Impactful data stories go beyond the numbers, charts, and graphs to deliver useful insights for your audience with a clear takeaway and a bias towards action. With effective data storytelling, you help users interpret your quantitative data so they can best understand what the numbers are telling them. This shifts the focus from raw calculations, code, and analyses to problem solving, effective communication, and clear takeaways and next steps.

While the path should be simple and linear, it can be challenging to craft an engaging journey from the numbers and insights to the takeaway and call to action. The following best practices can elevate your organization’s data strategy and how it collaborates and problem solves to drive stakeholders towards decisions and actions. In this guide, we explore what good data stories look like and the steps to take to ensure your story resonates with your audience.

Why is data storytelling such a sought-after skill?

LinkedIn has named data storytelling one of the most important jobseeker attributes for several years running. If you’re asking why storytelling matters so much when it comes to data, you’re illustrating the answer to your own question.

As humans, we always want to know the “why” behind the numbers. Narratives are more memorable than numbers and connect better with our feelings, a key quality for data actionability since emotions rather than logic drive most decisions.

Meeting your audience where they are by presenting an engaging narrative can help them better understand the insight and the quantitative data you are presenting, regardless of their data literacy level. This allows your audience to build an emotional connection with the problem you are emphasizing, unlocking a bias towards a decision or action.

To foster fluency in data-driven storytelling, scientists and researchers should develop these capabilities:

• Selecting the appropriate data for an engaging, credible story
• Aligning content with organizational needs and audience preferences and knowledge levels
• Understanding how the numbers highlight the business problem and corresponding solution
• Communicating this understanding in a concise yet compelling and visually attractive way
• Empathizing with the audience and anticipating how they’ll react to the data

6 Storytelling steps for data scientists

A problem well-stated is a problem half-solved. Get to the bottom of the issue with these data storytelling steps:

1. Define the problem to be solved

First, define the business problem you are working on so that your audience can narrow their focus to the area that needs attention. Having a clear definition also helps establish the purpose of your work and gets your audience into problem-solving mode early in your presentation.

Depending on the scope of the project in question, you may want to engage internal stakeholders, external partners, and/or subject matter experts in the process of defining the business problem.

2. Understand your audience

You should also consider the audience at this planning stage to ensure that your story connects with their knowledge level, preferred communication style, and priorities. Gather as much background information as possible so that you can understand and speak to their pain points.

Most importantly, you should pragmatically evaluate whether the problem you are presenting is relevant to the audience. If it is not, then consider expanding the audience to include the stakeholders who are deeply invested in the concerning topic.

3. Gather the right data

With your business problem in clear focus, begin to gather the necessary data. This often requires a combination of discovering relevant data sources with defining the key metrics that will clearly point towards the problem and possible root causes.

As always, data must be trustworthy and accurate. The data should also be insightful and guide the user to infer causation from the displayed correlation.

4. Craft your narrative

Next, outline your narrative. Just like any good story, it should have a beginning, middle, and end. Set the stage by establishing vibrant characters and a memorable setting.

5. Present the solution

Conclude the narrative with a focused call to action that leads your audience to the necessary next steps. The story should make the audience appreciate the problem and feel a need to act towards finding a solution.

6. Develop compelling visuals

Your narrative, i.e., the data story, connects the data to the problem with the use of relevant and engaging visuals to deliver the insights. Develop visuals that clearly support the narrative and its objectives. Graphic elements must add to your message rather than distract or detract.

During this step, it’s important to select the most impactful format for the data at hand. For example:

• Pie charts and stacked bar graphs effectively show parts of a whole
• Tree map charts work best for messages involving comparison and composition
• Waterfall and area charts can be used to show changes or compare evolution over time

Often, a data story requires more than one of these elements to communicate with your audience. As you make the choices of your visuals, make sure that each one is simple and focuses on the main takeaway.

To further strengthen your narrative without distracting from your presentation, it is often a good idea to add supportive visuals to an appendix. Your audience can choose to dive deeper if they are looking to answer a related question or confirm the strength of the takeaway.

Common data storytelling mistakes to avoid

Many of the frequent techniques used to craft data narratives actually prevent the clear communication of the topic at hand. Keep your narrative focused by avoiding these common storytelling shortcomings:

Presenting too much data

When we spoke with Alan Zhang, Staff Data Scientist at Coinbase, he reflected on his years of experience of how there really is no substitute for business knowledge. Good data practitioners have a deep understanding of the business model and know how to tackle the challenge of too much data.

