The Rise of the Data Smart Manager

The next generation of business leaders will be ones who master business, technology and data. Data Smart Professionals are much in demand, with McKinsey&Company predicting 1.5 million open positions for Data Smart Professionals in 2018 for US alone.  Let’s face it, not everyone can or should become a Data Scientist, it needs skills, aptitude, time and effort. Apart from recruiting new talent in the data space, organizations should work on re-skilling their existing workforce to take up this data challenge that we are facing.

Who are they?

Data Smart Professionals are those who know how to operationalize insights from data to drive businesses and decisions. Data Smart Professionals have a keen interest in driving data driven decision making at various levels in an organization. These Data Smart Professionals are also sometimes called data savvy Professionals. Our definition of Data Smart Manager is one who has a strong business acumen, cross-functional and people management skills, good understanding of data science principles, algorithms, stats, and math.

dt-2
Bernard Marr describes three levels of data maturity of an organization, namely using data to improve decision making, operations, and data monetization. A Data Smart Manager would be  responsible for maximizing ROI from data as companies progress in their data maturity journey. The need for Data Smart Professionals is emphasized by IBM as it is by McKinsey&Company.

Why are they important?

Asking the right questions and the ability to consume data insights are two most important functions of Data Smart Professionals.  Clark and Wiesenfeld describe the need for these “embedded generalists” who serve as “liaison” between data scientists and business partners to derive the real value of Data Science.The advantage of creating more Data Smart Professionals are many fold:                  1. They help improve the ROI from investments in data by operationalizing the insights, asking the right questions, and driving the data driven decision making culture throughout the organization. 2. Data Smart Professionals are relatively less costly to create in terms of time, as compared to training Data Scientists and analysts who will take few years of dedicated study in math, stats, programming etc. 3. Data Smart Professionals can be trained by re-skilling existing workforce who already possess domain knowledge and are accustomed to an organization’s culture. There of course, is possibility of training new graduates with data and business skills.

What skills do they possess?

While there are many skills that Data Smart Professionals possess, here are some important ones:

Asking the right questions: Data Smart Professionals ask the right questions. Scientific research is based on either inductive  or deductive logic. Deductive logic starts with a premise, collects data and then draws conclusions. The opposite of deductive logic is inductive logic, where the sequence is that first you have data, and then you make your conclusions from specific to general. Thanks to the data revolution, we now have huge datasets and there is a need to find patterns in them. In a sense this is a journey of finding questions based on the answers that are hidden in data.

Design of experiments: There is a difference between analytics and experimentation. Professionals involved in business decisions should increasingly rely on ‘test and learn’ than mere intuition. For analytics to succeed the right framing of problem is critical. The first step to problem framing is asking the right questions. The second is to design the right business experiment. For example A/B testing is one of the most effective ways of finding out which version of the website is working best. We need such type of test and learn experiments for data science problems as well.

Defining the right business metrics: If you throw data into an algorithm you will get some results, however, if the numbers are not relevant they have little value. Starting with the right business metrics is therefore crucial to any data science project in an organization. Bladt and Filbin in their HBR article suggest that good metrics are cheap,  consistent and quick to collect. Good metrics always put business goals first. For example, isn’t it plausible to think that 10000 Twitter followers are better than 2000 followers? Sure, more followers is a good sign. However, just having more followers does not directly translate into having more conversions. If conversion is your business metric then, more followers is not necessarily the best way to measure it.
Defining the right business metrics: If you throw data into an algorithm you will get some results, however, if the numbers are not relevant they have little value. Starting with the right business metrics is therefore crucial to any data science project in an organization. Bladt and Filbin in their HBR article suggest that good metrics are cheap,  consistent and quick to collect. Good metrics always put business goals first. For example, isn’t it plausible to think that 10000 Twitter followers are better than 2000 followers? Sure, more followers is a good sign. However, just having more followers does not directly translate into having more conversions. If conversion is your business metric then, more followers is not necessarily the best way to measure it.

Understanding of Algorithms, Stats and Math: A basic understanding of data science techniques and underlying Stats and Math is a must. A Data Smart Manager may not know how Eigenvalue is calculated, but should know Eigenvalue could help in determining the ideal number of clusters.

Data Evangelism and Leadership: It is also critical that Data Smart Professionals ensure the operationalizing of data driven decision making into the nitty gritty of an organization’s working. A Data Smart Manager needs to drive data driven decision making with an end goal in mind.

How can you become a Data Smart Manager?

There are two ways in which this quant crunch can be managed. One is to make existing domain experts more data savvy by training them on data skills and the interlock between data and business. The other way is to train people on both skills.


This article appeared first in the Silicon India Magazine


Dr. Marya Wani

Director of Programs- School of Data Science
Institute of Product Leadership

Become a Data Smart Manager today with an MBA in Data Science

Share on linkedin
LinkedIn
Share on twitter
Twitter
Share on google
Google+
Share on facebook
Facebook

Leave a Reply

Your email address will not be published. Required fields are marked *