Recent years have brought an explosion of data science jobs, and the demand is on the rise. Data analysts, data miners, Big Data scientists, and more similar job titles populate online job boards now. This is a new domain, and very few people, if any, occupying these positions have an academic qualification in this exact field.
Most data scientists come from other jobs or entirely different career paths. Some of them were engineers; others were programmers, statisticians or mathematicians. Mostly any job which required a solid understanding of logic and statistics can act as a launching point for a career in data science. Here is an outlook on a few trends identified, proposed and analyzed by the likes of PwC, Forbes and InData Labs.
As most companies are shifting toward a data-driven approach, there is a higher demand for data scientists, either for in-house development or for software providers offering SaaS products. There is a definite scarcity of these specialists; they need to have a rare combination of skills and a multidisciplinary thinking which has not been encouraged by most corporate environments.
While some companies are already ahead of the game and have attracted and trained top talent from around the world, newcomers to this industry will have a hard time sourcing the right people, and this will come at a very high cost.
The ideal candidate for these jobs is what Tim Brown of IDEO calls a T-shaped individual. This means a person who has a vast experience in a niche subject, but enough curiosity and personal skills to integrate that in a broader context. They are gurus in their area of know-how but are well-read and inquisitive so that they can work with others and create value.
Companies recruiting for data science jobs should be prepared to wait until they fill these positions for longer than when they are filling more general roles.
Another reason why candidates for data science jobs are hard to find in a typical corporate environment is that this multidisciplinary approach is more characteristic to entrepreneurs. They need to connect the dots, come up with ideas which have not been tested before and innovate.
Prior experience in number-dominated fields helps but is not enough. Also, the best results are achieved by those who have an inquisitive approach to problem-solving, instead of just using the beaten track.
Attracting or growing such talent will be a challenge for most organizations, as these people are more set to create their startups instead of spending time on common office jobs.
As data-centered skills are in demand, the academic world is catching up and designing data science programs. When these get traction, it will be presumably easier to grow such specialists, but for the next years, the dominant strategy will be to convert specialists from other related fields.
While it’s almost impossible to find the right candidate on the market, some companies will have to resort to recruiting from within the organization and help such hires to develop the necessary abilities.
The initial requirements for education are quite high. All data science jobs require a B.Sc. in Engineering, Mathematics or related fields. Some positions may even require skills associated with Ph.D.-level studies. Good news is that these skills can be trained, and knowing the internal workflows and problems helps once a solid computational base is set.
So when preparing your HR budgets for the next year, set aside a consistent amount for continuous education and training of your in-house data scientists.
The set of analytical skills required by data science jobs is usually found in people who have already completed academic degrees. Therefore, it is reasonable to expect to see a migration of the workforce from academia to the business world, also driven by higher pay rates.
The only problem of these latecomers, compared to their peers from engineering or other applied fields, is that some of them miss the contact with real-world issues and business context specifically. However, as such hires have the framework in place, they will get used to the nature of data tasks set by their corporate employers.
The world of job titles is confusing right now since many organizations tend to use titles as a way of employee gratification. It is not unusual to see employees in mid-level positions called VP just to give them an ego boost instead of a financial one.
In the world of data science, it gets more confusing due to the novelty of the job requirements. It is not uncommon to call a data scientist any of the following roles: those who perform SQL queries, data cleaners, machine learning roles (from architects to testers) and researchers. As you can imagine, each of these is, in fact, a separate job. However, until now these roles were not distinctive enough to get a title on their own.
As data science makes its leaps into the future, there will be less demarcation between data scientists per se and other roles, such as product managers. This already happens in companies like Quora, which are at the forefront of online innovation when it comes to data applications.
No talk about the future of data science jobs is complete without mentioning the impact on other jobs. As automation scale and pace increase, data scientists are like to put others out of their jobs. Although there will be no shortage of work in the future, there will be some imbalance between highly skilled positions and the low-end of the working spectrum.