We all probably agree that data science is probably one of the fastest-changing fields in US today, and what was state-of-the-art last year might have subtly become obsolete.  

Some of these trends are hard to miss – for example the growing supply of (not always prime quality) data scientists from some less-than-reputable bootcamps, as well as the general rise in the number (and scope) of companies comfortably talking about how big data or artificial intelligence is in their very immediate plans.

 

Everyone’s a Data Scientist

However, there’s a less-discussed trend that could still have major ramifications for the data scientist hiring market in 2017 and beyond – data scientists are coming from a wider range of educational and professional backgrounds.  

This is always been a factor to some extent, but especially in 2017 as some data scientist training programs are making major efforts to promote the data scientist career path in general, the diversity of people’s backgrounds within data science has probably never been larger.

This can be both good and bad – the good news is that it doesn’t hurt to be drawing from the expertise of people with very non-standard perspectives, especially when attacking projects and problems that have remained stagnant for some time.  

 

Less Qualified Candidates

However, the less optimistic perspective is the barrier to entry has been substantially lowered – it’s not at all uncommon to see people with essentially no real data scientist training or background to be marketing themselves as such.  

Longer term, this could lead to a shrinking gap in pay between data analysts and data scientists.  In any case, some employers (especially from larger companies) have started to become much more stringent on formal or traditional requirements for data scientist positions, for better or worse.

It’s an odd situation as probably the vast majority of companies, when asked if there is a large enough of qualified data scientists available, most would say no, there is not enough.  

Demand for data-savvy employees is far outstripping the available supply -McKinsey Global Institute Report, 2016

However, the key words here being ‘qualified’ and ‘available’ – most of those same companies would say that that there is no lack of data scientists that are either (a) available or (b) qualified.  In other words, in the dating world, we’d say there are a lot of single people, a lot of attractive people, but not a lot of single, attractive people.

 

Companies Raising the Bar

So why are many companies, who are struggling to attract top data scientists, simultaneously raising the bar?  Well for one, they don’t really have a choice – many would agree that making a bad hire is worse than making no hire at all.  

On the other hand, one has to wonder why, if companies are complaining so much about the lack of available talent in the data scientist market, we haven’t yet seen a marked upward shift in pay in the upper half of the data scientist market.

In any case, going forward, we can expect to see a steady stream of new people claiming to be highly qualified data scientists, and a steady stream of companies discussing the acute shortage of available talent.  I personally don’t have a solution, but one thing I’d like to see is the average duration of data science bootcamps to at least double.  

The complaints that many large companies have about the skills of data science bootcamp grads aren’t of some mystical variety – the (at least partial) fix could be adding substantially more time into the training programs to more thoroughly teach the fundamentals of computer science.  

Realistically, we probably won’t see much short-term resolution unless the students are willing to pay much higher tuition rates or the salary-based-tuition-repayment programs are willing to take on much more duration risk.