1장
The Art of Visual Storytelling: Transforming Data into Clarity
Have you ever sat through a presentation where the speaker proudly displayed a graph so complex that it might as well have been written in hieroglyphics? Cole Nussbaumer Knaflic's "Storytelling with Data" emerged from her mission to "rid the world of bad PowerPoint slides" after transforming data visualization practices at Google. This book has become the go-to resource for professionals across industries, with tech leaders like Laszlo Bock praising it as the practical complement to Edward Tufte's theoretical work. Unlike many data visualization guides that focus on technical aspects, Knaflic's approach centers on the human element-how to make numbers resonate emotionally and drive action. Her methods have been adopted by Fortune 500 companies and nonprofits alike, proving that effective data communication is a universal need in our increasingly data-driven world.
2장
Understanding Context: The Foundation of Effective Communication
Before creating any visualization, we must first understand the context-who we're communicating to, what we need them to know or do, and only then how data can help make our point. This critical foundation determines everything that follows.
When visualizing data, we must distinguish between exploratory analysis (hunting for insights) and explanatory analysis (communicating specific findings). Too often, people show all their exploratory work rather than focusing on key insights-like showing 100 oysters instead of just the two pearls found inside them.
Understanding your audience is crucial. The more specific you can be about who you're communicating to, rather than targeting general audiences, the more effective your communication will be. Consider your relationship with the audience and how they perceive you. Are you meeting for the first time, or do you have an established relationship? Do they already trust your expertise, or must you establish credibility?
After understanding your audience, determine what you need them to know or do as a result of your communication. Always have a clear purpose. As the data analyst, you're the subject matter expert in a unique position to interpret the data and guide people toward understanding and action. Even when explicit recommendations feel uncomfortable, they prompt productive conversations.
How you'll communicate affects your approach. With live presentations, you control the flow and can respond to audience cues. With written documents, the audience controls consumption, requiring more comprehensive information. Consider the tone you want your communication to convey-are you celebrating success, driving urgent action, or presenting something serious or lighthearted?
Only after clearly understanding your audience and purpose should you turn to data, asking what information will help make your point. The data becomes supporting evidence for your story.
Tools like the 3-minute story (what would you say if you had just three minutes?) and the Big Idea (a single sentence articulating your unique point of view, what's at stake, and forming a complete sentence) help distill your message to its essence. Storyboarding-creating a visual outline of your content-is perhaps the single most important upfront step to ensure your communication stays on point.
For explanatory analysis, clearly articulating who you're communicating to and what you want to convey before building content reduces iterations and ensures your communication meets its purpose. While pausing before creating content might seem to slow you down, it ensures solid understanding of your objectives, saving time later.
3장
Choosing the Right Visual: Making Your Data Speak
There are many different types of visual displays for information, but a handful will work for most needs. Looking back at over 150 visuals created for workshops and consulting projects in a year, only about a dozen different types were used regularly.
When you have just a number or two to share, simple text can be the most effective approach. Rather than diminishing impact by placing a couple numbers in a table or graph, use the numbers themselves prominently with minimal supporting text.
Tables interact with our verbal system-we read them rather than process them visually. They're ideal for mixed audiences where different people will look for their particular row of interest, or when communicating multiple different units of measure. The design should fade into the background, letting data take center stage.
A heatmap visualizes data in tabular format, using colored cells to convey the relative magnitude of numbers. This approach combines the detail of a table with visual cues that help readers quickly identify patterns.
While tables interact with our verbal system, graphs engage our visual system, which processes information faster. A well-designed graph typically communicates information more quickly than a well-designed table. The most frequently used graphs fall into four categories: points, lines, bars, and area.
Scatterplots show relationships between two variables by encoding data simultaneously on horizontal and vertical axes. Line graphs are ideal for plotting continuous data, particularly time series. Bar charts should be embraced precisely because they're common-familiarity means your audience can focus on interpreting the data rather than learning how to read the visualization.
