Subscribe to DSC Newsletter

You hear all the time that it is very hard to find and hire a great data scientist. Yet these scientists can't find a job, and are typically unemployed for many months after graduating or after being laid off - even those with 15 years of experience and stellar degree and accomplishments. 

So what's the problem? I'm suggesting a few possibilities:

  • Candidates lack business acumen
  • Candidates have poor communication skills, are asking too much money
  • Candidates have outdated skills - or University is teaching outdated material
  • Recruiters are very slow in the recruiting process - eventually candidates evaporate
  • Recruiters are very concerned about hiring because of the economy
  • Candidates should not apply for a job, but instead create their own career or enterprise
  • Recruiters don't know how to measure added value provided by analytic talent
  • Certifications, regulations (e.g. regarding data privacy) or US citizenship requirements is a barrier
  • Candidates will not relocate because they can't sell their house
  • Too many analytic people, we should discourage prospective students to pursue an analytic career

What do you think?

Views: 4307


You need to be a member of AnalyticBridge to add comments!

Join AnalyticBridge

Comment by Edmund Freeman on October 1, 2011 at 9:48am

I'm thinking about my own experience.


My first professional job was very difficult to find -- well over a year. Since then (15 years) I've never been out of work. It's never been that hard to find a new job, and sometimes jobs have come and found me. Part of that has been luck, living in the Puget Sound where there are a lot of analytic companies.

One aspect -- I've always been flexible about roles if I like the company. I've avoided putting myself in a particular box.

Comment by Vincent Granville on September 30, 2011 at 2:25pm

See another post on the subject:

I think candidates can feel there's no job because they are still looking in the wrong places (big job boards) or still act like 10 years ago: send a resume, and wait. This approach does not work anymore. And recruiters can have a feeling that candidates are rare because they've moved to different places (like Analyticbridge!) where candidates are of much better quality. Indeed, smart candidates post great articles and get hired without ever submitting a resume or applying for a job.

But barriers exist, e.g. for some analytic jobs you must be a US citizen. It makes it very hard, if you are recruiting analytic people with security clearance, to find available people. But in this case, the scarcity is created by artificial conditions (regulation).

Comment by Rick Wicklin on September 30, 2011 at 11:18am

@PeterWilliams SAS is a big company and we hire many, many, people with high-level programming skills. In fact, all of the SAS "solutions" are written using SAS, so obviously SAS programming skills are essential for those positions! However, we also hire people to program the algorithms and analyses that SAS surfaces to customers. Those positions require strong C/C++ skills, such as knowing how (and when) to call an SVD or gradient descent algorithm from libraries such as IMSL, LAPACK, etc. That is a much rarer skill, because grad students now use SAS and R for many tasks that (20 years ago) used to require C or FORTRAN. 


But I suspect this is not an issue related to @Capri's question. Most companies prefer the data scientist to have the high-level skills, because those are the skills that people use to analyze data.

Comment by Whitney K on September 30, 2011 at 10:23am

Are we really serious to think that the job market is saturated with too many people with analytical skills?  I find that hard to believe.  There are in fact NOT enough students pursuing a career in analytics, further highlighted by the MsKinsey Study on Big Data. The shortages in employees with data/analytical skills will be greater in the years to come.  This is being remedied by the number of new programs (usually distance learning programs) that offer degrees in Data Mining/Business Analytics.  There are a handful of universities now catching up to this demand for these skills and creating these programs.  

Maybe I'm out of touch with reality since I'm not currently pursuing a new job in analytics.  So is the market really that tough for analytics professionals?  

Comment by Jozo Kovac on September 21, 2011 at 2:59pm

Vincent you are right, with proper skills analytics expert can become enterprenuer. 

But who are his customers? 

Small companies may have troubles to pay his rates. Large may not want to do business with him, if he's not well known and unique. Medium companies can face multiple issues (data, strategy, budget ...) what need a small army or long time to be resolved.

Every problem has its solution. But still Analytics people have easier life as freelancers, pernament employee, consultants or researchers. Saying nothing about rewards.

Comment by Vincent Granville on September 20, 2011 at 8:58pm
Another point of view. Should analytic people stop looking for jobs, but instead create their own jobs: sell data, sell competitive intelligence, sell AaaA, and become entrepreneurs? After all most analytic professionals understand many things and can play several roles: sales, accounting, optimization, CRM, software engineering, marketing, social networks, ROI / finance, fraud detection, legal stuff, PR, etc. making them good candidates to become entrepreneurs without having to spend or raise money to launch their business.
Comment by Peter Williams on September 19, 2011 at 8:08pm


That is interesting.

Is the demand all for people with statistics and "low level" programming skills?

I am not seeing so much of that in Australia. There seems to be more demand here for statistics people with high level programming languages like R and SAS

Do you want people who can code at the level of SVD, gradient descent and the like?

Comment by Rick Wicklin on September 19, 2011 at 11:12am
At SAS, one of the difficulties are finding candidates that have depth in statistics and also demonstrated computational abilities in a low-level language like C/C++, FORTRAN, etc. There are many smart candidates graduating from statistics departments, but few have the computational skills that we are looking for.
Comment by Jozo Kovac on September 19, 2011 at 3:59am

Maybe problem lies in definition of what "Great data scientist" means.

I'm sure there's a huge difference between company management's point of view and real scientist's perspective.

Comment by Peter Williams on September 18, 2011 at 7:57pm

Have you double-checked your assumption about it being hard to hire data scientists?

Are you sure that the demand is for great data scientists as you state? If so then maybe the unemployed data scientists are not perceived as being great.

On Data Science Central

© 2021   TechTarget, Inc.   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service