The technology sphere is bursting with demand for data science roles in the market space. This comes as no surprise considering data science was listed as the sexiest job of the 21st century by the HBR. Analytics and data science roles have become the most sought-after jobs today, given the market offerings of higher pays than ever before, complemented by huge volumes of jobs at the entry-level.
This phenomenon is even more prominent in India due to the already high numbers of engineering graduates. This comes with the misalignment of ideas of what you want to do instead of what makes practical sense to land a good job. It wouldn’t be wrong to say that a high volume of engineering graduates aspire to enter this field naturally. Gradually, a higher proportion of folks from other related fields like statistics and economics have also made their way into the world of analytics and data science.
Sign up for your weekly dose of what's up in emerging technology.
The demand has picked up so exponentially that thousands of aspirants from various other unrelated fields are also willing to put in the efforts and risk to make a career transition to learn the pre-requisites and make their way to land a data science job.
With that comes the million-dollar question – “Can anyone become a data scientist?” followed by “How to land a data science job?”
The short answer is yes. However, one needs to consider several factors before that.
Data science is a diverse field that is still evolving- leading to challenges related to understanding the field and the several nuances that are involved. In this section, we will discuss some of the major challenges and how people from non-technical backgrounds can overcome them.
The first challenge is the complexity involved in defining and understanding data science in its entirety. With improper guidance, many folks tend to believe that data science is just learning Python or R, or learning a few algorithms, or just building data visualisations and dashboards. While all of these are not incorrect, they are not all rounded definitions either.
Data science is a vast field that requires multiple aspects to become what can be referred to as a “Full stack applied data scientist”, and honestly, there are not many who will fit the bill 100%. This is purely due to the evolving nature of this field, and that is fine. Draft and plan how you want to shape your profile and gradually add skills to the kitty based on your target jobs.
The second challenge is the approach and process involved in ramping up oneself to land a target role. We are already swamped by tons of advertisements claiming to make you a data scientist in 6 months or, in some cases, even a month. Yet, folks who have spent more than a decade in the industry would tell you that they still believe they know 50-60% of the numerous learning possibilities that exist.
Thus, the basic approach should be to get the foundational elements right. Make it so strong that you are confident in your ability to learn, understand and pick up any alpha beta gamma that may come up tomorrow. By the time you finish reading this article, there may be a few more analytics approaches, a few more variations of ML algorithms, few more data visualisations that would have been created, but if the foundations are strong, the hurdle to learn them as and when needed becomes an easier one to cross. So, pick up the right learning sources and focus on strengthening the selected elements needed for a target role.
That brings us to the third challenge – understanding the job description itself. This field is so vast and evolving so rapidly that it is tough to define it as one standardised job. While the title data scientist is most desired (and most popular), it is critical not to ignore roles such as business analyst, ML engineer, data engineer, BI architect and data analyst. And more importantly, who can be called a data scientist is a blurry line across organisations, teams and job profiles.
Make it a point to clearly understand the job roles and responsibilities than going by titles and designations. This will help you select which roles are suited best for you and what should be the target.
So how does one solve these challenges to transition to a career in data science and land a dream job?
Business problems have been present for years; data science and analytics are just a means of solving these better, the means to an end. So it is critical never to forget that while working on a problem. There are three technical skills that any aspiring data scientist needs to develop to work on a business problem: business skills, mathematical skills, and technological skills.
Let’s look at how you can gain and improve these skills.
The real trick to becoming a business problem solver is reading case studies and understanding various processes. Engage in activities like trying to understand the generated data, thinking of ways to analyse the data and using that to determine what outcomes could be obtained in this business problem using data science.
The actual portion is not much different from what you learned in high school when it comes to Math. Start by revising your high school mathematics and statistics but take a more understanding approach. Data science is not about mugging up algorithms and formulas that one can blindly execute—focus on understanding the backend workings of statistical models and algorithms. And most importantly, practice identifying what to use & when to use it, given the scenario of data and insights/outcomes needed. Do this to train yourself in solving complex scenarios when something fails rather than plain vanilla ideal world execution.
To start with upskilling yourself in technology, pick up one data engineering, data science, and data visualisation platform. SQL – Python – Tableau/ Power BI is the recommended combination. Practice as much as possible. Troubleshoot for various complex scenarios when a particular package needs to be tweaked, or a complex data operation/transformation needs to be carried out. Do this while remembering the basic syntaxes and packages.
Along with the technical skills are the three essential soft skills that you should build. Technical skills can only get you so far, making it important to communicate them or make decisions based on your findings.
While learning design thinking, it is important to remember that you are building it for the end-user who will consume the solution, given that design and craft end to end business solutions rather than one-off data tasks or analysis or model building.
Build your decision-making ability by not stopping at the execution level. Instead, interpret what the numbers tell, question the outputs and solve problems for scenarios when there is a judgment based on the model outputs.
Communication is one of the most crucial soft skills to attain, and while it is not difficult, it is important to upskill yourself in being a better communicator. Start by practising. Practice presenting solutions, explaining complex algorithms to a business audience, creating readouts for the technical audience and giving mock interviews. My trick is- if you can explain it to your grandmother, you have understood it well enough.
Once you have got the first two aspects covered, the challenge of understanding job roles is quite easy to solve. Read job descriptions, talk to people from the same org/team to understand the job roles and responsibilities. Then, understand what the day in the life of a person in that role in that company looks like.
Once you have understood the life cycle, focus on identifying where you will play a “responsible”, “accountable”, “consulted”, & “informed” role.
Accordingly, select the roles to target, the skills needed and then try to fill the gap by picking up learning modules across aspects suggested in the solutions to the first two challenges. Understand how the titles are structured across different types of organisations and teams and pick the ones that are up to your alley based on skills, expectations and growth opportunities to spread the spectrum towards a “full-stack applied data scientist.”
This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill the form here.