Is data science STILL the sexiest job of the 21st century?

When Forbes coined the phrase “sexiest job of the 21st century” to describe data science, way back in 2012, the role was in its infancy.

At the time, the ability to discover new insights from data, using a combination of statistics, machine learning and computer programming, was expected to be the new competitive differentiator. Data scientists, the rare breed that combines these skills with business insight, were expected to be in high demand and draw high salaries.

The demand for data scientists

Certainly, in South Africa, jobs are available for data scientists, and, in most cases, pay decent salaries.

Yet, in many cases, the shortage of expertise is leading corporates in South Africa to employ graduates with little or no work experience.

The goal is to build data science capability at a relatively low cost.

Finding data the biggest time sink

And, in many cases, the promise of discovering the next big business wave is replaced by the drudgery of data management.

One new data scientist, four months into his first job, with a listed bank, complains that rather than doing data science, he is spending all of his time simply trying to identify and source data.

He is not alone!

In conversations with his university classmates, this is a common experience.

Privacy concerns

Privacy and similar legislative concerns may mean that months are spent negotiating to access data once it is identified. In the absence of sound data governance, many organisations err on the side of caution, refusing to share operational or customer data with the data science team.

And, once the data is delivered data quality issues may make the data untenable for meaningful analysis.

Another data scientist, working in the insurance industry suggested that data cleansing and preparation took up around 90% of her time, rather than the generally accepted average of 60%.

Data scientists leave before they add value

Certainly, in South Africa, data governance and data quality issues mean that data science is taking a back seat to data management. These frustrations mean that many data scientists move jobs before they add significant value.

In the meantime, some research suggests that data science is a prime candidate for automation – suggesting that this career path may not be long-term.

Is data science still the sexiest job of the 21st century?

What do you think?

Response to “Is data science STILL the sexiest job of the 21st century?”

Mike Palmer

April 25

Data quality is certainly a big issue, but there’s no reason why it should occupy so much of someone’s time. There are a lot of solutions available to cleanse, match and validate data. Granted, some of them are expensive and slow, and you have to balance that against the cost of someone manually checking and correcting data. In some cases, it may well be cheaper to employ a recent graduate than buy data management software or services from the likes of IBM. But dig a little deeper and you’ll find some real gems.

DataMatch Enterprise from Data Ladder, for example, is fast, accurate and inexpensive. In an independent test run by an Australian University, DataMatch Enterprise outperformed IBM and SAS data quality tools in terms of speed and accuracy. And the cost is much less than a data science graduate.