Progressing from analysis to interpretation, data scientists combine their problem-solving and storytelling skills.
By Russ Banham
Fifteen years ago, data scientists worked primarily in the fields of statistics and academia; today, they are in high demand by companies across industry sectors. The reason? Digital transformation. Companies are automating manual processes and using advanced analytics tools like machine learning (ML) to mine mountains of data in search of business insights. And that shift has transformed the job of a data scientist.
In addition to scraping, cleansing and prepping data for analysis, data scientists are developing storytelling skills to impart the significance of their findings in business terms. In turn, their work is informing better decisions by function leaders, from finance and accounting to sales and marketing.
Finding meaning isn’t enough
“The fundamental challenge with all data scientists is the meaning you derive from the data, but that’s not enough [for business leaders] to make a decision,” says Matthew Chu Cheong, founding data scientist at Stable Auto, a provider of data analytics improving electric vehicle (EV) charging station performance.
“You need to understand why and how the algorithm came to its conclusion,” continues Chu Cheong, whose Ph.D. research at the University of Texas at Austin focused on the development of “smart grid” decision-making algorithms. “And then you need to tell this story in words or visualizations to decision-makers to affirm that this is a compelling conclusion they can believe in.”
Paul Magnone, co-author of the book, “Decisions Over Decimals,” which focuses on the uniting of data intelligence with human judgment to drive better business decisions, shares this perspective.
“There are a select few who understand the power of data, know the questions to ask, connect it to their larger business strategy and use it to engage customers and achieve revenue objectives; anchoring this team is the data scientist,” says Magnone. “Data scientists are the people who can coax treasure out of messy, unstructured data.”
This “treasure” excites Ann Irvine, whose Ph.D. in computer science from Johns Hopkins University focused on ML algorithms. She serves as chief data scientist at Resilience, a provider of cyber risk solutions helping companies evaluate, manage and quantify their cyber risks.
“I suppose I’m a data scientist, a phrase that is not that old,” says Irvine. “I started as a software engineer doing analytics on the products I contributed towards. Then, I spent time with sales and [product] delivery teams making sure I understood how things worked, and then I did demos for potential customers.”
She adds, “As we continue to democratize data for more people to access it, the role of the data scientist appears to be shifting away from just engineering towards storytelling. I’m seeing an intersection of these skill sets.”
Chu Cheong agrees. “The world of business is becoming more data-centric, whether we like it or not,” he explains. “Data science is not just this technological niche; it’s a way of saying this is how our business is running, and I have analyzed the data streams that inform it.”
As data becomes democratized for employees to use in self-service analytics, more holistic skill sets amass, with one or two predominating as singular areas of expertise. As Magnone puts it, “In their quest to transform a firehose of information into intelligence, businesses are building teams who have the core skills to enable their organizations to make faster, more productive decisions.”
Resolving the supply-demand problem
The added competencies of data scientists are elevating the strategic importance of hiring them. Back in 2011, Harvard Business Review said that a data scientist was the most desirable job of the 21st century, due to their “training and curiosity to make discoveries in the world of big data.” In 2022, HBR reported that the job is even more in demand.
The problem is there aren’t enough data scientists to fulfill every company’s wishes. A recent survey by Anaconda states that 90% of companies are experiencing a talent shortage in the number of data scientists they seek. Meanwhile, 6 in 10 recruiters in a 2022 survey by Upwork say the toughest role to hire is data scientists.
Not just any data scientist. The study by Anaconda affirms growing interest in hiring data scientists who do more than crunch data, who also, “through reporting and presentation (make) data actionable.”
Magnone concurs with this evolution in the role, citing the progression from pure science to imaginative storytelling, or as he puts it, a “data artist [who] creates graphs, charts, infographics and other visualizations so people can quickly understand complex data. Great visualizations expand the dialogue by bringing more people into the conversation.”
He could be describing the five data scientists on Irvine’s team at Resilience. Not all of them are data scientists per se, a term that entered the lexicon only recently, she points out. They include former academics (Ph.D.s in math and chemical science, respectively), a former labor organizer and self-taught data scientist, a former high school math teacher, and a computer expert in Geographic Information Systems.
“They’ve been trained to use and interpret our advanced analytics to create our cyber risk models,” she says. “They’re all excellent communicators and writers. When they finish a project, they put it into a narrative that includes data visualizations. I send them into the business regularly to have conversations regarding their findings.”
Other chief data officers can learn from Irvine’s recruiting experience, fulfilling their demand for data scientists in individuals with more unorthodox backgrounds that have key competencies like curiosity, attitude and an abiding interest in diverse work, the skillset defining the modern data scientist.
Russ Banham is a Pulitzer-nominated business journalist and best-selling author.