While speaking about the skills and attributes of a data scientist, mentioning emotional intelligence as a necessary trait is often frowned upon. To be a great data scientist, you do need excellent technical and analytical abilities. But emotional intelligence? Really? 

This couldn’t be further from the truth. Emotional intelligence is, in fact, a vital ingredient to becoming a highly successful data scientist. Technical and analytical skills are, of course, critical requirements, but they go only so far in helping answer ‘how’ to solve a given problem—grasping the ‘who’, ‘what’ and ‘why’ of a problem is key in delivering real value to a client. That requires skills beyond Python and Mathematics.

Here are some questions that one must ask while solving a data science problem:

• What is the real problem the client is facing beyond the stated problem?
• Who are the stakeholders that are impacted by the problem?
• What is the driver for each stakeholder to get the problem solved? 
• What are the potential options to pursue?
• Who will be positively or negatively affected by each solution, and why?
• How should we communicate to ensure that the solution lands well with the stakeholders?
• What are the risks of implementation failure, and how do we mitigate that risk?
• Are there any unexpected consequences that the stakeholders might face?

And so on.

None of these questions can be answered by simply wearing a technical hat. Behind every problem statement a data scientist faces, there are real people. Real problems. Real concerns. Real expectations. Delighting a client with your solutions, therefore, is not just about writing a high-powered ML algorithm. It’s also about dealing with the business and the emotions, solving the right problem, and helping relevant people get on board.

Demystifying emotional intelligence

First, the dictionary definition of emotional intelligence. It is ‘the capacity to be aware of, control, and express one’s emotions and handle interpersonal relationships judiciously and empathetically.’ Daniel Goleman, a science journalist who wrote the 1995 best seller ‘Emotional Intelligence’ outlines a model with five building blocks:

• Self-awareness
• Self-regulation
• Social skills
• Empathy
• Motivation

Some individuals may have better emotional intelligence innately than others, but the reality is that these skills can be acquired and developed with focused efforts. For example, some people may have a natural flair for displaying social skills, but even those who are awkward at this will get ahead with behavioural training and conscious efforts focusing on effective social interaction. 

In my view, while all these five building blocks are important, there are two crucial ones for a data scientist – empathy and social skills. We can’t possibly think of solving a client’s problem effectively without bringing in a blend of both these skills in ample amounts. 

Empathy, the ability to be in someone else’s shoes

Empathy is often confused with terms such as ‘sympathy’ and ‘compassion’. It is much broader than those characteristics. It is the ability to understand and feel a situation from another person’s frame of reference. 

Empathy should be one of the most important tools in a data scientist’s toolkit. When a data scientist starts seeing problems from their client’s perspective, they start internalising and experiencing the pain points, the burning need, and the perceived risks of failure seen by the client with a lot more intensity and clarity. This personal experience of putting oneself in the client’s shoes makes all the difference in tackling the right problem the right way. It paves the way for superior problem solving and, therefore, a more apt solution. When you lack empathy, you simply follow the written word of the problem statement, which can often be half-baked and shallow. Empathy compels us to take the right action that helps ease someone’s pain. It is key to building authentic, trusted relationships. 

What should a data scientist do to build an empathy mindset? Working on a few simple behavioural elements can go a long way. Here are a few tips:

• Resist the temptation to share your views and ‘grab the stage’ during a conversation; ask open questions more and listen to understand others’ perspectives.
• Pay attention to non-verbal signals such as body language – our body language conveys a lot more about our feelings than actual words do.
• Let go of the impulse to prove yourself right in a conversation and get into discussions with an attitude to learn.
• Internalise the thoughts and feelings of your client and absorb the root cause of their needs.
• Be vulnerable and express how you feel. If you are nervous or anxious, say it. You don’t have to appear super-human. It makes it easy for others to share their feelings with you.

Social skills, the bridge that connects

As a data scientist, you often deal with those ‘hard’ problems that the business has not been able to solve themselves. In other words, you don’t often have any low hanging fruits because the client has already explored those options and benefitted from them even before they approached you. To solve deeper problems that really matter, you must get your clients to invest their time and energy in talking to you. They will deep dive and join you in that exploration only if they think they can trust you. How you communicate and interact with them, verbally and non-verbally, will give them definitive cues on how far they can trust you and be open with you. In other words, your social skills can make this journey of exploration a lot easier.

Similarly, you may have formulated the best data science solution in the world, but unless you can get your client to buy into your solution, it is no good. An outstanding data scientist should also have the skills to visualise, tell the story, communicate effectively, and persuade the client to agree with the approach and solution.

The ability to connect with others goes way beyond good command over language or the ability to create visually attractive slides. Understanding the mental make-up of your client and framing your words accordingly is equally important. For example, you would need a very different approach delivering a message to a client who comes across as demanding or starts panicking at minor details, compared to someone who is relatively easygoing or self-assured. 

Here are a few ways to develop social skills:

• Try maintaining eye contact in conversations and create a goal to hold eye contact at least for a few seconds during the engagement.
• Give sincere compliments to someone who has done a job well; it shows that you are friendly and can help in starting engaging conversations. The operative word here is ‘sincere’.
• Practice active listening with full attention to the person you are speaking with, and paraphrase to ensure that you understood the point correctly.
• Keep yourself aware of current events, news, and trends, to find topics of common interest as ice breakers with your clients.
• Finally, simply engage with as many people as you can. Discard fears and approach each conversation confidently. Make mistakes. The only way to learn social skills is to practice social skills. 

Emotional intelligence is an essential skill one should acquire to become an outstanding data scientist. It goes beyond that – it is a life skill that helps one become a better professional, leader, and human being.

This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill the form here.