They apply that business knowledge to narrow their analysis scope to the sources that are most likely to deliver impactful insights into the problem at hand. Too much data not only distracts an analyst from getting to that golden insight but also creates an abundance of often overlapping insights that become repetitive or simply “too much” for the audience to process.

Using wrong or irrelevant data

Often, ineffective data storytelling uses the wrong data to draw readers’ attention. Frame the story in a way that emphasizes the solution to audience pain points without molding the available information or infusing it with personal bias.

Irrelevant data also distracts your audience and takes away from the core message you want to convey. Avoid delving into areas that don’t support the key objective you defined in the planning phase of your story: the business problem your data can solve.

Skipping steps in crafting the narrative

Crafting a story without an audience in mind also leads to issues. If you don’t know who your audience is, you can’t deliver an effective call to action that speaks to that person.

The lack of an engaging storyline also limits the actionability of data. Let’s say your company’s analytics dashboard shows a big change, such as a flood of traffic to a specific web page or a dip in sales in a month when you usually thrive. While a simple chart can illustrate this information, it doesn’t give your audience any insight about what happened and why—the details you can flesh out with a compelling narrative structure.

Delivering lifeless visuals

Data stories commonly consist of static slides combined with heavy narration, lacking the live, interactive features of platforms designed specifically for data visualization. While modern data analysis software integrates tools to create rich visualizations, these features rely on artificial intelligence and predictive machine learning. While the resulting images often look impressive, they lack the emotional core to engage audiences and produce data actionability.

These shortcomings can hamper organizational change and limit the nearly endless potential returns businesses can derive from data exploration. To take advantage of business insights, your data team must be able to deliver the three Es of data storytelling, as established in Forbes:

• Explain what you’re showing the audience and why it’s important
• Enlighten the audience by showcasing patterns, trends, and insights they wouldn’t be able to independently uncover
• Engage the audience with a compelling narrative that increases understanding and encourages action

Simple strategies can transform a bland visualization into an engaging narrative. For example, see the strategic use of color in the example below to draw the audience’s attention and focus on important points in the data to illuminate key insights.

How notebooks help tell the data story

If you don’t know about computational notebooks, you’re missing out on a powerful collaborative tool your company can use to craft compelling data stories. When we talk about notebooks in a digital sense, think beyond the battered spiral Mead notebook you remember from high school.

Computational notebooks like Jupyter Notebooks allow users to work with code, build visualizations, and write a narrative in a webpage-like fluid experience. Notebooks have been extensively used by data analysts and data scientists to publish public data analysis projects. In fact, as of Dec 2020, there were approximately 10 million notebooks published publicly on GitHub.

Open-source computational notebook tools facilitate collaboration beyond each organization’s data team. The publisher can explain their project while empowering the reader with descriptions as they step through the code. Open-source notebooks, therefore, enable analysts with varied experience and use cases to access, analyze, and visualize data in real-time in a single centralized location.

Noteable was purpose-built for effective data storytelling

Noteable was built to empower everyone with data. With that mission in mind, we are focused on helping data practitioners deliver higher impact while working in a familiar Jupyter-like interface. Noteable enables organizations and data teams to quickly work with data, collaborate around insights and data points, and deliver a rich narrative supported by visuals.

With Noteable, data teams can complete an end-to-end analytics project with rich enterprise features, including:

• Commenting tools
• Interactive data explorer
• No-code visualizations
• Data connections
• Secret storage
• Flexibility of using Python/R/SQL interchangeably
• Markdown-based text formatting
• Publishing
• … and more

These features make Noteable one of the most comprehensive data analysis and presentation tools in one, empowering data teams to create dynamic, interactive data stories that effortlessly explain, enlighten, and engage.

Notable’s user-friendly interface deploys powerful collaborative features such as:

• Smart security and version control
• Real-time commenting and collaboration
• Auto-save functionality

Plus, Noteable’s managed cloud infrastructure offers single-sign-on, access control, and privacy features so you can work securely while seamlessly collaborating across teams.

Intuitive, no-code tools with native SQL support make data exploration possible for all types of audiences, even individuals who don’t typically work in a scientific setting. This is critical since experts project that the future of BI will increasingly involve collaboration from nontraditional team members as the importance of data to meet business objectives continues to expand.

Noteable is transforming the data storytelling experience by making data more accessible, usable, and valuable to everyone. Watch our demo video to learn how to craft compelling data presentations that deliver meaningful insights and tangible results.