Creation of Value Through Data Science: A Note to Future Data Scientists

Without a proper understanding of business, the results of a data analysis remain useless. That’s the main challenge of future data scientists. Coordination with business, understanding of main trends and constant improvement is key success factors initiating the creation of value through data science business. 

Assume you know that you are very good at maths and statistics. You are confident in using all ETL (Extract, Transform and Load) tools and BI tools. You think that you have enough experience in data cleaning and visualization. So, in your opinion, you are the perfect data scientist. Bad news… maybe you are not.

You have the modern tools and know how to use them, but when it comes to figure out why you are using them, the story begins to change. When you go for a job interview with your amazing data analytical skills, you will be appreciated but there will be some questions to be answered:

• Do you understand that “data means business”?
• Do you know the business basics and understand the economic and industrial dynamics?
• Are you able to create “shareholder value” by solving the problems with the tools you have?
• Are you able to communicate the results of your work and give actionable insights to the management of the company?

Business understanding and experience is overlooked, simply assumed or just briefly mentioned in advice on becoming a data scientist, yet it is a big part of what makes an effective practitioner. Data science for business exists to solve real problems where data is integrated with the discoveries and/or solutions. You should realize that you have to apply your skills in data analysis to solve the business problems.

Consider yourself as a captain of who can use his ship with eyes closed shut. If you do not know the destination, none of your passengers would care about your skills in steering that ship. If they find themselves “in the middle of nowhere”, you are in trouble. Besides, you should also know how to arrive at that destination, the dangers waiting for you along the way, the optimal path to minimize the cost, the smoothest route to comfort your passengers. The passengers will also need to know where they are at a specific moment and other related information about the cruise. You should find the best way to communicate the status and the problems faced, if any. You should also keep in mind that if the passengers are not happy, they can find another ship to travel next time, your company will lose money and you’ll have a bad reputation as a captain.

Therefore, as a data scientist, your role is very similar to this captain. Now you know that using the tools you have or how advanced the tools are not enough to achieve your responsibilities, here are the things to complete to call yourself a business oriented “data scientist”:

• Get familiar and be keen about the strategic plan of the company. Before starting to data collection and analysis process, you can use this knowledge as a guidance to solve the problems put in front of you.
• Always understand the characteristics of your target beneficiaries and tailor your process accordingly. Each of the beneficiaries who use the results of your work has different needs depending on their coverage of the overall company strategy. You should shift the shape of your mind in order to perceive problems from their point of view.
• Understand the basics of the problem by communicating with the business and let the problem drive the process of choosing the related data rather than trying to fit the “available data” to the problems at hand.
• Take a leading role in promoting the data quality in the organization.  Do this from the business point of view, by showing how the “quality data” can contribute to objectives of the company to the people on the other side of the table.
• Communicate the results of your analysis effectively to the related business decision makers. Support your result with informative dashboards, graphs and charts in a way carefully precise and relevant. Put narratives to better explain the story behind the facts considering the mindset of business decision makers.
• Know how decision-making process works. Try to give an idea about the processes to be changed for improvement. Share the data driven problems you faced throughout your analysis and give recommendations to resolve these problems. As a data scientist, you can have a strong impact on the company’s desire to learn and improve.

When someone asks you what your job is, you should confidently answer that you add value to the business based on the data.

Having all these capabilities at once is not easy for everyone, it is a challenge and especially when working in large firms with hierarchical or specialized structures. While the general business knowledge can be quickly learned on the job and in school, the deeper intuition comes with experience.

The similar challenge is also valid for the “corporate controllers”. Their knowledge of the accounting and business will not be sufficient to solve the analytical problems they face in the normal course of their business.

Financial ccontrollers have the challenge to learn how to use business analytics tools but it’s also a fresh chance to differentiate themselves from their colleagues. The quality of planning, forecasting and performance management considerably increase with the use of statistical methods and models. They can show previously unrevealed causal effects or generate more precise forecasts.

However, to use the methods of modern advanced data analytics, financial controllers may have to go beyond the functional limits of the data scientists.

It is essential for controllers to act as designers and coordinators of the business analytics process. To do so, they must be able to act as a “link” between management, data scientists and IT experts.

No need for any confusion. The only truth is that the data analytics is all over the business, so there is no escape. Do not isolate the knowledge of data analytics from the business and do your business using the data analytical technologies.