A Data Scientist is not just Data or Science ...
Two incidents separated in time and space is the motive behind this article.
A nephew of mine was interested in a lateral shift into the Data Science arena after spending over a decade in the IT sector. We had a discussion on whether domain was more important or the math. This was a few months back. I was emphatic that domain was far more important. Somehow I was not able to convince him,
The second incident was today. I was chair of a panel discussion in the Cypher 2018 conference in Bangalore. The topic the panel deliberated on was "Capacity and Capability building for Talent in Data Science". One of the benefits of being a chair of panel discussions is that you get free coaching on thoughts troubling you - without the panelists realising it. My nephews discussion was still not resolved in my mind. The panelists helped create some clarity. Virtually every panel member highlighted one very important aspect of being a Data Scientist. A Data Scientist is not merely an engineer who understands how to massage data using a few tools and algorithms. Perhaps the most important quality of a good Data Scientist was that the person understood business. In fact one of the panelists went so far as to say that a good Data Scientist would make a good Business Manager and vice versa. It suddenly dawned on me that the "domain" that I was trying to get my nephew to specialise in was "business" in all its flavours. For a techie this is close to blasphemy. However, Increasingly this realisation is hitting the profession. It is important to understand data and related technologies. However it is imperative to understand business.
Data Scientist as a nomenclature seems to somehow do injustice to the role.
- A Data Scientist does far more than manage Data.
- Organisationally the designation Data Scientist appears as if the role falls under the CIO's control. Ideally it should report into the COO if not the CEO. At the least they should be collocated with the business.
- The word "Scientist" in "Data Scientist" seems to imply that the person needs to have a technical background. Actually many economists and social scientists could play the role far better.
- The designation also seems to indicate that the role is technology and tooling heavy. In reality it is business heavy.
- A technical person tends to start at the Data end of the story. And the old truism that if you massage data long enough - it will tell you what you want it to say, plays out. In reality a good Data Scientist ought to constantly think about the business insights that would be most useful and then seek out the data that can give him/her that.
Is it time to change the name of the role. Should we title the position "Business Insights Engineer"? The optics of this designation more accurately reflect the role. The person is supposed to glean business insights and yet has to be tech savvy.
- The word Engineer indicates more the process than the competence.
- Your potential talent pool is suddenly larger - you move away from merely looking for technically qualified (read engineering graduates) professionals.
- This designation could also help position the role within the right context in the organisation.
- It could help professionals understand that there are multiple career paths in the area of Data Science. One on the technical axis (Data Curator, Data Engineer, Visualisation Engineer et al); and the other on the business axis (from Business Analyst, Business Process Manager to Chief Operating Officer).
If we are to truly benefit from the role of what we call a Data Scientist today, my opinion is that we should strengthen our focus on the Business side of the story. And ensure that those who occupy such positions have enormous Business Acumen in addition to being Tech Savvy, and comfort with Data.
Is it really necessary to change the designation? It helps far more than most would imagine. Role holders understand their responsibility more clearly and potential role holders realise the kind of competences they need to pick up. Critics would ask "what's in a name?". I merely stake claim to calling a rose, a rose!
As an Experienced Certified Coach, I help Corporate Leaders, Entrepreneurs & Business Owners to Scale-up rapidly
5 年Well written.... understanding the data and applying that knowledge to discover trends in order to take informed decisions is important
Technology Team Lead at Amdocs
5 年I read this piece and imagined you standing right in front of me. Glad to receive this insight.?
Digital Innovation in Railways and Mobility, Propulsion Expert and Power Electronics Expert; Monitoring and Diagnosis in Railways and Infrastructure
6 年Fully agree with your views, remember a lot similar presentation you made some 14 years ago at Bangalore SPIN in the context of Software Quality. One suggestion to the above article and proceedings, "Business" needs one more "off shoot" - i.e. technology and engineering under it, the question could be why ?, I am in the field of Analytics/AI/ML for engineered entities be it Traction Equipment, Signalling, Energy, Thermal Power, Distribution...... and it is? nightmare to get a "Data Scientist" to align with all this, it is high time people with engineering experience and specially "computational engineering" start doing Data-Science, that would bring a lot of meaning to the merger (Data) Science + (Real) Engineering.??
SAP LE WM Consultant ★ Supply Chain ★ Business Process Design & Improvement ★ Business Analytics
6 年Appreciate your views Bhaskaran. Isn’t it natural that without the right context, Data is just a series of 0s and 1s. Unless Data is paired with right business context, deriving right wisdom ( a.k.a business value) from it won’t happen, isn’t it?
Operations | Ideation | Strategy | Data Capture and Quality
6 年Potentially there is opportunity to develop a new role as well... An intermediary between the quantitative data scientist and the decision makers. Data/insight communication often needs a creative view that not all DS have developed.