How to Improve Data Literacy: 7 Proven Methods
As a company focused on helping organizations develop critical skills that their employees care about, we work on a lot of data literacy training programs. To succeed, these programs must foster organizational sophistication in and around data as well as progressively expand the individual data capabilities of employees over time.
On an individual level, data literacy takes root based on four abilities:
- being able to understand the data in front of you (what Jordan Morrow calls reading data)
- being able to manipulate and work with data
- being able to tie data to your job
- being able to drive change with data
Of course, the fourth step involves multiple skills: analyzing data in a way that produces credible and actionable findings, presenting your results in a compelling way, and being able to rally stakeholders to support and act on your conclusions.
At the organizational level, we look at a symmetrical frame to the individual one that mixes technology and culture. Here, too, there are ascending levels of sophistication:
- descriptive analytics, which enlighten organizations on the current state of affairs
- diagnostic analytics, which gets at the why of what’s happening in the organization (why were sales off in the Southeast but exceeded goal in the Northeast?)
- predictive analytics, which lays out what might happen in a data-supported way
- transformational analytics, which defines what should happen to take the organization to the next level
Manifesting data literacy as culture
It’s one thing to create individual change, it’s an entirely different challenge trying to figure out how to become data literate as an organization. Aligning these individuals in a way that informs and evolves a culture is a much taller order. Many data leaders have discovered that it’s not a foregone conclusion that as people become more literate, the data culture grows. To make cultural change take hold, we recommend that data leaders consider these seven strategies.
- Start small and scale up. Data literacy initiatives often are kicked off because a C-level executive has said, “The future of our organization depends on us being data literate. Let's go off and do it.” The bias is therefore to go off with a big bang, but these data culture initiatives often fail because they stretch so wide they can’t possibly go deep enough to drive meaningful change. Just as you don’t want to go live with a new technology across every employee’s laptop on the same morning, it pays to look at data literacy on a divisional or even a team level as you venture out. Go deep on a more limited basis, figure out how to do it well given your company’s unique dynamics, then scale up.
- Create a common language. Even a casual look into the acronyms that define a field demonstrate that common language is essential to driving change. Skills are important, but it’s when we are able to talk about why we use the skill, and how, that the right vocabulary starts to show up in people’s working language. A common language is also where each individual change can affect and reinforce someone else’s change, which creates positive momentum across the board.
- Frame questions linked to metrics. In a data-driven organization, asking why and how a team decided to prioritize one project over another, how they’re going to measure their efforts, and what success looks like is far more likely to produce concrete, quantitative measures of business impact. When you’re able to dig into the questions with descriptive and diagnostic tools, you can home in on whether that revenue growth you’re seeing is attributable to a rise in customer onboarding, to existing customer expansion, or to higher conversion rates. These are much more satisfying outcomes than anecdata responses that are often no more than a gut sense anchored by a few stories from the field.
- Make career development real. In our experience, data leaders who set the expectation that part of each employee’s job is to build skills consistently and constantly, then deliver the means for them to do it, dramatically outperform those that don’t. This means data literacy training that allows people to grasp concepts, then develop the habit of applying them through action. Too much training still consists of an employee sitting in a conference room or watching their screen as a paid actor explains something. This is why the average online training completion rate other than what is enforced by compliance standards hovers at a miserable three percent. You don’t learn how to ride a bike by watching a video of someone else ride one. You learn data by doing.
- Avoid dashboard proliferation. We’ve seen a dynamic where immature data companies find themselves awash in dashboards that measure without meaning anything. Data leaders should focus on setting a standard where the organization rallies around a core set of dashboards and visualizations that everyone can drill down on and don’t require a CDO to present. This also prevents the analytics team being bombarded with requests for dashboards that the rest of the company may then accept with blind faith. What you want in a data culture is people interpreting and analyzing the way data is presented, whether that’s from third parties or their own colleagues.
- Bring in senior leaders as SMEs. Rather than dropping senior execs into the deep end of visualization tools on day one, try involving them as stakeholders who can offer the business context behind the analytics. This creates a much more comfortable dynamic, with the added benefit that simply being a part of these programs starts to improve data literacy and teaches executives some of the core skills they will need.
- Teach in cohorts. We are a cohort-based learning company because we’ve found that when people have the opportunity to teach each other, it creates a much safer space to show vulnerability. No one can know everything in our super specialized world. Encouraging people to ask mentors if they can walk the group through how something works doesn’t make people feel exposed about a skill they may believe they already should possess.
Finally, don’t expect to improve your organization’s data literacy in a week, or even six months. Just as corporate cultures need constant tending, data cultures need reinforcement, practice, debriefs, and all the other tools that foster lasting change. Follow a realistic path based on common sense, and both individuals and the entire organization will flourish.