As data increasingly becomes the lifeblood of organizations great and small, executives and project managers are on the lookout for data scientists equipped not only with the right R, Python, and machine learning skills, but also a healthy dose of business acumen. 

Kathy Rudy, chief data and analytics officer at IT research and advisory firm ISG, suggests that, while skills in statistical theory and up-and-coming technologies such as artificial intelligence (A.I.) are important for data scientists and analysts to possess, you also need to understand which business-centric questions to ask, as well as how to present data and insights in a way that’s ultimately useful to the business. 

“Let’s say I work in D&A for an airline, I’m a wizard at A.I. tools and create great reports on how many flights people take, average searches before actually purchasing a ticket, average ticket spend, average flights per person per year—the basic data the system delivers,” she explained. “All great information, but what does the business need to know? Maybe it’s the average number of empty seats on a particular route or number of people on a waitlist per flight to determine if they need to add flights to a route? Just knowing how to work an Xbox does not mean you know how to play the games.”

That’s a point of view shared by Nisha Krishan, Everest Group senior analyst for digital transformation services, who points out the data sets required for growing e-commerce are different from those required for those of an insurance firm. “How do you understand the challenges the industry faces?” she said. “That can only come from the busines skills rather than the tech skills to generate insights and play with the data.”

As Krishan explained, the business expectations for data scientists are changing. The modern data scientist has to help product owners meet business objectives, and that means being ready to argue for (or against) particular business strategies. “The role of the data scientist is not just to build those algorithms, but to provide those business insights and direction, and communicate that to the business stakeholders, as well,” she said. 

That includes building strong communications skills that let you argue with data in a way that business users can understand. “Your technical language may not be easily understandable by other stakeholders, so how do you articulate your algorithmic findings?” Krishan said. “Data scientists should be able to argue with data in a way that eventually leads to a data-led culture within an organization.”  

Learning the Business

But how can data scientists learn solid business skills? Rudy suggests that data scientists and analysts try to embed within the business units they serve.

“I used to consult to IT service desks, and we’d create what we called ‘Dipping Sessions’ where agents would actually sit within the business on two-week rotations to understand how people did their jobs and the IT services they needed to do them,” she said. “In the case of data scientists, they need to know how the people they support use data and analysis to make better decisions.”

Do data scientists and analysts need business-based certifications? That’s an excellent question. Krishan points out that training and certifications are only the first step toward achieving business acumen. “People think certifications mean the talent is ready, and overreliance on certifications is a big problem,” she said. “Talent deployment means the person has the right skills and business acumen to do the job—that doesn’t come from certs alone.”

Data scientists should supplement any certification training with constant reinforcement, Krishan added, be it through mentorships or specific programs to help them understand the industry and the specific company. “It has to be attacked from all sides,” she said. 

As organizations define the data they need to successfully make fact-based decisions, the business side of data analytics will only grow. “Data is information, and when analyzed and used correctly, it can change how you build a business,” Rudy noted. “That’s not to say there won’t be bumps along the way, but as you learn how to read data and understand what data are required, the dependance on data for sound decision-making will only grow.”

Ultimately, data scientists need to understand the industry they serve, the metrics that drive the industry, the business unit they support, and the strategy of the organization. “What is the business trying to accomplish and what analytics—not just data—can support that mission?” Rudy asked. “If a data scientist can proactively bring insights to the business and demonstrate a strong understanding of that business, including the drivers of growth and competitive advantage, they’ll be golden.”