beautiful library

21 November 2022


Data Storytelling: Building a Compelling Narrative from Complex Data & Analytics

Gleaning insights from big data might seem intimidating—how are you supposed to sort through huge quantities of data to find the most relevant information? It’s like finding a needle in a (very large and extremely convoluted) haystack. 

But, at the core of insight-garnering is storytelling. Data storytelling can be the difference between a mess of numbers on a spreadsheet or a clean, succinct visualization that helps the audience understand what their next best action is. In fact, data storytelling is so important that Gartner predicts it will dominate the business intelligence sector by 2025.  

In this post, we’ll walk you through how to create a compelling, impactful story that highlights your data’s most critical insights and uses its power for good. 

What Is Data Storytelling? 

There’s more to data storytelling than building a compelling narrative based on data analytics. Storytelling doesn’t just inform—it influences and persuades your audience to believe in your conclusions and act accordingly. 

Data storytelling is an enormous responsibility. Data itself may not show bias, but our analytics methods and how we visualize the results can. Transforming data into a compelling narrative is equal parts data visualization and commitment to recognizing what is essential to the story. 

6 Steps to Transform Data into a Compelling Narrative 

It can be difficult to know where to start—these are the most important steps to ensuring quality data storytelling. 

1. Identify the Story That Needs to Be Told 

Storytelling begins with actually finding the story. You can ask a question or form a hypothesis, discover what data is available for the given query, and then dig in. Here are a few approaches to use: 

2. Choose the Right Data 

There’s an old but established maxim in data science: “Garbage in, garbage out.”  

Your insights or conclusions are only as good as the data you choose. It’s important not to become overwhelmed with data when constructing a story or to overfit the data and skew the results. 

For example, say you’re trying to understand a potential roadblock in your customer journey. Sales data, customer usage data, and customer feedback data would be relevant, but employee experience data would not. Understanding the role of data in your query helps to avoid cherry-picking data or minimizing scale, two things that can skew your data story. It allows you to choose the best approach for answering the question that addresses your audience’s needs.  

3. Tailor to the Audience 

One of the biggest pieces of telling a great story has nothing to do with the data itself, but instead, who you’re telling the story to. You can have the same result from a query but align the narrative with what the audience finds most valuable.  

Consider reporting sales forecasts for a particular seasonal product. The business team might be more interested in the broader business implications of a potential shortfall in inventory. In contrast, the operations team might need a greater perspective on how different operational decisions could impact the outcome. 

4. Build the Narrative 

Now that you have your audience in mind, the right data, and your driving question or concern, it’s time to build the narrative. You might be tempted to jump right to your findings (and that may be appropriate based on your target audience), but you can also tell a linear story, saving the “So what?” for last.  

For the story, you’re looking for the following: 

5. Visualize for Easy Understanding 

A picture is worth a thousand words. This is especially true in data science when you’re dealing with extremely high volumes of data. Visualizations enhance comprehension. Telling the story means knowing which visualization brings the most essential points to the surface and does not obscure the truth. Common data visualizations include: 

It’s important to choose the right visual for the story. For example, a scatter plot might look interesting visually, but a straightforward line graph will allow decision-makers to clearly see the most relevant information in a sales forecast. A scatter plot could shine, however, when it comes to visualizing customer preferences and their relationship to a certain variable.  

6. Interweave Data Throughout the Narrative 

Data on its own isn’t enough to tell a story, but it does provide the foundation for it. Interweaving the data throughout the story helps the listener understand the context, see where the insights might be headed, and avoid overwhelming them with data all at once. In addition, keeping data at the forefront of the story can help prevent internal biases from showing up. As we said in the beginning, data itself doesn’t show bias, but our methods of interpreting the story can cause implicit biases to appear.  

Place data at crucial points in the traditional story arc—the introduction, the rising action, the climax, and the resolution. This flow captures attention and helps highlight the critical context of the data and the conclusions we can draw from it. It also helps make clear how we arrive at certain conclusions using the data. 

Telling a Great Data Story Is a Critical Skill 

Telling a great story free of bias is a critical skill required of today’s data scientists. Telling a story with data means understanding profound truths about what the data says and deciding how and why to tell the story.  

With the right visualizations, data analysts and scientists can help others collaborate and “speak the language of data” more easily and leverage insights for decision-making within the organization. Data may be one of a company’s most valuable assets, but the story helps realize its full potential. 

RapidMiner works with Tableau to enable dynamic and impactful data visualizations. Simply craft the story and decide the data you need to tell it—we’ll handle the rest. Learn more and see it in action by downloading our webinar. 

Related Resources