Minerva Studio - Fotolia
Research shows that data-driven organizations are more successful, but employees often lack needed data literacy skills.
According to a 2020 survey by Sapio Research, 80% of decision-makers believed opening up access to data has a positive impact on their organizations and 74% said employees have access to the data they need. But 53% of respondents reported employee resistance to using the data.
Meanwhile, research has consistently shown that data-driven companies are more successful. A 2019 survey by McKinsey & Company, a management consulting firm, showed that companies where employees consistently use data in decision-making are one-and-a-half times more likely to report revenue growth of more than 10% in the past three years.
The difference comes down to data literacy.
"It is crucial in today's world where data is omnipresent," said Shreeni Srinivasan, director of enterprise analytics and applications delivery at Sungard Availability Services. "Data literacy can empower employees to make fact-based analytics decisions that are more grounded in reality than the ones made on instincts or gut feeling."
According to the same McKinsey & Company survey, the share of executives at high-performing organizations who understand data concepts is 44% higher, the share of managers who understand data is 39% higher and the number of frontline employees who understand data is 12% higher than other survey companies.
However, there are significant obstacles to achieving data literacy. According to Gartner, 50% of organizations lack sufficient data literacy skills to achieve business value, and 35% of chief data officers said poor data literacy is one of the top roadblocks to the effective use of data and analytics, just behind cultural challenges and lack of resources and funding.
What is data literacy?
Data literacy is the ability to write and comprehend data similar to how we view literacy with reading. This can include an understanding of where data comes from, communicating information derived from data to others and knowing where to use different analytical tools and methods.
"When enterprises have more data literate employees, they understand that data is no longer just the domain of the data team," said Andrew Stewardson, data manager at Farm Credit Services of America, a credit provider to farm and ranch operators based in Omaha, Neb. "Having a higher level of data literacy means that we can better serve our customers."
Stewardson's organization took an unusual approach to data literacy training by creating an internal persona, Walt, to answer data-related questions from employees.
"The key to encouraging data literacy training was making Walt relatable to various people within the organization," said Michael Meyer, data engineer at Farm Credit Services of America. "We also created a blog where users could ask questions about all things related to data."
That took the pressure off the data teams to drive change, he added.
"Just putting data into the hands of individuals in an organization doesn't automatically increase data literacy and make an organization data-driven," Stewardson said.
In fact, rolling out data projects without paying attention to data literacy can be a costly mistake.
For example, Penny Wand, director of technology at West Monroe Partners, a management and technology consulting firm based in Chicago, was working on a project for a manufacturing firm to roll out pricing strategy analytics.
"People just rejected it," she said. "They didn't understand the results."
The project was a failure, and not only did the company lose the time and money it spent on creating the analytics, it also lost millions of dollars in lost opportunities, Wand said.
"It cost them money because they couldn't optimize their pricing strategy," she said. "They lost money by not being able to put into action what they learned with the data."
A lot of people have been out of school for a long time, and their math and data analysis skills aren't on a level playing field, Wand said. And this doesn't just hurt basic analytics projects.
"Forget about AI -- without some level of data literacy, you're not going to get there," she said.
Who's responsible for data literacy education?
Unfortunately, there aren't currently a lot of best practices and guidelines to follow when it comes to teaching employees in data literacy skills, Wand said.
"We are in the infancy of formal data literacy programs," she said.
One approach is to make education specific to the role the employee has at the organization, she said. That's what Coursera is doing with its Data Science Academy. Another approach is to get the education to people as they need it -- as they're using their data tools -- instead of a formal training program.
Then there's the challenge of measuring success.
Many companies are addressing the data literacy skills challenge by moving data scientists out of data science departments and into business units where they work closely with end users, said Bryan Coker, principal consultant of data and analytics at AIM Consulting Group, a Seattle-based management consultancy.
"I think that's a trend almost everywhere," he said.
Another approach is to offer targeted, hands-on workshops to employees with a focus on the specific analytics tools used in the company and the specific business challenges around data the company and its employees face.
These workshops can be run either by the tool vendors or by independent consultants, said Justin Richie, data science director at Nerdery, a technology consultancy.
"I'm a very firm believer that people learn best by doing," Richie said. "So having the ability to have hands on a keyboard or laptop and making something is all about creating that contextual knowledge of doing it yourself. It's better than sitting in a college-sized auditorium and hearing someone speak for eight hours."
Or, these days, the Zoom equivalent, he added.
How can enterprises encourage data literacy skills?
Not everyone wants to rush out and learn math.
"That would be a good workday for me, just taking math classes," said Jeff Herman, data science instructor at the Flatiron School in New York.
Other employees may need some support.
At Herman's previous job for a railroad company, data scientists led training sessions for other employees about data and how to use it to better do their jobs.
"We'd talk about basic statistics," he said. "We'd talk about different databases: Here's a database of locomotive data, here's data about where the trains are going, here are the financials, and here's how to access the data."
Companies looking to do something similar should look for communications skills when hiring data scientists.
Jeff HermanData science instructor, Flatiron School
"People who can communicate with nontechnical stakeholders and be comfortable leading training," Herman said.
The Flatiron School also has a free data science prep course in addition to its regular data science curriculum, he said. Khan Academy also offers a lot of free courses covering everything from basic statistics to data analysis.
But it's not just about making the training available, Herman said.
"Companies need to talk about what the benefit is to the employee," he said. "It's not just a benefit to the company but for you as well. It's going to open more doors for you, make you more marketable."
Additionally, data literacy starts at the top, with the executive team.
"They have to be comfortable with the idea of making decisions based on data," Herman said.
Benefits of improving data literacy
When the only people looking at data are data scientists, important insights might be missed. For example, at the railroad company where Herman used to work, one key indicator was locomotive idle time.
"We thought it was wasteful of fuel," Herman said.
But when other people at the company -- outside the data analytics teams -- started using Power BI, they were able to see that same data from a different perspective.
"People closer to the operations of locomotives knew that there were specific reasons it had to be idling sometimes," he said. "They were able to make a dashboard for when a locomotive was idling when it shouldn't be and zero in on the instances where we could actually save fuel."
It's no surprise that in a recent Forrester Research survey, 90% of surveyed data and analytics decision-makers saw increasing the use of data insights in business decision-making as a priority.
But with only 48% of decisions being based on quantitative information and analysis, there's quite a bit of room for improvement.