The goals of data storytelling are to inform audiences and influence decision-making. Catherine Cote (2021) notes there are three elements of data storytelling:

1. The data. 

2. The visualization. This can be charts, graphs, pictures, videos, diagrams and other representations of the data.

3. The narrative. This communicates insights from the data. Insights include, but are not limited to, relationships, context, or change. 

Not all data sets are well suited for storytelling. Dykes (2019) developed the following figure to assess if a data project should employ storytelling as a method for dissemination. 

To use data storytelling as a method, the following five elements must be employed. 

1. Storytelling should be used with data that are insightful, meaning it goes beyond providing interesting information and instead conveys an observation or understanding. 

2. It should be explanatory, where the data displayed is predetermined and has limited interactive components (more on the difference between exploratory and explanatory in the next section). 

3. The data should be concrete, meaning a specific conclusion is being made and there is little opportunity for viewers to interpret the data in multiple ways. 

4. It should also be finite, where the data represents a fixed moment in time so that users can glean insights and understandings at a much deeper level.

5. Lastly, it should be curated meaning the elements highlighted from the data were chosen for a specific purpose for the intended audience. 

While the five elements are helpful in determining if storytelling should be employed in a project, it is important to have an understanding of the common types of visualizations and why explanatory is the most appropriate method for storytelling. In explanatory visualizations, the researcher builds the narrative and communicates it to the audience. Explanatory visualizations tend to focus on specific topics and are simpler in design. 

The USA Facts website provides a series of explanatory visualizations meant to address common concerns among the American public. 

In the visualization, Is Congress representative of the American people?, the researcher directs the audience towards the central story and specific insight: the representation of women in congress. The story is in the title. There is one simple filter in the visualization: Congress representation vs US population. Although originally in one visualization, two are presented here to illustrate the difference. To contrast, an exploratory visualization would likely include a filter with multiple demographic categories: race/ethnicity, income, gender, and so forth. And the title would be generic, something along the lines of “Representation in Congress by Demographic.” 

This visualization also uses effective visual cues. The different colors are intuitive and effective in displaying a contrast. The shape of the visualization is similar to the congressional seating pattern, immediately placing the actual congressional house and senate chambers in the minds of the audience. The original data source is provided underneath the visualization, establishing trust and credibility. However, it is in a gray font so as to not distract from the central story.

In exploratory visualizations, the researcher presents the data to the audience. The audience discovers patterns, trends, insights, and a general narrative on their own. Exploratory visualizations are generally more expansive. Traditional dashboards are exploratory.

This exploratory visualization from the National Survey of Student Engagement (NSSE) displays sense of belonging data for college students. 

Note the title is descriptive, not narrative. “NSSE 2020 Sense of Belonging” does not convey a story, but a topic. This visualization has much more detail than an explanatory visualization. It offers a comprehensive view of the data, including detailed and granular information. Users can access underlying data and additional details as needed. This requires more work by the audience, but may incentivize deeper investigation.

The audience is required to craft a story by using the filters. For example, an individual may be interested in telling a story about belonging for first-generation students. The audience draws their own conclusions. Some audience members may even export the data for their own analyses.

In exhibitory visualizations, the audience must rely entirely on their own senses and interpretation skills. Kirk (2019) likens these types of visualizations to art in a museum. The audience is on their own in terms of understanding the content and context of the visualization.

This exhibitory visualization from Data Pointed relies on the capacity of the audience to interpret the data. This visualization is presented in the context of an article (see the Read More button). If this were an explanatory visualization, the researcher may create a story title that directs the audience to a specific narrative. If it was exploratory, the visualization would include filters and other interactive features so the audience could create their own narrative. 

For most data projects, explanatory or exploratory visualizations are typically used. The following chart can help in determining which type is best for the intended outcomes. 

Much like works of fiction and nonfiction, data stories often contain similar constructive elements. In literature, stories regularly incorporate four components:

1. A setting or plot.

2. Characters. Usually there’s a main character or protagonist. 

3. A problem or conflict that needs to be addressed or overcome. 

4. A resolution or transformation.

Setting

In Everyday Business Storytelling, Janine Kurnoff and Lee Lazarus (2021) note that establishing a setting “builds critical focus.” Providing background information about the data source and other relevant factors helps the audience understand where the data comes from and why it matters. When the setting is clear, the audience can relate the data to their own experiences or knowledge.

Decision-makers rely on contextual information to make informed choices. Establishing the setting enhances decision-making by focusing on data that is relevant. Setting can be established through the following: 

• Clear and detailed titles and subtitles. In explanatory titles, make sure the titles are engaging and communicate a narrative, not only describe the data.

• Annotations and labels. 

• Narrative text. Text plays an important role in data storytelling. Don’t be afraid to use it to enhance the story. 

• Reference points. Add data sources and technical notes. However, don’t make them the central focus of the story. Place them at the bottom and in a gray font.

