Subscribe to DSC Newsletter

27 types of data scientists - where do you fit?

Three metrics can be used to segment the population of data scientists. Each metric has three levels: high, medium, low. Hence the 27 (= 3 * 3 * 3) types of scientists.

Here are the metrics in question:

  1. Soft skills: sales, business acumen, communications
  2. Leardeship: vision, intuition, guessing/interpreting client needs, big picture - strategy oriented vs. tactical / detailed oriented
  3. Knowledge: experience, craftsmanship, broad and deep knowledge (deep in some areas) vs. specialized only or absent.

Where do you fit? Employers tend to hire among a subset of those who rank high on all three metrics (if you are high on all three metrics but independent and making good money, you won't be hired, but you are not looking for a job anyway).

Our survey about "how hard it is to find a data scientist position" revealed that 70% of analytic people find it difficult to impossible, see .

I'm interested in providing training to help candidates acquire the skills /craftsmanship required. More on this later.

Views: 7326


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

Join AnalyticBridge

Comment by Dorothy Hewitt-Sanchez on March 14, 2012 at 1:25pm

I am commenting on your statement: (if you are high on all three metrics but independent and making good money, you won't be hired, but you are not looking for a job anyway).


Once upon a time I had a wonderful career but the market just stop for me.   I was independent and making good money.  So why does that count against me?  Why do people feel that I do not need a job?  I kid you not, I was in an interview and the interviewer told me,” You do not need a job”.  I was so hurt because I do not understand why she would say this.  So I feel that people are not hiring me because they have this crazy idea that I do not need a job.  Am basically pushed into poverty because people are saying I do not need a job.  This does not make sense to me.  

Comment by Stephen Crosbie on March 13, 2012 at 10:00pm
Great thoughts Rex et al!
Comment by Rex Pruitt on March 11, 2012 at 7:50am

Nice post Vincent! I really like the use of the 1/3 approach. Seems a lot of things are in thirds. Anyway, my personal career experience supports your "Data Scientist" metric breakdown and may help apply a time-line for either internal corporate development and/or a pattern for candidate evaluation.

  1. Early in my career I achieved #1 through education and training at the university and corporate sponsored level.
  2. For me, #3 came next through application of what I had acquired as skills and the personal realization of my talent (separation between Strengths & Weaknesses) driving emphasis on strengths development. This phase was filled with achieving success with skills and knowledge to the benefit of the organization being serviced.
  3. Then #2 played out, whereby, I was able to embrace knowledge and be in a position to share it to help the corporate vision, strategy, and my peers. It was a transition from departmental and process level success to corporate level success with recommendations that help drive the business at the Executive and BOD levels.

My business career is 30 years long and I would suggest that to achieve the stature and maturity of a well-rounded data scientist (if we must give it a title) that actually will contribute to “bottom-line” benefits; the individual needs 10 years equivalent in each segment.  Let’s face it; this is ultimately about delivering results. That would place the best candidates in the 40-50 year old range depending on how much “fast-track” occurs in acquiring skills and knowledge.

I'm sorry for using myself as an example and humbly submit that I'm only trying to apply the suggested metrics as a case study approach. My career is the only one that I'm qualified to do that with... (-;

Anyway, is the thought in your original post that there are 27 types of data scientists? Or, that a true data scientist (DS) has 27 qualities that qualify them for that title, thus establishing a measurable career path (I.e., DS-1, DS-2, DS-3, etc.,)?

Personally, I think titles are somewhat abstract and generally feed the ego and/or a labeled sense of accomplishment. Too often people are given a title they do not deserve or they deserve one they cannot get. I see this most often with Management titles...LOL!

Comment by Martin Wells on March 10, 2012 at 7:03am

Thanks, Vincent. I think the 3x3 segmentation is very useful for stratification and certainly points out why 'data scientist' can mean very different things to different people or industries. I have often thought along these lines - but use the 'classic' data scientist Venn diagram to describe metrics. 

I believe that essentially combining your first two metrics into one - termed Domain Expertise, but taken to include business/social accumen, vision, strategy, etc. - and splitting your single technical metric into two - one related to programming/hacking skills and the other math/statistics - more accurately represents the breakdown of data scientist function in industry today. 

Nonetheless, a useful reflection and discussion for employees AND employers as they contemplate the 'right fit'.

Comment by Vincent Granville on March 9, 2012 at 4:46pm

Yes Gregory. Also this special segment made up of professionals that rank high on all three dimensions represents far less than 1/27 of the data mining community. Also, this segment is made up of people like you who:

  1. have no interest in being hired and routinely turn down 98% of all employment opportunities (I can't speak for yourself, so feel free to reply if you disagree),
  2. and who actually may not be a good fit in the corporate world,
  3. and who make more money in their current situation than in any data scientist position in the corporate world or government.

I guess that's why employers complain that it is so hard to find talent. And why candidates complain it's hard to find a job.

On Data Science Central

© 2021   TechTarget, Inc.   Powered by

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