Mastering Data Storytelling in Data Science

Have you ever watched a movie that had great actors, costumes, sets, cinematography, and special effects but a terrible plot? If so, you probably left the movie theater or walked away from your television feeling disappointed. On the other hand, if you've ever seen a low-budget movie with a great storyline, you probably ran around telling all your friends and family members about the amazing movie you just watched.

The value of a good story is important to keep in mind when you're presenting your data team's findings to others in your organization. Many data science teams make the mistake of approaching their presentations as if they're creating a big-budget film. They create beautiful graphs and slides, assuming that these visuals will engage the audience and tell the story — everything that needs to be said, they think, is conveyed by the visuals. But that's not how it works. Making something beautiful doesn't make it interesting or memorable. You need to tell a compelling story. Your story is the star of the show. Your visuals are the supporting cast.

What Is a Story?

By definition, a story is simply an accounting of incidents or events. In the context of data science storytelling, I like to think of a story as the means to making connections — connecting the dots in the data to reveal its meaning and significance and connecting the data to the audience to teach it something new.

Spinning a Yarn

To "spin a yarn" is to tell a tall tale. The phrase originated in the 1800s to describe the process of repairing rope onboard a boat. This time-consuming task involved weaving together numerous fibers. Seamen then began using the phrase to describe the telling of a long, imaginative, and typically improbable tale. Various threads must be woven together to create an entertaining and memorable story.

When your data science team sits down together to spin its own yarns, be sure to weave together the following threads:

Engage the audience by stimulating their curiosity.
Entertain the audience with stories, humor, mystery, visuals, and so on.
Educate the audience by revealing something significant that they do not already know or are not already aware of.
Transform the audience by inspiring them to embrace the recommended change in strategy, decisions, or behaviors to the benefit of the organization and themselves.

Engage

Imagine a typical presentation. The title of the opening slide is "Fourth-Quarter Sales Projections." The audience is already yawning.

Now, imagine if the slide contained only the name of the presenter. She steps forward, introduces herself, and begins to tell a story. She starts by saying "Over the past several months, sales have been rising steadily, but our team couldn't figure out why." The audience is instantly hooked. They had expected a long, boring presentation but are now about to see a mystery unfold.

Entertain

After hooking your audience, you need to keep them entertained. Starting with a mystery goes a long way toward holding the audience's attention until the big reveal at the end, but you can use other communication tools and techniques to keep the audience engaged and entertained, such as:

Relevant humor
Personal stories
Literary devices, such as similes, metaphors, and analogies
Audience participation
Interesting visuals (slides, photos, video)
Interesting questions that your team struggled to answer

Educate

In the world of data science, one key purpose of a good story is to educate the audience — to convey interesting and relevant information or insights, something the audience didn't already know. At the end of your story, you don't want anyone in the audience asking, "So, as a result of your analysis, what does our organization know now that it didn't already know?" Even worse is if the audience listens closely to the story and walks away from the presentation wondering "So what?" or "Who cares?"

When composing a story, the data science team should be sure that the story is educational as well as entertaining.

Transform

The ultimate purpose of a story is to transform the audience — to convey interesting, relevant information or insights that transform strategy, decisions, or behaviors in a positive way for the organization. When composing a story, the data science team needs to identify the main point it wants to drive home and the transformation it hopes the story initiates. In many cases, the story should end with a call to action, stating explicitly the transformation that needs to occur.

