The softer costs of being a Data Scientist

November 23, 2020

Data Science is hot. Everyone wants to learn and become a Data Scientist. The path is pretty straightforward – learn some linear algebra, basic calculus, inferential statistics, probability, Tensorflow & PyTorch, and you are done. Voila, a new Data Scientist is born!

True, but what does a data scientist really produce? Hopefully useful insights from data that were not available to people till then. Now how do you think people generally take to “Truth” pronounced by a geek from looking at some data. In my experience not so well at all. General responses have ranged from the mild “The person does not know what he is talking about, just ignore him/her” to being branded as a “Snake Oil” salesman and being derided across the Enterprise.

It is never easy telling people that they are wrong and that there is a better way to do things, especially when things have been done in a certain way for years together and the company has been reasonably successful through those years.

Now if we want to succeed as a Data Scientist and have a reasonably long career, we have to build this muscle, the ability to face rejections and mockery, and still persist with the insights from data. I am finding more and more that folks coming into this field of data-intelligence do not seem to be aware of this. They seem to think their responsibility ends with writing some sophisticated python code that spits out an “intelligent truth”! Unfortunately, that is where the story actually begins.

By the way, it is not all gloom, there are ample examples where “data-based” insights have been accepted and implemented with exponential benefits. The key for folks working in the “data-intelligence” domain is to persist and ensure that frontline business folks accept and implement the same.