PDF Summary:Storytelling With Data, by Cole Nussbaumer Knaflic

Data visualizations—charts, graphs, and dashboards—are supposed to help us understand information better but often leave us more confused than enlightened. In Storytelling with Data, Google veteran Cole Nussbaumer Knaflic reveals that effective data visualization isn’t about creating the most complex charts or using the fanciest tools. Instead, it’s about making intentional choices that help your audience understand and act on your insights.

Our guide breaks down Knaflic’s framework for turning raw data into clear, actionable stories that compel your audience instead of confusing them. You’ll learn how to interpret your data effectively, craft narratives that resonate with your audience, and use design principles to enhance understanding. We’ll supplement Knaflic’s key practical examples and additional context to help you apply these principles to your work. We’ll also explore how techniques from other storytelling traditions—from fairy tales to film—can enhance your data presentations.

(continued)...

(Shortform note: Researchers have found that stories make information uniquely engaging and memorable on a neurological level. According to neuroscientist Paul Zak, compelling narratives engage more regions of the brain, making the information more impactful, compared to just stating facts. Stories that create tension and allow us to emotionally connect with characters capture our attention and evoke empathy. The brain also releases oxytocin, a neurochemical linked to trust and cooperation, when we become immersed in stories, motivating us to help others—or to work together toward a shared goal, like acting on the information data stories tell us.)

Distill the Data Down to Its Most Basic Form

Even the most interested listeners can only absorb so much information, which is why Knaflic emphasizes the importance of reducing complex information to its essential elements. The principle is simple but powerful: Before you can tell an effective story with your data, you need to identify the handful of key points that really matter. Knaflic offers two frameworks for this process of distilling your data down to its most fundamental form:

• The “3-minute story”: A concise narrative that you can tell in a short amount of time. This forces you to focus on what’s truly important and eliminate everything else.
• The “Big Idea”: A single, compelling sentence that captures your core message. Think of it as the headline you want your audience to remember.

(Shortform note: Knaflic’s “Big Idea” might be more digestible than a “3-minute story” for most audiences. Psychologists say our attention spans are getting shorter, dropping from around 2.5 minutes in 2004 to just 47 seconds in recent years. Experts attribute this to more frequent switching between tasks and digital distractions, which not only reduces our attention span but also increases our stress levels, makes us more prone to errors, and reduces our productivity since it takes mental effort to refocus. The rapid pace of modern media like TV, films, and online videos has also conditioned us to expect constantly shifting stimuli, eroding our capacity to focus on the same task—or story—for long.)

You can use a 3-minute story or a big idea to communicate your message succinctly. For example, imagine you’ve analyzed a year’s worth of customer feedback data for a subscription service. Instead of presenting every detail, you might distill it down to this Big Idea: “Our response time to customer complaints directly predicts their likelihood to renew.” Your 3-minute story would then focus on the key data points that support this conclusion, leaving out the non-essential details. By starting with this distillation process, you can then build your visualizations around the points that matter most, making your critical insights clear.

(Shortform note: Knaflic’s 3-minute story and big idea are variations of “elevator pitches”—speeches that concisely convey an idea or proposal within a very short timeframe. The term seems to have originated with Elisha Otis’s 1853 demonstration of his new elevator safety brake. Otis gave a brief “pitch” for his invention by cutting the cable while standing on the elevator platform. The phrase gained popularity in the early 20th century Hollywood film industry, where screenwriters would seize brief elevator rides to pitch movie ideas to busy executives. The term later expanded beyond the entertainment industry, becoming widely used in the business world to describe a quick, persuasive proposal.)

Choose a Narrative Structure

The second skill you need to turn the numbers into a story is choosing a narrative structure. Every great story needs a structure, but not every audience needs to go on the same journey. Some audiences will need you to talk through your process from start to finish, while other audiences just want to know what you recommend, based on your research. Knaflic explains that the narrative structure you choose should match your audience’s needs and their relationship with you. She identifies two primary approaches to structuring your data story:

The Chronological Approach

A chronological approach to your data walks your audience through your analytical process step-by-step. You might start by explaining the question you needed to answer, then go on to how you gathered your evidence, what you discovered, and finally what it all means. Knaflic explains that this structure works best when you’re building credibility with a new audience, your methodology matters as much as your findings, or your audience needs to understand your reasoning to trust your conclusions.

(Shortform note: The chronological approach Knaflic recommends mirrors what literary scholars have discovered about successful storytelling through AI analysis of narrative structures. Many acclaimed stories follow a “journey” or “quest” archetype, where protagonists move step-by-step through a process of discovery, challenge, and resolution—much like how analysts progress through data collection, analysis, and insights. Just as readers care more about a hero when they witness their full journey, audiences have more confidence in data conclusions when they can understand the analytical journey that produced them. This might explain why Knaflic finds the chronological approach particularly effective when methodology needs to be explained.)