Constructing the Narrative

While the purpose of a story is to engage, entertain, educate, and transform the audience, you can use various narrative techniques to tell the story. The following five narrative techniques are particularly helpful when you're trying to explain data science concepts to your audience:

Anecdotes: An anecdote is a short personal account of something that's relevant to the larger topic. The key words here are "short" and "relevant." An anecdote is often useful at the start of your storytelling session.
Case studies: A case study is a specific instance of something used to illustrate a point. Case studies are especially helpful in presenting a problem and possible solutions. Describing a similar problem in the past that was solved can be an effective way to recommend a solution for a problem the organization is currently experiencing.
Examples: An example is a person, place, thing, or occurrence that's representative of a category or general rule. Examples are helpful for clarifying or supporting a more general idea.
Scenarios: A scenario is a fictional sequence of events that challenges the audience to consider possible outcomes. Scenarios often begin with the word "Suppose . . .". For example, "Suppose you have been waiting in a long checkout line for 15 minutes and suddenly realize that you forgot to grab the very thing you came for." You then challenge the audience to think about what they would do.
Vignettes: A vignette is a brief, evocative description, account, or episode. Think of it as a short scene in a movie. You want the audience to imagine themselves in the scene you're describing. For example, "Imagine how you'd feel if you were in a hurry to pay your bills online. You sign in only to discover that your bank completely redesigned its website. Nothing is where you expect it to be."

Storytelling with your data and analytics is an effective way to engage, entertain, educate, and transform your audience. Although having attractive data visualizations certainly helps, the story you tell will have a greater impact and leave a longer-lasting impression. By following the guidance in this post, you should be better prepared to tell great stories with your data. In subsequent posts, I will provide additional tips and suggestions.

Frequently Asked Questions

What is data storytelling?

Data storytelling is the practice of using data analysis and visualization to tell a compelling narrative. It combines the art of storytelling with the science of data to communicate insights in an engaging and understandable way.

Why is data storytelling important for a data scientist?

Data storytelling is important for data scientists because it allows them to effectively communicate their findings to a non-technical audience. This skill helps ensure that the insights derived from data analysis are understood and can drive informed decision-making.

What are the key components of data storytelling?

The key components of data storytelling include a clear narrative, compelling data visualizations, and actionable insights. These elements help in presenting data in a way that is both engaging and informative.

How can data visualizations help in telling a story with data?

Data visualizations, such as charts, graphs, and maps, can help in telling a story with data by making complex data more accessible and understandable. They highlight key trends, patterns, and outliers, allowing the audience to quickly grasp the main points of the data story.

What are the benefits of data storytelling in data science?

The benefits of data storytelling in data science include improved communication of data insights, increased engagement and understanding from the audience, and the ability to drive actionable decisions based on data. Data storytelling is a powerful tool for making data accessible and impactful.

Can you provide some data storytelling examples?

Some data storytelling examples include visualizing the impact of climate change using real-time data, presenting sales trends with interactive dashboards, and creating infographics that explain complex data sets. These examples utilize the power of data visualization and narrative to communicate insights effectively.

What are the three key elements of effective data storytelling?

The three key elements of effective data storytelling are a clear and concise narrative, visually engaging data visualizations, and actionable insights derived from the data. These elements work together to create a compelling data story that resonates with the audience.

How can data analysts and data scientists improve their data storytelling skills?

Data analysts and data scientists can improve their data storytelling skills by practicing the art of storytelling, learning to create effective data visualizations, understanding their audience's needs, and continuously seeking feedback on their presentations. Leveraging data visualization tools and staying updated with current data storytelling techniques also helps in honing these skills.

What role do data visualization tools play in data storytelling?

Data visualization tools play a crucial role in data storytelling by enabling the creation of visually appealing and informative charts, graphs, and dashboards. These tools help transform raw data into visual formats that are easier to interpret and communicate, enhancing the overall impact of the data story.

Why is having good data important for successful data storytelling?

Having good data is important for successful data storytelling because the accuracy, relevance, and quality of the data underpin the credibility of the story. Good data ensures that the insights presented are trustworthy and meaningful, helping to effectively convey the intended message to the audience.

This is my weekly newsletter that I call The Deep End because I want to go deeper than results you’ll see from searches or AI, incorporating insights from the history of data and data science. Each week I’ll go deep to explain a topic that’s relevant to people who work with technology. I’ll be posting about artificial intelligence, data science, and data ethics.

This newsletter is 100% human written 💪 (* aside from a quick run through grammar and spell check).