight analyst” and other jobs where the main function is to transform available data into insights that the upper management will use to make informed business decisions.

And with this growing demand for data professionals, more and more people are looking to specialize in this area that promises to be one of the most prominent in the market during the days to come, and with the amount of resources we have available today, this exercise has become easier than ever. Companies like Udacity, Coursera and others offer a multitude of courses related to data science and often these courses are developed with industry leading partners such as IBM, Google and Microsoft, which often ensure that the technologies they teach are the are the ones that are going to be used by the industry in years to come.

But where does the biggest mistake lie?

The biggest mistake lies precisely in the fact that many data science aspirants do not give due importance or simply ignore the fundamentals. When I talk about fundamentals I mean the old and basic statistical concepts that many of us have learned years before. And that, whether you believe or not, is the foundation of the data science we know today as a discipline.

During my years as a working professional in the Mining Industry, an industry that generates massive amounts of data in a day-to-day basis, I can definitely tell that the focus of many aspirants before they dive deep into data science, is usually whether to use Tableau, Power BI; or whether to program using Python vs R, basically putting more focus on the tools and completely ignoring the theoretical part of the discipline. And while those questions are all valid, I can say with absolute certainty that as long as you don´t have solid foundations in place, the odds of not asking your data the right questions become astronomical, and wrong assumptions lead to wrong conclusions, that in the end result in very beautiful dashboards that will add less or nothing to the business, or in the worst case scenario will lead to wrong business decisions.

Of course there are other mistakes in the mix!

Using low quality/unreliable data, using the wrong processing methodology or even choosing the wrong visualizations for presenting a given information are other common mistakes that can lead to wrong interpretation and poor decision making as well, but since all these errors are mainly technical, they are often easy overcome, especially when the professional has a very solid foundation of this important and exciting discipline that is data science.

Image Credits: Marketoonist by Tom Fishburne