Leading With the Call to Action

Leading with the action you want your audience to take puts your recommendations front and center. You might begin by explaining what your audience needs to do, then explain why (using data to support your argument), and then explain how the audience can make it happen. Knaflic explains that this structure works best when you have an established relationship with your audience, your audience is primarily interested in outcomes, or time is limited and decisions need to be made quickly.

(Shortform note: Writing an effective “call to action” is a strategy many marketing experts rely on. They’ve found that an effective call to action not only ​​leads with the desired action or outcome, but also creates a sense of urgency to motivate your audience to act quickly. To create a sense of urgency, highlight the benefits of the action to show how it will add value or solve a problem, and use clear, concise language with actionable verbs. You may also find it helpful to test different variations of your call to action, so you can refine it over time.)

To choose the right narrative structure, think about who your audience is. For example, imagine you’re analyzing customer churn data. With your data science team, you might use the chronological approach to explain your machine learning model’s methodology and findings. But with your CEO, you might lead with a call to action, like, “We need to invest in customer service training to reduce our 25% churn rate” and then back it up with supporting data. Whichever structure you choose, remember that your visualizations shouldn’t stand alone. Always provide clear context that connects your data to your audience’s needs and decisions.

(Shortform note: Data analyst Brent Dykes (Effective Data Storytelling) explains that when you put the insights from your data into context, you not only help your audience better understand the numbers but also put your interpretation through a valuable vetting process that can help you find any flaws. Dykes recommends six ways to provide a frame of reference for your data: comparisons that highlight similarities or differences, historical trends to show improvements or declines, analysis scaling up or down the impact to a shorter timeframe to convey significance, background information on factors that influenced the results, examples that make your data more relatable, and validation of surprising or anomalous details.)

Borrow Techniques From Other Storytellers

Data analysts aren’t the only people who need to tell compelling stories. Novelists, journalists, and filmmakers have been perfecting the art of storytelling for centuries. Knaflic suggests that we can learn from their tried-and-true techniques to make our data stories more engaging and memorable. She points out three powerful storytelling techniques that work just as well with data as they do with drama:

Repetition

Remember how fairy tales often repeat key phrases three times? Repetition isn’t just effective in stories intended for children: It’s a proven memory technique that works for data stories too. Knaflic explains that in the course of presenting your story, you can:

1. Tell your audience what you’re going to tell them
2. Relate the point you’re trying to get across
3. Remind them of what you told them

For example, in a quarterly sales presentation, you might start with an idea like, “We need to expand our West Coast operations.” Then, you could support this recommendation with a detailed market analysis. Finally, you could conclude by reiterating the West Coast expansion recommendation and summing up all of the evidence you gave in favor of this course of action.

A Logical Framework

Just as every good movie has a clear structure, your data story needs a logical framework. Knaflic recommends thinking about both horizontal and vertical logic:

• Horizontal logic: how your story flows from one point to the next (like scenes in a movie)
• Vertical logic: how each individual element (like a chart or slide) is organized to guide understanding

To check your story’s flow, Knaflic recommends going through your presentation and writing down the main point of each slide on a sticky note. Then arrange these notes to see if your story flows logically, just like a filmmaker would do when storyboarding a movie. This exercise can help you spot any flaws in the logical framework of your story and figure out how to make adjustments.

From Storyboard to Story: Lessons from Wes Anderson’s Films

Director Wes Anderson’s meticulous storyboarding process demonstrates how repetition and logical frameworks work together in storytelling. Before filming The Grand Budapest Hotel, Anderson created detailed animated storyboards that he narrated himself, allowing him to test and refine both the narrative flow (the “horizontal logic”) and the composition of individual scenes (the “vertical logic”).

This iterative process—moving from rough sketches to animated storyboards to final film—creates opportunities to see how the film can best reinforce key story points. Anderson even uses repetition, in much the same way that Knaflic recommends, within scenes: When introducing important information, he often presents it multiple ways (like showing both a character reading a letter and the letter itself). His storyboards also demonstrate how a clear logical framework helps audiences follow complex narratives: Each scene flows naturally to the next while maintaining its own internal coherence.

A Second Set of Eyes

Next, Knaflic recommends getting feedback from someone unfamiliar with your data and the story you’re trying to tell with it. She explains that getting a new perspective on your presentation can reveal:

• Where your story loses momentum
• Which points need more explanation
• What might confuse your audience

(Shortform note: Even if you don’t have another person to give you a second set of eyes on your story, as Knaflic recommends, you can still get the benefits of a fresh perspective. Psychologists have discovered the “vicarious construal effect,” where imagining an experience from a new perspective helps us see something as if it were new to us. By adopting the viewpoint of someone encountering your data for the first time, you can experience the novelty they feel and see your story as if for the first time. Researchers find that trying to see something through another person’s perspective can even help us appreciate things more—a valuable strategy to fall back on when you’ve been working with the same data for a long time.)