Characters

Characters humanize your story. This is particularly important in social science and education fields. As Kurnoff and Lazarus (2021) note, “the more your audience learns about the situation - and the effects it has on your character - the deeper their interest grows.” Characters can motivate decision-makers to take action by showing the effects of decisions or interventions on real people.

Including multiple characters can provide different perspectives and add depth to data stories. This can be achieved by showing how various groups or individuals are affected by the data. In the exploratory belonging visualization presented earlier, filters allow the audience to explore intersectionalities among students. In higher education, characters are usually students. But they can also include faculty, staff, community members, and other stakeholder groups.

Here are a few practical ways to include characters in your data stories: 

• Use interactive filters. The audience can select a character on their own. Multiple filters allow for viewing the intersectionalities among characters.

• Visual elements. Colors, shapes, and other visual cues can highlight characters’ unique stories. 

• Pull quotes. If your data set has qualitative evidence, you can use quotes from characters to highlight quantitative data. 

Problem / Conflict

Kurnoff and Lazarus (2021) write “conflict gives your audience a reason to care. Conflict is reassuring and leads to the promised land. The path to killing status quo is conflict.” 

Just like in any good story, a problem or conflict creates a structure around which the narrative can be built. It gives the visualization a clear purpose and direction. This is obviously easier to accomplish in explanatory visualizations. However, it can be established in exploratory visualizations. For example, a comprehensive enrollment dashboard can show ups and downs in enrollment, revealing challenges and opportunities. 

Crafting a conflict helps the audience understand the consequences of the data, making the information more relatable and impactful. It shows why the data matters and how it impacts real-world situations. A clear problem or conflict also motivates the audience to think about solutions and take action. It creates a sense of urgency and importance. Resolution is a natural consequence of conflict. Problems and conflicts often evoke emotional responses, making the data more memorable and impactful. 

Here are a few practical ways to establish conflict in your data stories: 

• Incorporate conflict escalation. Humans are designed to think temporally and linearly. Use time and linear thinking to build conflict. Build stories around moments.

• Highlighting discrepancies. Use the visualization to highlight gaps and disparities that indicate a problem.

• Comparative analysis. Compare different datasets or time periods to show where conflicts arise.

• Trend analysis. Show trends or patterns that point to ongoing issues or emerging problems.

• Use of annotations. Add annotations to highlight critical points, providing context and explaining why they represent problems.

• Avoid neutral words. Bill Schander (2023) created an exercise for distinguishing between neutral and active words. “Chair” only activates the language part of the brain. However, “coffee” activates multiple parts of the brain: smell, memory, sight, etc. 

Resolution

Resolution highlights the impact of addressing the problem or conflict and inspires the audience to take action. Resolution clarifies the message by summarizing key points. 

A resolution also provides hope. For example, disaggregating student success by race, sex, or gender identity can be discouraging. A compelling narrative can provide a roadmap and inspire action by identifying an important problem that needs to be addressed.

Here are a few practical ways to include resolution in your data stories: 

• Before and after comparisons. Show data before and after to highlight the impact of specific actions or display a trend.

• Trend reversals. Illustrate how negative trends have been reversed through targeted efforts.

• Compare and contrast. Present alternate solutions and pathways. This provides the audience with a choice and gives them control. 

• Use repetition. Repeat the narrative to enhance audience engagement.

Taking reference from the literary world is not the only method by which to construct data stories. Dykes (2019) recommends following a six step approach to developing your data story project. 

Data foundation: Start by scouring your quantitative and qualitative data to identify which elements are the most relevant and appropriate to your story. At first this may be overwhelming, but try being serendipitous with your approach; explore and play around with the data. While more time consuming, some great stories may emerge. Some of the greatest scientific discoveries have been made through serendipity, so don’t be afraid to add it to your toolbox. 

Main Point: Every data story needs a central insight. Stephanie Evergreen writes we should always ask ourselves “What’s the point.” The point can be clarified by asking questions. What message are you trying to convey? What does the audience need to know? At this moment, what are the most important narratives?

In The Data Storyteller’s Handbook, Kat Greenbrook (2024) encourages researchers to clarify the purpose of their visualization by identifying one of three purposes: 

1. To discover insights in data. 

2. To inform others of the data. 

3. To educate others about the data. 

Explanatory Focus: Once the key metrics are identified, dig deeper on each to provide perspective on how and why something is occurring. Listen to stakeholders and decision-makers. Take an investigative approach and tease out what problems or stories are most compelling to them. If struggling where to start, try looking at organizational statements and issues. You can also glean insights by reviewing mission statements and strategic plans. These documents often include organizational priorities and problems that need to be addressed. For instance if school enrollment is down by 15%, providing data on economic trends or national admissions data can give a greater viewpoint on the institutional data you are reporting. 

