On April 16, 2013, industry experts, advisors and visionaries convened on Twitter to discuss the bourgeoning role of the data scientist during our bi-weekly #IMChat. Named “Sexiest Job of 21st Century” by Harvard Business Review, data science is an amorphous and complex field, encompassing business knowledge, analytics expertise, technological skills and visualization capabilities. So what exactly does a data scientist do and do you really need one?

During this lively exchange, all those questions were answered and more. Moderated by Information Management Editor-in-Chief Julie Langenkamp-Muenkel (@JulieLangenkamp), the chat featured Krish Krishnan (@datagenius), among other data luminaries. The recap below lists the questions posed and a number of the most insightful tweets:

Do you really need a data scientist to handle big data?

@datagenius: #bigdata needs more analytical discovery and processing prior to reporting demands. Hence #DataScientist

@gwenshap: Data scientists often analyze small samples, data engineers turn their models into big data products

@InfoMgmtExec: I don't buy the idea that Big Data requires a new role e.g. Data Scientist to make it successful. The fundamentals haven't changed.

Is a good data scientist as difficult to find as claims say?

@gwenshap: The requirements are high: Actuaries who know machine learning, software development and distributed systems.

@JulieLangenkamp: McKinsey reports the US will need 140K to 190K people with deep analytics skills by 2018

@datagenius: Yes finding skills in the market today is not easy. lack of skills, lack of understanding the need & lack of role definition

What is a data scientist, and what does a data scientist really do?

@KirkDBorne: Focus on the Science in #DataScience: a Data Scientist infers hypotheses from data, design a test, evaluate, modify hypothesis...

@JulieLangenkamp: Part analytics expertise. Part tech skills. Part creativity and intuition to find relationships in data. Part observer.

@InfoMgmtExec: Combination of Intellectual Curiosity, Mathematical Analysis/Modeling, Data Mining, Data Manipulation & Problem Solving skills

@Eric_Kavanagh: A Data Scientist applies a wide range of methods and disciplines to extract insight from very large data sets.

How do you recruit a good data scientist?

@WhitneyAEden: You could take a look at the graduates of innovative BI/analytics programs bit.ly/11mKtu1

@datagenius: Good #DataScientist is a relative definition. In many cases we see the need to grow in-house talent. There are consultants too

@gwenshap: Current hiring practices seem to focus on PHDs, overlooking experienced data analysts who don't have the academic credentials.

@datagenius: There are universities like NorthWestern offering courses at Masters Degree level in #analytics to help the skill gap

How do you set up a data science team? And where do they fit in the org chart?

@datagenius: They will be a shared services team & will work closely with LOB Execs & SME's. Jointly funded by CIO/CxO

@InfoMgmtExec: Organizational Design for Analytics Success is critical. It is cross-functional and business lead. IT only a stakeholder.

@WhitneyAEden: Ideally, they should be equally involved in the biz and tech sides of the organization

@robinbloor: Where does the 500lb gorilla sit? Anywhere it wants to. Should be true of real data scientist teams

What kind of an environment best fosters success for data science/data scientists?

@WhitneyAEden: One that fosters exploration and collaboration. Data scientists need the bandwidth to explore

@gwenshap: data driven companies change behavior based on insights from data. Anything else is a waste of data science.

@datagenius: #DataScientist can succeed in any environment that needs non-traditional approaches to data processing

@InfoMgmtExec: Analysts require strong leadership, continuous engagement, independence, community & collaboration. Analytics no different.

@lrott: Don't silo your data team. They need access to every part of the organization and have the freedom to ask [questions]

Join us for the next #IMChat on April 30th at 1pm ET.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access