Step 3: Use Design to Tell Your Story

Knaflic explains that you can use thoughtful design choices to help your audience understand and connect with your data. Our brains process visual information in predictable ways, automatically noticing and interpreting certain visual elements—what Knaflic calls “preattentive attributes”—before we consciously think about them. By understanding how to use these attributes strategically, we can create visualizations that naturally guide our audience’s attention and make complex information easier to grasp.

(Shortform note: What psychologists call “preattentive features” are visual elements our brains automatically process before we consciously think about them—like how you can instantly spot a red dot in a field of blue dots without having to search for it. Preattentive features naturally capture our visual attention because they stand out from their surroundings in some way—for example, they’re different with respect to color, size, or movement. Data visualizations are most effective when they’re designed to work with, rather than against, the brain’s natural tendency to notice preattentive features first—since your audience automatically processes this information, they can grasp the concepts you’re communicating more intuitively.)

Guide Your Audience’s Attention

According to Knaflic, to create clear visual hierarchies, you must organize information so that the most important elements stand out while supporting details remain visible but don’t overwhelm. You can achieve this by leveraging three key preattentive attributes:

Size is one of your most powerful tools for creating a visual hierarchy that helps you tell your audience where to focus. Our brains automatically interpret larger elements as more important, so you can use relative size to signal what matters most. Keep elements of similar importance at similar sizes, then make crucial information proportionally larger. For instance, in a dashboard showing various sales metrics, you might make the total revenue figure significantly larger than supporting statistics.

Color provides another way to direct attention and create hierarchy. Used strategically, color can help you tell multiple stories within a single visualization. For example, in a sales performance dashboard, you might first use color to highlight which products are meeting targets (green) versus falling short (red), then switch to using different colors to distinguish between product categories (blue for electronics, purple for appliances, orange for furniture). This sequential approach helps your audience focus on one story at a time—first understanding performance issues, then exploring patterns across product lines.

Position also plays a crucial role in visual hierarchy. Knaflic says that in Western cultures, we tend to scan information from top-left to bottom-right, so placing critical information in the top-left area ensures it gets noticed first. You can also use contrast and isolation to make key elements stand out: A single data point in a bright color against a neutral background will naturally draw the eye.

For example, imagine you need to create a visualization about whale conservation that helps your audience quickly understand which species need the most urgent protection. To accomplish this, you could leverage all three preattentive attributes: First, make the critically endangered North Atlantic right whale (fewer than 500 individuals remaining) the largest element in your visualization, while showing more stable populations in smaller sizes. Second, use red to highlight critically endangered species, orange for endangered species like the blue whale, and green for recovering populations. Finally, position the most urgent cases (like the North Atlantic right whale) in the top-left.

Illuminating Insights: What Medieval Manuscripts Teach Us About Design

The visual principles Knaflic describes aren't new: Medieval manuscript scribes utilized these same preattentive attributes centuries ago. In illuminated manuscripts, size hierarchies helped readers navigate texts. Large decorated initials marked new chapters, while smaller decorated letters indicated less important transitions. Color served multiple functions: Red ink (called “rubrication”) highlighted important sections, while gold leaf drew attention to the most crucial elements. Position was equally strategic: Important text typically appeared at the top of the page, and the left margin was considered the primary starting point for decoration.

These medieval artists weren't just creating beautiful books—they were developing sophisticated information design systems that guided readers through complex texts. Their success demonstrates how deeply these visual principles are embedded in human perception, transcending both time and technology.

Make Information Easy to Process

Knaflic says that beyond guiding attention to specific elements, your design should help audiences understand how different pieces of information relate to each other. Three strategies can help you accomplish this:

1. Choose the Right Orientation: Horizontal bar charts work particularly well for categorical data, as they follow natural left-to-right reading patterns and provide ample space for labels. This reduces cognitive load (the difficulty of processing information) by working with, rather than against, how we naturally process information.

(Shortform note: While Knaflic's advice about horizontal charts following left-to-right reading patterns works well for Western audiences, it raises questions for data visualization elsewhere in the world. In languages like Arabic, Hebrew, and Persian, which read from right to left, the “natural” direction isn't so straightforward. Many Arabic-language publications maintain left-to-right data visualizations, while others flip their charts to match their reading direction. Effective data visualization must balance universal cognitive principles with cultural context. Some patterns, like the tendency to associate “up” with “more,” appear to be nearly universal, while others, like directional flow, might be more culturally influenced.)

2. Use Space Thoughtfully: Spacing is a design element that allows you to organize information in digestible chunks. Like paragraphs in writing, strategic spacing helps readers process information in meaningful chunks and understand relationships between different elements. (Shortform note: You can make data easier to digest for your audience by using consistent spacing, which helps your audience to see patterns in your data with minimal cognitive effort.)

