Data Science for Managers and Directors

Over the last few years, there has been a meteoric rise in the number of online courses and training programs to prepare the next generation of data scientists. This is in response to the talent gap and the perceived demand for data scientists. Though finding a data scientist who can mine large data sets and derive actionable insights is still a challenging task, the increasing number of bootcamps, MooCs, post-graduate in data science and online training programs are somehow addressing the issue. In our opinion, there are many companies out there that are not necessarily short on data science talent, it is that the top management of the company does not properly use the talent they already have in the company. Business directors, managers, and VPs must adapt to work with data scientists if they want their company to make the best use of data; in a nutshell, they need to be more data fluent.

While the structure of many business organizations is highly complex, this simplified diagram clearly shows some of the important roles in a data driven organization.

The roles between a data science team and the senior leadership team meet the critical function of translating - interpreting important business plans into tangible initiatives, framing projects for data science professionals, interpreting and observing results from the data scientists, and communicating those results back to the senior leadership team in a convincing manner. In order to interpret clearly, a person should be well versed with both data and business. We think that this translation process can be further segregated into 4 different stages: clarify, iterate, interpret, and communicate.

Clarify: Abstract Strategy to Concrete Projects

Today many business organizations have started embracing data to become a "data driven organization".  However, the problem here is that “becoming data driven” is such a fuzzy concept that many get confused in identifying where to begin. While senior leadership envisions the big-picture goals for a business, data science professionals should come up with tangible steps to help in achieving these ambiguous strategic business goals. 

Though it is hard to ignore the temptation to look at available data and envision what can be achieved, a good data science manager should always begin from the end point, focusing on creating an experience that perfectly aligns with the business goals, and only then examines how to technically achieve it.

Iterate: Managing the Data Science Process

It is not always necessary that you be a data scientist to manage a project in data science. But it is useful to be familiar with the characteristics of data science tasks so that every project can be effectively planned and managed.  For example, there are some tasks that may sound easy, but are deceptively difficult. The best way to keep a project moving in the right direction is to be at ease with fast prototyping, iteration, and pivoting speedily towards a conclusion that meet the needs of a senior leadership team. 

Interpret: Consuming Your Data Responsibly

Interpreting data-driven results has never been an easy task. Reports that comprise charts, numbers, and correlations may lead to acceptance of results as scientific even if they contain serious errors. For managers, directors, and VPs, it is important to be able to question a data-driven result, since this leads to an understanding of what to trust, and eventually making the best actionable decision from the data. When interpreting data-driven results, a healthy skepticism is considered a manager's best friend.

Communicate: Sending Results to the higher level

Data science managers should always keep in mind that a senior leadership team is not as engrossed in the data as data scientists are. Thus it becomes essential that data science managers should effectively communicate quantitative information and help the senior leadership see the big picture as clearly and quickly as possible.