Riding the Third Wave of Analytics: Three Skills for Data Science Success

Two decades ago, trained data professionals spent an exorbitant amount in using statistical and analytical software such as SPSS and SAS. That was the very first wave of analytics transformation – monetization.  This was soon followed by commodification, as expert programmers started developing analytical algorithms on open source. Today, we are witnessing the third wave – democratization of analytics – that is finally getting the algorithms into everyone’s hands.  UI-based tools like Exploratory and Dataiku leverage open source technology, and are widely adopted by business users across marketing, finance, manufacturing, etc.

A significant consequence of the current transformation is the emergence of a new breed of data scientists from wide ranging educational and experience backgrounds. As organizations further democratize data science, three key skills will distinguish the talent of tomorrow – business thinking, software and statistical abilities, and soft skills.

Acquiring a business mindset

One of the biggest challenges for a data scientist is the ambiguous nature of requests they get from internal as well as external clients. In such a scenario, using a business mindset can prove to be highly valuable. Instead of fishing for signals in the data alone, thinking like a business stakeholder and working closely with business units can help data scientists decipher answers that are aligned with business objectives. Coursera, for instance, created small collaborative sub-teams of 2-4 data scientists to work closely with different functions and business units to better understand their business goals.

Let’s say the business ask is a vague statement like, “we want to increase sales.” This is an opportunity for the data scientist to add value by working with business stakeholders to find a solution for the abstract business problem. Depending on the business they are in, the data scientist can breakdown the problem (sales, in this case) into smaller subsets –  sales by product line or new and existing clients – so the trends become easily visible. The data scientist can also look at targeted marketing with proper segmentation, scoring and ranking of the buyers. There are multiple ways to slice a problem. Understanding the business mandate as well as industry trends not only helps in effective resolution but also in creating an innovative data culture that adds value to the organization.

Augmenting technical data skills

As per a recent ET report, there has been a 400% rise in demand for data science professionals in India across industries, but the supply has grown only by 19% in the past year.  The problem becomes more acute as the unstructured data available in most organizations continues to grow. To address this, many enterprises have started data science practices and training programs within their organizations or partnered with educational institutions/edtech firms to grow and mentor data science professionals.

A case in point is the Data University at Airbnb that has trained over 500 employees. Airbnb data scientists work directly with engineers, designers and product managers, creating a data-driven culture within the organization.

Developing data storytelling capabilities

The father of modern physics, Albert Eisntein famously said, “If you can’t explain it to a six-year-old, you don’t understand it yourself.” This is very relevant to the realm of data science as well. One of the more challenging tasks for data scientists is to present the insights buried under humongous amounts of data and complex calculations in the form of a simple story that even non-technical business folks can understand.  This is where soft skills such as people and communication skills come in handy. They are critical to not only convincing business stakeholders and getting them on board with the findings, but also liaising with other data-driven teams within the organization to enable vital business decisions.

LinkedIn, for instance, closely evaluates candidates’ communication, project management, and influence skills when hiring for data scientist positions, as it considers these skills to be as important for success as hard skills.

More than 39,000 analytics jobs are expected to be created in India alone by 2020. Intelligent machines based on Artificial Intelligence (AI) and machine learning (ML) can provide data science hard skills to a certain extent. But people skills and a business understanding is something machines cannot replicate.  While data can tell multiple stories, it needs creativity and talent to discover all the possibilities from the given data to improve products and services. Organizations that help their employees imbibe the right business, technical and soft skills through training and L&D will be able to democratize data, and create a data-driven organization for competitive success.

References:

https://towardsdatascience.com/what-is-the-most-effective-way-to-structure-a-data-science-team-498041b88dae

https://economictimes.indiatimes.com/jobs/indias-demand-for-data-scientists-grows-over-400-report/articleshow/64930355.cms

https://neilpatel.com/blog/how-airbnb-uses-data-science/

https://www.adma.com.au/resources/how-coca-cola-uses-data-to-supercharge-its-superbrand-status

https://www.datascience.com/blog/skills-for-linkedin-data-scientist-candidate