To Be a Data-Driven Enterprise, Become Data Literate First
These seven steps will help your enterprise build data literacy for every employee.
- By Claudia Imhoff, Romain Duboc
- September 14, 2020
In many organizations today, we hear CEOs, CDOs, and other executives crying out for their enterprises to be more data driven. Certainly, this is a worthwhile goal and critical for the very survival of organizations in these uncertain times. However, just throwing data and analytics at business users does not make a company "data driven." Mere access to these valuable assets does not mean that users (even intelligent users) will be able to use the data and analytics in a perceptive and shrewd fashion and come away with the insights and knowledge they need to improve the organization.
What is needed to correct this situation? The entire enterprise must become "data literate." Just as literacy is the ability to read, write, and comprehend the written word, data literacy is the ability to derive meaning from data and analytics. Data literacy is formally defined in Wikipedia as: "The ability to read, work with, analyze, and argue with data. Much like literacy as a general concept, data literacy focuses on the competencies involved in working with data. ... It involves understanding what data means, including the ability to read graphs and charts (visualizations) as well as draw conclusions from data."
Data-literate people know what data is appropriate to use for specific purposes, think critically about the information obtained by data analysis, recognize misrepresented or misleading data and analytics, and understand the underlying methods and results of analytics.
Additionally, the need for data literacy and data-driven decision making has generated a new breed of employees who serve the enterprise as data interpreters -- people who translate esoteric visualizations and data science results into meaningful business intelligence for less-data-literate personnel. These invaluable resources explain what is happening to the enterprise and the consequences of its actions through data storytelling using common business language and simplified graphs and charts.
A data-literate organization makes its decisions based on real facts and analytics rather than conjecture, bias, intuition, or gut instinct. A better understanding of predictions and analytics leads to better strategies and more successful business outcomes.
This article describes how to become data literate and what a modern organization needs to support data literacy. It provides practical steps your organization must take to become both data driven and data literate.
A Quick Guide to Becoming a Data-Literate Enterprise
Once you've grasped the essential elements of data literacy, it's time to move on to the hard issues. How does an organization become a regular stakeholder in the promotion of data literacy? How do you take this idea from the budding stage and ensure it continues to blossom far and wide, affecting the work of every employee?
Let's start by examining two key principles.
1. C-suite leaders must embrace new skills and attitudes to inspire others and change behavior.
Data-literate organizations begin their journey by ensuring their employees are educated about analytics and are part of the creation of those analytics.
To become data literate, you need people. The more you put people at the heart of this change, the more they will help the entire organization take the leap to data literacy. McKinsey says,
Crafting and implementing a big-data and advanced-analytics strategy demands much more than serving up data to an external provider to mine for hidden trends. Rather, it's about effecting widespread change in the way a company does its day-to-day business. The often-transformative nature of that change places serious demands on the top team.
We understand well how this organizational shift within the company implies a revolution of the management style itself, even of the deep company culture. You may wonder, whom do you choose to lead this change? To find the answer, we need to step back and ask, "What is the fundamental purpose of data for a business?" Because it enables better, faster decisions and business growth, it makes sense that data literacy must serve business's core objectives.
It's critical for the top management teams to advance their own knowledge so they'll be able to change their teams' mindset, reorganize roles, and create new ones to adjust to the diversity of data-related opportunities impacting all the organization.
Even though the C-suite must upgrade their skills to drive and inspire action, employees must also change their attitudes and approaches to using analytics. Employees need a thirst to discover, curiosity, and a willingness to learn. The role of a leader is to encourage the entire organization to craft the right initiatives and help employees embrace data literacy in their day-to-day activities.
2. Your data literacy program must use agile principles to engage and educate your team.
Your program must be agile and responsive, adapting to learner demands and expectations. This is a far cry from outmoded and rigid platforms that are built on clunky technical density. However, while designing a truly creative and playful data literacy program, don't forget about the essentials. All the tools and techniques of any traditional literacy program -- and its various delivery channels -- must be part of this initiative.
Using context, any idea can be made more relevant, immediate, and personal to the learner. Furthermore, to ensure that data literacy is not just more knowledge that is absorbed but rather a culture that lasts, it is important to use the fundamental pillars of the agile manifesto in the learning process from the group's point of view.
The more employees are involved in the learning process and exchange with each other, the more they will be able to identify the foundations of data literacy, take ownership of its principles, and thus be able to apply them to the management of their own data. Eventually, human-centered methodologies such as design thinking can be a great help here. (Read more on design thinking in this guide). This is fundamentally inclusive, widening the impact of your data literacy program, connecting with the learner's cognitive ecosystem, and creating a continuous cycle of knowledge assimilation.