Certain graph types should be avoided in effective data communication: pie charts, donut charts, 3D effects, and secondary y-axes. Pie charts are fundamentally flawed because humans can't accurately attribute quantitative values to two-dimensional space. When segments are close in size, it's nearly impossible to determine which is larger. Never use 3D effects unless you're actually plotting a third dimension of data. Secondary y-axes create interpretation challenges as viewers must determine which data corresponds to which axis.
The right graph is always the one easiest for your audience to read. Test your visuals with colleagues to see where they focus, what they observe, and what questions arise.
4장
Fighting Clutter: Creating Space for Clarity
Every element added to a page increases cognitive load-the mental effort required for your audience to process information. We should identify and eliminate elements that don't add sufficient informative value, as clutter makes visuals appear unnecessarily complicated and risks losing audience engagement.
Cognitive load is the mental effort required to process information. We've all experienced excessive cognitive load when confronted with busy slides or complicated graphs that make us disengage rather than invest time deciphering them. As designers of information, we must be strategic about using our audience's limited mental processing power, avoiding extraneous elements that consume resources without aiding understanding.
The Gestalt Principles, developed in the early 1900s, help us understand how people perceive visual order and can guide us in distinguishing between signal (valuable information) and noise (clutter) in our visuals:
Proximity: We perceive objects that are physically close together as belonging to the same group.
Similarity: Objects with similar color, shape, size, or orientation are perceived as related.
Enclosure: Objects physically enclosed together are perceived as a group.
Closure: People perceive a set of individual elements as a single, recognizable shape when possible.
Continuity: Our eyes naturally seek the smoothest path and create continuity.
Connection: Objects physically connected are perceived as a group.
The shift from center-aligned to left-justified text creates the biggest visual improvement in layout. Center-aligned text creates sloppy edges that make even thoughtful layouts appear messy. Creating clean horizontal and vertical lines through proper alignment helps guide the viewer's natural "z" pattern eye movement.
White space in visual communication functions like pauses in public speaking-essential for audience comfort and comprehension. Despite people's tendency to fear empty space and fill it unnecessarily, strategic white space draws attention to non-empty areas.
Clear contrast signals where viewers should focus, while lack of contrast creates visual clutter. As Colin Ware noted, a hawk is easy to spot among pigeons, but becomes harder to find as bird variety increases-similarly, when too many elements differ, none stand out.
When decluttering, consider removing chart borders (the Gestalt principle of closure allows viewers to perceive boundaries without explicit borders), gridlines (to create greater contrast that makes data stand out), data markers (when they create unnecessary cognitive load), and cleaning up axis labels (removing trailing zeros and abbreviating when possible). Label data directly to eliminate the cognitive work of going back and forth between legend and data, and leverage consistent color by making data labels the same color as the data they describe.
5장
Focusing Attention: Guiding Your Audience's Eyes
Visual perception isn't just about our eyes capturing light-it's primarily about how our brain processes this information. Three types of memory are crucial for visual communication design: iconic, short-term, and long-term memory.
Iconic memory operates unconsciously and extremely quickly-a survival mechanism evolved to detect environmental differences. Information remains here for just a fraction of a second before moving to short-term memory. Short-term memory can only hold about four chunks of visual information simultaneously, which is why complex visuals with numerous elements force audiences to work hard. Information either disappears from short-term memory or transfers to long-term memory, which combines visual and verbal components.
Preattentive attributes allow audiences to see what we want them to see before they consciously process it. When used strategically, these attributes can draw attention to specific elements and create information hierarchies. Our brains are hardwired to quickly notice differences in our environment, making these attributes powerful communication tools.
Without visual cues, we must read text sequentially. However, preattentive attributes like color, size, and style (bold, italics) can direct attention to specific content with varying degrees of emphasis. These attributes can also create visual hierarchies, making information scannable.
Size signals relative importance-elements of equal importance should be sized similarly, while critical information can be emphasized by making it larger. I learned this lesson early at Google when designing a dashboard where one data element occupied 60% of the space simply because it was available first. We realized this unintentionally emphasized less important information and redesigned the layout.
