Insights on Data Science and Analytics by Srinidhi Rao


What are the leadership skills needed to drive effective change in organizations to achieve data and analytics excellence?

There are many skills that set apart effective analytics leaders from the rest. Among them, I would rate people skills at the top. A leader should be able to set the right vision for analytics in, align key stakeholders and drive adoption in the organization. In most cases adopting analytics involves investment, change management and short- term business risks. Can the leader rally people on the idea of analytics despite these challenges?

Second, in line would be business knowledge. No analytics leader can deliver value without adequate knowledge of the business. Business knowledge is required for identifying the right problem, socializing it and getting investment, translating that to analytics problem, translating analytics solution to business solution and finally for demonstrating value. Can the analytics leader talk numbers and impact in business language?

The third and the most underestimated skill is that of following through. The analytical transformation is a journey. It involves workflow changes and behavioural and cultural shift. Once stakeholders are aligned with analytics goals, the leader has to follow through at regular intervals, hold people accountable for their commitments, to drive sustained action and derive value.

The final skill that I would emphasize is empathy and team building. Good Data Science talent is hard to come by and even more difficult to retain. We need leaders who can create the right ecosystem to nurture talent and help them thrive, and hold them together as a team, while still catering to delivery pressures from the business.

Can artificial emotional intelligence be a substitute for real emotional intelligence?

There is no such thing called artificial emotional intelligence. Emotional intelligence, by definition, involves empathy, that machines aren’t built for. That said, machine intelligence can support the human effort to empathise and make it more effective. For example, if you are preparing for an important meeting, an AI program can summarize the Twitter activity of the person you are meeting and offer a synopsis of the sentiments expressed in those tweets. This will help you empathise with the person better and engage in a more effective meeting. Similarly, as we have seen in the case of Amazon or Flipkart, an AI program can help analyse your shopping patterns and determine how often to send you offers so that you have the highest chance of responding without feeling spammed. On the other extreme, you have chatbots like Mitsuku that can make you feel as if a real person is talking to you. They can even make emotionally sensitive and profound comments like “I am sorry to hear about your loss. I know how it feels.” Then again, this is an example of emotional intelligence being “programmed” rather than empathy being inculcated. The moment the Turing test fails is when the trust will be broken. To summarize, while AI can help to increase the experience of interactions, and even mimic emotional intelligence,  it can never replace human empathy.

What does sales analytics mean?

Sales analytics is essentially analytics that helps make your sales organization more effective.  At an operational level, the most prevalent use cases are - identifying the highest propensity-to-convert leads, forecasting demand in order to support staffing and inventory planning, pricing recommendations, insights on discounts, computing optimum cycle time to convert a lead,  etc. At a more strategic level, sales analytics can provide valuable insights such as the optimum mix of organizational spend on branding vs pricing, optimum assortment mix for each market, identifying whitespaces in product portfolio for which there is an unfulfilled customer need etc. Many organizations, big and small, have used analytics to increase the effectiveness of their sales strategy and operations, resulting in a direct impact on both top line and bottom line. 

What role does analytics play in measuring the engagement level of employees?

This is an interesting and important question. While analytics is widely used in functions where the dollar value impact is tangible, such as sales and revenue management, the average investment into people analytics stands at less than 5% across Fortune 500 organizations.  If we, as a community, believe that people are the most important assets of an organization, this trend should change.

Specifically, in the context of employee engagement, analytics opens up a lot of possibilities. For starters, employee engagement is a tricky attribute to measure. Analytics can actually help you align on what is the right metric(s) for your industry and organization to measure employee engagement. Second, it can help identify what drives employee engagement in your organization. For example, a major US bank figured out that employees who hang out together in groups are more engaged with the organization and show better performance. 

With insights like this, right behaviours can be nudged. Third, once you derive broad patterns to identify attributes of employees that are engaged more and stay longer, those insights can be fed back into the recruitment or on-boarding process. For example, a tech giant found that employees from a creative background were more likely to be disengaged in the company, as they could not relate to the technical folks around them. Based on this input, the tech company improved its on-boarding process to ease -in non-tech folks into the organization.

How can data scientists and analysts improve their communication skills?

Data Scientists are generally seen to be a breed more introverted than people in other professions. But that should not hinder them from being good communicators. Great thoughts and insights that are not well-communicated do not hold any business value. 

Storytelling: The most important aspect of communication for a Data Scientist is being able to tell a story from data. There are many good courses on Udemy and Coursera as well as videos on YouTube, on how to become a good storyteller. While you can get very good inputs online, practising and fitting the theory to one’s personality and style is the key.

The power of PowerPoint: Data scientists should learn to use Powerpoint effectively to support their stories. While stories can be impactful, it is the slides and videos in a presentation that make the stories “stick” and give credibility and recall value. A picture is worth a thousand words and a video, a million words. 

Pre-suasion: As much as it might seem like it is a far cry from their personality, Data Scientists should use pre-suasion as a technique to influence change. However powerful the data and insights are, decisions and actions are eventually taken by human beings. One important aspect of a good communicator is to get key people to support the analysis and recommendations. This can be done by involving key stakeholders during the analysis and taking their feedback, rather than doing the project in a silo and waking up to reality much later. It’s all about carrying people along!

Srinidhi Rao, the Senior Partner at TheMathCompany. He has over 15 years of consulting experience helping Fortune 500 companies improve organizational effectiveness through use of cutting-edge problem-solving tools and techniques such as advanced analytics, design thinking, systems thinking, game theory and behavioural sciences.

Post a comment