Master LLMs with our FREE course in collaboration with Activeloop & Intel Disruptor Initiative. Join now!


Data Scientists might become a commodity — High time to mend ways……

Data Scientists might become a commodity — High time to mend ways……

Last Updated on August 2, 2021 by Editorial Team

Author(s): Supriya Ghosh


Data Scientists might become a commodity — High time to mend ways……

Image by Author

Data science is ‘the hottest job of the 21st century with an above-average paid salaries.

According to AI & Data Science Salary Study 2021 by AIM Research, the salary of a data science/analytics professional is 44 percent higher than that of a software developer and 36 percent higher than that of an IT developer.

Companies and Regions across the globe are offering a handsome package to hire skilled data, science professionals. Almost all industries are utilizing data science in one way, or the other, and data scientists are in demand in most countries. The topmost countries paying utmost to the Data Scientists are the USA, Switzerland, Norway, Australia, Canada, Germany, South Africa, France, Netherlands, and the UK. The list by no means is exhaustive.

But a big question is, does that mean it will always remain the same? In fact, this is a topic of debate.

Today data science and machine learning has brought a wave of transformation which a decade back nobody could ever imagine. Most of the companies are now at least knee-deep into using data science technology and applications.

Source —

Data Science has a myriad of applications in almost all fields and areas, like GIS, Defense systems, product recommendation systems, business decision making, intelligent vehicles, computer vision, health care, retail, e-commerce, manufacturing, telecom, aviation, smart homes, smart factory, smart offices and building, banking, etc. The list goes really long.

Value-adding analytical capabilities are spreading their leg in every sphere of activity with new tools and techniques coming up in an unusually short time. The saying goes as “Necessity is the mother of invention”. Hence the new problem is accompanied by a new invention, and this is all done very quickly. Organizations have unparalleled access to the components necessary to build these advanced capabilities. As costs and fees for these components are dropping day by day, it is making consumption of analytics easier and less expensive — that is, more and more organizations are able to incorporate analytics into their processes.

Source —

As analytics is becoming readily available to all organizations, so are the Data scientists just like commodities.

What does this mean?

To understand this, let me explain commodity at first.

Anything that can be easily traded (bought or sold) on the commodity market is a commodity.

And every hot technology and skill ultimately transition from being an emerging to a commodity if not utilized wisely.

Source —

Data Scientists are tending to take on the characteristics of a commodity in the coming years. With growth at its peak in this field, everybody now wants to join this scientific rat race without being left behind.

This reminds me of a dialogue from the movie “Incredibles” which I will paraphrase as:

“Everyone can be super or special !!!

And when everyone is super or special, then no one is.”

As more companies are adopting and adapting analytics, what will the forthcoming commoditization of data scientists mean?

There are high chances that future data scientists might be traded like a commodity.

The thought of such an era is fearful and needs deeper analysis and investigation in itself.

Imagine what will happen when data scientists are not able to offer new and distinguished perspectives anymore?

By new and different perspectives, it means a new set of outlooks and benefits other than the existing proven ones.

Data will continue to get generated multifold but techniques for processing such data might become stagnant and may not offer additional distinguished perspectives because of the amateur skills of the data scientists which may remain unmatched with the algorithms, tools, and techniques.

Source —

Let me elaborate more.

One of the latest visible trends is that many of today’s data scientists have come into the field for handsome salaries without giving a second thought. Unfortunately, they are driven by the glamour quotient and monetary benefits. Data science is about defining and solving business problems but on one hand, they fail to understand or appreciate the business aspects of problems and on another hand, they lack exposure to the engineering aspects of developing solutions as well as domain knowledge.


What can be done differently to avoid this situation?

How can future data scientists shape themselves so that not to turn themselves obsolete?

What skills do they need to invest in?

Below is the non-exhaustive list.

1. Understanding the business aspects of problems as eventually everything needs to be realized in terms of business benefits.

2. Developing domain knowledge

3. Learning, thinking, and talking business language. Most of the time, simple business rules learnt adds much value.

4. Not focusing on fashionable techniques without having a solid foundation in algorithms, mathematics, and statistics. E.g., Many data scientists assume they only need to understand deep learning without having a comprehensive understanding of what assumptions are made about data, how they need to be evaluated, what impact turning parameters has, where exactly deep learning outweighs machine learning, etc.)

5. Not using deep learning when one doesn’t need it, and in 80–90% of business cases, data scientists don’t need it.

6. Always being mindful and gaining problem-solving skills more and more.

7. Learning every day — being like a Learning Machine but remembering to take baby steps and follow Kaizen Principle.

8. Spending most of the time understanding the domain, business problems, data, and how to add value with analytics.

9. Prediction is easy, creating business impact is hard. Hence additionally, teaming up with the marketing team to deeply understand the business context, customer needs, etc.

10.Many a time sacrificing model’s performance for better business performance.

11.Providing recommendations for insightful actions as recommendations can be commodities but action is not.

12.Always focusing on decisions to be influenced/improved.

13.Always exploring opportunities to add value from analytics.

14.Focusing on selection criteria.

15.Focusing on success criteria or KPIs.

With this, I conclude. This is entirely my perspective which I have mentioned. Please do not consider this as a final and bound reality. This topic anyways remains open for long debates.

Data Scientists might become a commodity — High time to mend ways…… was originally published in Towards AI on Medium, where people are continuing the conversation by highlighting and responding to this story.

Published via Towards AI

Feedback ↓