Linear Sequence: In a data story, you want the information to unfold sequently. For instance, if you are creating a student performance dashboard, you will want to start with pre-matriculation metrics, followed by key performance indicators earned during the program. Formatting, hierarchy, and other layout decisions should guide the audience through the story. Storyboarding can also be a helpful process to determine sequencing. Ann Emery (2014) has a process for storyboarding. She focuses on building the narrative by layering from different elements of your data, one chart at a time. Bill Schander (2023) recommends stepping away from the computer and sketching out your data by hand. This is a process of trial and error, but over time stories will emerge. 

Dramatic Events: All the data in the story should lead up to this event. Viewers should be presented with a general background or a presentation of a problem or opportunity, then the story should have contributing factors and influences relevant to the problem/opportunity, followed by the aha moment where the key finding or insight is evident. This dramatic event brings viewers to the end of your story where you share solutions or next steps. When leading up to this dramatic event, always be thinking about what your audience wants or needs to hear and ensure you have all the required elements to present the full picture of the story.

Visual Anchors: Visualizations should be selected so that lay viewers are able to glean trends, patterns, and irregularities in the data. Choosing the correct visualization enhances the storytelling. More information on selecting and creating visualizations can be found in stage two of this toolkit.

For those who work in academia, it is common to want to apply your training in academic writing to your data story projects. Resist that compulsion! A typical academic paper may have this outline: 

• Abstract

• Introduction

• Lit review

• Hypothesis

• Theoretical framework

• Conceptual framework

• Methodology

• Sampling

• Data collection

• Data analysis

• Results

• Conclusion

However, you do not need to follow this outline for your data stories. In data storytelling, audience interpretation is a matter of inference, not an understanding of methodology. No one knows the data better than you. Start with the story and relegate the methodology to the background. 

Once you have customized your visualization for the target audience, there is only one proven way of confirming that you were successful: ask your audience for feedback! Establish relationships with representatives of common target audiences and ask them for formative feedback (that is, feedback you can use to improve a visualization while it is still under development) as well as summative feedback (that is, feedback on the final visualization that you can use to improve future efforts). If a significant number of individuals agree that a visualization is clear, you can be confident in its use. If, however, they don’t understand the data, meaning, or message, you can be certain that others will also find the visualization to be ineffective. In cases in which it is not appropriate to survey representative audiences, colleagues who understand the needs of potential viewers can serve as proxy reviewers to assess clarity, readability, and applicability.

One approach for transcending organizational silos is to establish a data-sharing and evaluation team, where they develop or expand their data governance structure. For example, if the team is working together to improve enrollment management, they would develop a protocol for collecting and sharing the data, including who should have access to the information and how it will be used.

Data storytelling is more than just presenting data; it is about weaving a narrative that resonates with your audience and drives action. By integrating the principles outlined in this section, you can transform raw data into meaningful stories that highlight key insights and foster informed decision-making. Remember, the power of a data story lies in its ability to connect with the audience on an emotional level, making complex information accessible and engaging. As you continue to develop your data storytelling skills, keep in mind the importance of clarity, context, and creativity. With these tools, you can effectively communicate the value of your data and inspire positive change.

Visualization specialist, Mafe Callejon (2023), shares top tips and tricks to create interactivity for a broad audience using Flourish.  The presentation covers: Why data storytelling matters, Getting started with data storytelling (Starter kit, Extra magic), Help and resources.  The following are the full presentation and the session slides: 

• Data storytelling: Transform your charts into an engaging experience (45:53)

• Session slides

Apex Global Learning for Excellence shares 22 data storytelling examples. These focus on short, one page narratives, in an infographic style.

• Examples with resource links

VISME provides an overview of data storytelling with examples and downloadable templates.

• Examples with downloadable templates

Microsoft Power BI provides an overview of how storytelling, visualizations, and dashboard can be interconnected.

• Examples using the Power BI interface

• Callejon, Mafe. 2023. Data storytelling: Transform your data into engaging experience. [YouTube]

• Cote, C. 2021. Data Storytelling: How to Effectively Tell a Story. Business Insights (Harvard Business School).

• Dykes, B. 2019. Effective Data Storytelling : How to Drive Change with Data, Narrative and Visuals. Wiley and Sons.

• Emery, A. 2014. How to present data when you’re presenting: Storyboarding your data visualizations in videos, webinars, presentations, and more.

• Greenbrook, K. 2024. The Data Storyteller’s Handbook. Rogue Penguin. 

• Gupta, Y. 2020. Dashboards and Data Stories Explained! Medium.

• Kirk, A. 2019. Data Visualizations: A Handbook for Data Driven Design. Sage.

• Knaflic, C. 2024. Storytelling with You: Plan, Create, and Deliver a Stellar Presentation. Wiley.

• Knaflic. C. 2015. Storytelling with data: A data visualization guide for business professionals. Wiley. 

• Kurnoff, J., & Lazarus, L. 2021. Everyday Business Storytelling. Wiley. 

• Schander, B. 2023. Master the Concepts of Data Visualization & Storytelling. LinkedIn Learning.