3. Use Text Sparingly but Strategically: While visuals are the heart of data storytelling, well-chosen text enhances their impact. Use clear, informative titles that convey key messages (like “West Coast Sales Drove Q4 Growth” rather than “Sales Data 2023”). Add labels and annotations only where they illuminate key points—think of them as spotlight operators, directing attention to what matters most.

(Shortform note: Data journalist Amanda Cox’s work at The New York Times exemplifies how using annotations and text, rather than simply presenting raw data, helps audiences more easily understand data visualizations. Cox’s visualizations use words and annotations to highlight relevant patterns and expert interpretations of the data, rather than leaving the interpretation up to the audience. This makes it easier for audiences to identify and grasp the key takeaways from the data.)

Keep It Clear

Knaflic emphasizes that every element in your visualization should earn its place. Question whether each gridline, label, or decimal place truly adds value. When dealing with complex data, consider breaking it into smaller, focused views rather than creating overwhelming graphs where multiple lines tangle together. The goal isn’t to oversimplify but to help your audience see the signal through the noise.

(Shortform note: We can see Knaflic’s principles applied—and sometimes challenged—in data journalism. FiveThirtyEight, the news site founded by Nate Silver (The Signal and the Noise), is known for making complex political polling and statistical analysis accessible to general audiences. It prioritizes clarity over complexity, uses standard charts readers can easily parse, and uses scale to highlight insights. But FiveThirtyEight diverges from Knaflic’s advice to remove all non-essential information by often including context and methodological explanations to build trust with readers. This suggests that which information qualifies as “essential” might depend not just on the data itself, but on your audience's relationship with that data.)

How to Get Started With Data Storytelling: Start With the Basics

Knaflic explains that before you dive into advanced or novel visualization techniques, it’s crucial to master the fundamentals. Building a solid foundation in core visualization methods helps you to focus on establishing a shared understanding with your audience.

Knaflic recommends starting by learning to create tried-and-true visualization types like bar charts, line graphs, scatterplots, and tables. Master best practices for creating these visuals, such as choosing the right chart type, removing non-essential and potentially distracting details, and using preattentive attributes strategically. Once you’ve built proficiency with the basics, you can explore more advanced techniques if needed.

For example, let’s say you want to visualize product pricing data over time for your company and its closest competitors. A simple line graph allows you to clearly show pricing trends without overwhelming your audience. You can use line thickness, color, and direct labeling to highlight key insights. Only after mastering this fundamental visualization should you consider more novel approaches like a slopegraph or animated transitions between time periods.

Knaflic stresses that the basic principles of using data to tell a story are tool-agnostic: The practice relies on the same best practices, whatever software or platform you use to process your data and create your visualizations. While proficiency in tools like Microsoft Office is now a basic expectation across many roles and industries, what can really set your work apart is the capacity to transform data into compelling narratives that resonate with your audience and help them to take action.

(Shortform note: When you’re just starting out, it can seem like there’s an endless array of charts to learn to build. Experts say the most essential are bar charts, line charts, scatter plots, and box plots, which encode values through position, length, and area. As you progress, you’ll notice common variations on these basics, like stacked or grouped bar charts, area charts, dual-axis charts, and density curves, which incorporate additional variables or smoothing techniques. When you’ve mastered these, you can move on to specialist charts tailored for specific use cases, such as funnel charts for conversion tracking, bullet charts for performance benchmarking, and map-based visualizations for geospatial data.)

Keep Learning—and Get Others on Board, Too

Knaflic believes that data visualization and data storytelling are skills that can be taught and learned, just like any other skill that you might use in your career. She thinks it’s important to invest in training experts in your organization so you can improve your team’s skill in storytelling. You can invest in a formal training program like a workshop, a course, or a guided practice session. Or, you can provide learning resources and opportunities for members of your team who are interested in specializing in this area.

Whatever route you choose, Knaflic says you should aim to create an environment where your team actively develops, practices, and shares data storytelling skills. That might involve establishing a process for getting feedback on presentations, having coaches mentor other members of the team, discussing best practices, or running challenges to get your team to learn and apply new techniques. Developing these skills takes commitment and investment. But the payoff is immense and will enable your organization to extract maximum insight from data and communicate findings with clarity and impact.

(Shortform note: Data storytelling and visualization are valuable skills to develop, but some experts argue that cultivating a truly data-driven culture requires more than just technical training. It also involves embracing data literacy so you can use data for decision-making across all roles and levels. Data literacy doesn’t require everyone to have advanced technical skills, but rather to be comfortable using data to make informed decisions and improve their processes. Insufficient data literacy poses risks such as misinterpretation of data, poor decision-making, and missed opportunities for innovation and growth, while investing in data literacy for everyone can yield benefits like more innovative thinking.)