Now that we have established the need and support for data literacy, let's turn our attention to how an organization can begin to improve the data literacy of every employee. Recognize that the following list is a starting point; it must be customized to fit your particular business and business environment.
1. Understand your business and its data/analytics needs. Of first order is to get a high-level understanding of what your organization does, its strategies, and the kinds of data and analytics needed throughout to ensure proper decision making. Much of this information may be obtained from annual reports, executive strategies, and even executive compensation plans.
2. Assess your organization's current data literacy level. A quick way to determine a company's data literacy is through a survey sent to representative personnel at all levels in the company. You can use the survey results to divide the employee population into two broad categories -- information producers and information consumers -- each with distinct data literacy needs. (Note: some individuals can be both.)
Information producers include:
- Data scientists: people who are educated in and knowledgeable about the advanced forms of data manipulation and analysis and can perform these analyses with R, Python, or other modern data science technologies
- IT professionals: those responsible for implementing and maintaining the technical environments for data and analytics
- Business analysts: people who work in various departments, performing simple analysis, may be "homegrown," and may lack proper training and access to sophisticated data analysis tools
Information consumers include:
- Company executives: people who understand and assess the data well enough to feel comfortable in making decisions; they may need training to ask the right questions and learn to cut through vendor hype
- Operational personnel: those who perform the day-to-day activities that manage or fulfill the tasks of running the organization
This baseline establishes the level and type of education needed for each employee group.
3. Begin educating each group according to their needs. We covered some outlines for the educational content needed in the Quick Guide section earlier in this article. As a first step, everyone should be exposed to a critical or logical thinking course that teaches problem solving and decision making. There are many such courses offered online. Some are even free. You can offer more advanced courses depending on the employee's interest, which may lead the person to become the much-needed data interpreter. Further education may be provided on topics such as decision making itself, why data is important to each employee, how to access data, and, ultimately, how to use it in their daily work decisions.
In addition, you need to train your employees in the technologies you are using. This training should be specific to the individual and the tools they use regularly, and it should be offered on a recurring schedule. Users need to understand the data itself -- where it came from, how it was processed, how it can be accessed, whom to contact for access, how to manipulate it for further discovery, where to find its metadata (definitions, lineage, usage tracking, and so on), and how it was used in calculations, algorithms, and statistics.
4. Empower employees to make decisions. Executives must believe in their employees enough to give them the freedom to make good decisions daily. Every employee is an analyst at some point during their workday, but without analytical and critical thinking skills or access to the right data and analyses, they can easily make the wrong decision. Therefore, it is critical that users have access to the right data and analytics at the right time to make the right decisions.
5. Support employees' decision-making ability with appropriate technologies and deployment methodologies. There are many technologies today that fit the needs of the various categories of business people. Their data and analytics needs -- as well as their level of data literacy -- must be taken into consideration when choosing the proper technologies to support their decision making. Unfortunately, no one technology will work for all levels of data literacy. Your organization needs multiple technologies for a complete and modern analytics environment. Remember, your ultimate goal is to build bridges between those technologies to offer a unified, understandable, and actionable environment for all employees.
6. Develop data storytelling and data visualization skills. On the last mile of this journey are the stories you weave and the insights your soon-to-be-literate employees can deduce instantly. Data storytelling must be engaging, empathetic, and, in the long run, profitable for your organization. In short, data storytelling creates an experiential framework.
Data visualization appeals to a learner's core cognitive capacities and a rich UI enables smarter and faster absorption. Developing the appropriate visualization for a particual audience is another skill to be learned by the data interpreter. This means that the data must be quickly assimilated; visualized using clear, concise, understandable graphs; and readily actionable on all levels (from the frontline personnel to the executive C-Suite) and in all business sectors.
Moreover, although data visualization serves these objectives of understanding and use, it also guarantees its use and adoption by users. In today's work environment, mobile applications are necessary to take data right into the everyday work of any employee, where anything they need/want to know is only a click or a swipe away.
7. Begin analyzing the organization's decisions for feedback into the analytics assets themselves. Tracking the decisions made should not to be used as a punishment for making a bad decision. Rather, the analysis of the decisions should be used to improve the analytics assets that were used in the decision-making process. A bad decision means that an analytics asset was wrong, misleading, or misinterpreted. In any case, an analysis of the overall decision-making process can be used to improve the assets and/or indicate the need for further education or training.