So You Want To Hire a Data Scientist

Tom Davenport, our chairman, along with co-author DJ Patil recently wrote, “Is Data Scientist Still the Sexiest Job of the 21st Century?”  This is an update to the original article from ten years ago.  It got us thinking about the organizations we work with and the wide range of data science resources and capabilities among them.

Something we hear often is, “we need to hire a data scientist” without much thought about what that means.  We reply, “what kind of data scientist?”  The field is so broad now that, while there are some good generalists, there is a lot of specialization among practitioners.  This is like saying, “I need to see a doctor.”  Well, what kind of doctor?  A general practitioner?  A cardiologist?  An oncologist?  A neurologist? You get the idea.

Data scientists may be especially experienced in marketing analytics.  Or supply chain.  Or predictive maintenance.  Or machine vision.  Or natural language processing.  And so on.  Layer on top of that a particular focus on a specific industry.  So, it is not as easy as just “hiring a data scientist” — it takes some thinking just to lay out the specifications for what you need.

Tom and DJ’s article provides a good sense for how to manage a diverse data science team given the current state of the profession.  Adding to that are these other considerations:

• Data scientists need peers to collaborate with and exchange ideas.  If you are going to hire only one data scientist you will probably be doing so again within a year.
• Data scientists need new challenges regularly and may have a short attention span.  If they are not able to complete an assignment and move on to another, you will lose them to another organization that can provide them with that new opportunity.
• Data scientists need the right tools (AI/ML packages and frameworks).  As Winston Churchill said, “give us the tools and we will finish the job”.
• Data scientists need to be involved with, and understand the importance of cultural and change management.  It is professionally disappointing to see one’s feasible, valuable, and even elegant solution fall by the wayside, unused.
• Data scientists should not be confused with data engineers, data architects, and other associated professionals.  While there are a small number of individuals who can play all of those roles, this is rare.  An organization that does not understand the distinction between these roles will likely not foster a rewarding environment for a data scientist.

Tom and DJ predict, “The amount of data, analytics, and AI in business and society seem unlikely to decline, so the job of data scientist will only continue to grow in its importance in the business landscape.”  It has taken a long time for business managers and human resources professionals to fully understand how to manage analytical talent.  In fact, leading HR organizations themselves are starting to apply analytics to talent development.  Now is the time to take the development of an organization’s data science capabilities as seriously as developing senior executive management.