Color, when used sparingly, powerfully draws attention. Too many colors prevent anything from standing out-like trying to spot a hawk among diverse birds rather than among pigeons. Use color consistently throughout your communication so your audience learns that the highlight color indicates where to look first.
About 8% of men and 0.5% of women are colorblind, most commonly struggling to distinguish between red and green shades. Avoid using these colors together, or include additional visual cues like bold text, varying saturation, or plus/minus signs. Color evokes emotion, so choose colors that reinforce the tone you want to set.
Most audiences naturally scan pages in zigzag motions starting from the top left, making this area prime real estate for your most important information. Position elements in ways that feel natural for your audience to consume.
6장
Thinking Like a Designer: Form Follows Function
Form follows function in data visualization-first determine what you want your audience to do with the data, then create a visualization that enables this easily. In design, affordances are inherent aspects that make an object's intended use obvious-knobs afford turning, buttons afford pushing. When sufficient affordances exist, good design becomes invisible.
Effective highlighting should be limited to about 10% of your visual design to maintain impact. Various highlighting techniques include: bold text, italics, underlining, case and typeface, color, inversing elements, and size differences. These preattentive attributes can be layered for maximum impact on the most important elements.
When designing data visualizations, eliminating distractions is just as crucial as highlighting important elements. As Antoine de Saint-Exupery said, "perfection is achieved not when there's nothing more to add, but when there's nothing to take away." Not all data are equally important; summarize when detail isn't needed; remove elements that wouldn't change anything if eliminated; and push necessary but non-message-impacting items to the background using light grey.
Creating a clear visual hierarchy helps guide your audience through information in the intended order. Using the same preattentive attributes that highlight important elements, you can pull some items to the forefront while pushing others to the background.
Accessibility in data visualization means creating designs usable by people with diverse abilities and varying technical skills. Research by Song and Schwarz found that when information appears complicated, people perceive it as more difficult and are less likely to engage with it. For data visualization, this means keeping designs approachable by making them legible, clean, using straightforward language, and removing unnecessary complexity.
Text plays crucial roles in data visualization: labeling, introducing, explaining, reinforcing, highlighting, recommending, and storytelling. Every chart needs a title and every axis needs a label-these aren't optional elements. Without proper labeling, viewers waste mental energy figuring out what they're looking at rather than understanding the information.
People perceive more aesthetic designs as easier to use than less aesthetic designs-whether they actually are or not. Studies show that attractive designs are more readily accepted, promote creative thinking, foster positive relationships, and make people more tolerant of problems. To create aesthetic visualizations: (1) Be smart with color-use it sparingly and strategically; (2) Pay attention to alignment-organize elements to create clean vertical and horizontal lines; (3) Leverage white space-preserve margins and don't stretch graphics to fill empty space.
For a data visualization design to be effective, it must be accepted by its intended audience-yet people naturally resist change and cling to familiar formats. When introducing new visualization approaches, articulate the benefits, show side-by-side comparisons, provide multiple design options, and identify and convert influential audience members who can help bring others on board.
7장
The Power of Story: Making Data Memorable
Stories captivate audiences by taking them on an emotional journey that facts alone cannot achieve. A good story grabs attention and creates a lasting impression that can be easily recalled and shared. Business communications can harness this power, drawing lessons from plays, movies, and books to help tell more effective data stories.
Aristotle's three-act structure (beginning, middle, end) provides a foundational framework for storytelling, commonly known as setup, conflict, and resolution. Robert McKee, an acclaimed screenwriting expert, contrasts two methods of persuasion: conventional rhetoric (bullet points and statistics that engage only intellectually) and storytelling (which unites ideas with emotions). Kurt Vonnegut's advice on writing offers valuable lessons for data storytelling: find a subject you care about, don't ramble, keep it simple, have the courage to edit ruthlessly, be authentic, communicate clearly, and prioritize your audience's understanding.
