Navigating Data Literacy In The World Of Augmented Analytics

Artificial intelligence (AI) capabilities like machine learning (ML) and natural language processing (NLP) continue to improve, and augmented analytics products can reliably automate many tasks related to seeing and understanding data. With powerful tools that can surface insights from data, executives are often left wondering: Does this technology actually reduce the need for data literacy training efforts in their organizations? No, rather the contrary.

Data literacy—the ability to read, write, and communicate data in context—is more important than ever. It is critical in helping organizations develop a data-driven way of working and empowering employees to augment AI skills with their own creativity and critical thinking.

There are additional factors to consider in the role of data literacy for an organization’s growth and success. Hiring, training, and retaining data scientists and analysts is difficult–plus, their skills are often nuanced and expensive. According to 365 Data Science, most data scientists probably won’t spend more than 1.7 years at their current workplace. Data scientists and analysts, who are highly trained, often receive requests for tasks such as building a clean data source for sales or churning out basic reports. With their specialized abilities, their time and skill set would be better served working on modeling and developing workflows for higher-value, complex business questions.

When executives invest in AI and augmented analytics technology, the business user—a more casual user of data compared to a dedicated analyst—can access the answers to their questions and the information they need to do their jobs well without worrying about the mechanics of doing so.

Exploring how AI-enabled solutions can support user tasks and find the right user experience has enormous potential to set the tool and the user up for success. For instance, an AI tool can automate some of the more tedious tasks around data preparation and then provide the results to the human, who can further analyze and visualize the content based on their analytical needs.

Advancements in Augmented Analytics Help People Answer Questions Faster

Augmented analytics solutions can make it easier for business users to understand data, which helps companies maximize the value of these costly technologies. For example, augmented analytics can understand customer interest and offer predictions about consumer preferences, product development, and marketing channels. They can also provide additional context about trends, values, and variances in one’s data. Sophisticated algorithms can suggest additional visualizations that can be added to a dashboard, along with text explanations and context generated in natural language.

Here are some examples of solutions that can help elevate your workforce.

1. Data Stories. Tableau Cloud now includes Data Stories, a dynamic dashboard widget feature that employs AI algorithms to analyze data and write a simple story about it in either a narrative or bulleted form. The stories weave together narratives about data beyond mere charts and dashboards at a register accessible to business users for answering many of their questions. This reduces the level of data literacy a business user needs to understand the information most important to them. Data Stories surfaces the simple questions a user asks when they first look at a bar chart or a line chart: Was this number that looks like an outlier truly an outlier? How has that number changed over time? What’s the average? The data still needs to be interpreted—it’s not the entire story—but it’s a big step toward unlocking the insights in data.

2. Show Me. Augmented analytics features also allow for smarter encoding defaults. For example, Show Me recommends chart types and appropriate mark encodings based on data attributes of interest. Users can then focus on the high-level takeaway they want to communicate and share these charts with their audience as part of their visual analytical workflow.

3. Natural language understanding. With sophisticated research, large training sets for language models, and improved computing capabilities, natural language understanding has also significantly improved over the years.

People can ask analytical questions without having to understand the mechanics of constructing SQL queries. With better intent of understanding, natural language interfaces can answer questions with interactive charts that users can repair, refine, and interact with as they make sense of the data.

4. Machine learning. Augmented analytics related to ML has also made strides. These models can learn sophisticated and complex analytical tasks such as data transformation operations that are personalized to a specific type of user or a group of users. Furthermore, many augmented analytics experiences now have user interfaces that feel intuitive, reducing the complexity of training and applying a model in a user’s analytical workflow.

Although AI has incredible capabilities, it will never completely replace humans. Gleaning high-level takeaways from lower-level statistical properties can be complex and rather nuanced. People have a higher level of creative cognition; we are inquisitive; we can distill these high-level takeaways from data.

Recommendations for Fostering Data Literacy

In order for organizations to unlock higher-level insights from their data, employees—business users and analysts alike—must be educated about how they should analyze their data and have best practices for visualizing and presenting data. Here’s how organizations can develop best practices in promoting data literacy and augmenting AI with analytics tools.

1. Invest in training.

Having both the right tools and the right education/training is critical for any organization. In a Forrester Consulting study on data literacy, only 40% of employees said their organization had provided the data skills training they are expected to have.1 Individuals and organizations should expose people to better training in terms of the best practices of seeing and understanding their data. Workplaces should offer courses around data visualization and data literacy so employees can understand patterns and learn the best ways to create and represent charts.

To train your employees, you can enlist great third-party programs by companies like Qlik, Data Literacy, Coursera’s Data & Analytics Academy, EdX, Datacamp, Khan Academy, General Assembly, LinkedIn Learning, and more. Tableau offers self-driven learning, live, virtual training classes, and a free course on data literacy. Similar projects that incorporate training, some of which are free, include Data to the People, Storytelling with Data, The Data Lodge, The Data Literacy Project, and others.

Executives should also consider: How can your employees be trained, not just in the language of charts but also as a broader paradigm?

One downside of building tools that have a lot of augmented capabilities—which include AI and machine learning—is they can look deceptively simple, and they can get users very quickly ramped up. But undertrained users could generate a chart or takeaway insights from a chart that could be misleading or misguided in some way.

It’s important to educate people on the language of visual representation and the science behind it so that they, at the least, are data informed, if not data literate. For example, how do people identify what an outlier is? How should they design dashboards that are trustworthy? They should also be able to understand the distinction between correlation and causation. This will ensure the data is accurate and can be used for analysis.

2. Make data-driven decisions.

Moving from data orality—where people talk about making data-driven decisions—to data literacy—where people have the ability to explore, understand, and communicate with data—requires democratizing access to data visualizations. This entails a focus on individual learning and applicability, but it should be more of an organizational change. The true democratization of data literacy takes into account the entire ecosystem of data. It recognizes the proliferation of charts in users’ daily lives and works to make them intelligible broadly.

People ought to be making decisions based on data and not just on subjective opinions; this goes back to the importance of training that educates users on the distinction between correlation and causation. How should data-driven decisions be made? What is the medium of presenting data and the key takeaways so the discussion can stay objective to make effective decisions? For example, tech companies should use user telemetry data to determine what features to build, usage characteristics, and identify any friction in the user experience.

3. Develop and maintain adequate infrastructure.

To support the first two recommendations, executives must ensure their organization has built an adequate, scalable infrastructure to house and govern its data. They should also help their organizations identify and gain access to AI technology that addresses their customer problems and needs.

Furthermore, decision-makers must be thoughtful and deliberate about data privacy and trust. It cannot be an afterthought; it must be taken into account seriously right from the beginning. The responsibility of data privacy and trust should be distilled all the way down to the individual user, which comprehensive data governance and management policies can cover.

Continue Focusing on Data Literacy Efforts

Investing in AI and augmented analytics tools like Data Stories is an excellent step toward empowering business users to unearth answers from their data, but these tools will complement data literacy efforts rather than replace them. Furthermore, the right forms of investment in both AI technology and training can effectively support humans to do what they are best at: ideating and creating solutions while solving customer needs, all centered around data.

Continuing to focus on data literacy throughout your organization will ensure that more of your employees—the casual business user and the sophisticated data analyst—are asking the right questions about your data that will lead to further insights.


An analytics partner like Tableau offers breadth and depth in capabilities as well as role-based training—making it a flexible partner on the journey to discovering what works best for your company. Learn more about Tableau Cloud.


Set up your business users for success. Learn more about Data Stories here.