The beginning introduces the plot and builds context, setting up essential elements like the setting, main character, problem, and desired outcome. This section should answer why the audience should care. Creating tension between "what is" and "what could be" engages the audience by giving them a stake in the solution.
The middle develops "what could be" and convinces the audience of the need for action. This section might include background information, external context, illustrative examples, supporting data, consequences of inaction, potential solutions, benefits of recommendations, and why the audience is uniquely positioned to act.
The story must conclude with a clear call to action that specifies exactly what you want the audience to do with their new understanding. An effective approach is to tie the ending back to the beginning by recapping the problem and the need for action, reinforcing any sense of urgency to motivate the audience to act.
The order of your story must be deliberately chosen based on your audience's needs. Consider whether to use a chronological approach (following the analytical process from problem to solution) or lead with the ending (starting with the call to action). Whatever flow you choose, ensure your story has a clear structure.
Repetition helps transfer information from short-term to long-term memory. The "Bing, Bang, Bongo" approach leverages this by telling your audience what you're going to tell them (introduction), telling them (main content), and then summarizing what you told them (conclusion).
Several techniques can help ensure your story comes across clearly. Horizontal logic means that reading just the slide titles throughout your deck should tell your complete story. Vertical logic ensures all elements on a single slide reinforce each other-the content supports the title, and visuals align with the text. Reverse storyboarding involves reviewing your communication by writing down the main point from each page to ensure it resembles the storyboard for the narrative you intended to tell.
8장
From Theory to Practice: Applying the Principles
The storytelling with data process transforms raw information into compelling visual narratives through a systematic approach. By understanding context, choosing appropriate visuals, eliminating clutter, focusing attention strategically, applying design principles, and crafting narratives, we move from merely showing data to truly storytelling with data. This transformation requires careful consideration of audience needs, business context, and desired outcomes at each step.
When facing specific visualization challenges, multiple solutions often exist, each with distinct advantages. For instance, when dealing with a "spaghetti graph" (a line graph with numerous overlapping lines), three effective strategies emerge: emphasizing one line at a time using preattentive attributes like color or thickness; separating lines spatially either vertically or horizontally into small multiples; or combining both approaches by separating lines while still emphasizing one series at a time. The choice depends on factors like the number of lines, the importance of direct comparison, and the story you want to tell.
Alternative visualization methods often prove more effective than traditional choices. When seeking alternatives to pie charts, several options present themselves: showing the numbers directly when simplicity is paramount and exact values matter most; using simple bar graphs for easy comparison and quick visual processing; employing 100% stacked horizontal bar graphs when part-to-whole relationships matter while maintaining comparability; or creating slopegraphs to effectively show percentage changes through line slopes. Each alternative serves different analytical purposes and narrative goals.
Design considerations shift dramatically with background color. When working with dark backgrounds (sometimes required by brand guidelines), the contrast principles reverse fundamentally - white becomes the most attention-grabbing color against black, while gray recedes into the background. Colors typically avoided on white backgrounds (like yellow) become highly effective attention-grabbers against black. Understanding these reversals helps maintain visual hierarchy and emphasis in dark-themed visualizations.
Animation serves as a powerful tool for addressing the "slideument" problem (presentations that must serve dual purposes as live delivery and standalone document). By controlling audience focus during live presentations while allowing for comprehensive takeaway materials, animation bridges this gap effectively. Starting with a blank graph and progressively revealing only the relevant data points as the narrative unfolds prevents the audience from jumping ahead and maintains their attention on the current point. This technique can be particularly effective for complex data sets or multi-step analyses.
Logical ordering of information requires strategic thinking in data visualization. When ordering categorical data in bar charts, consider what arrangement will best highlight your key insights rather than defaulting to alphabetical or size-based arrangements. For instance, ordering by magnitude can reveal patterns and extremes, while grouping related categories can show relationships and hierarchies. Time-based ordering might be appropriate for temporal data, while custom ordering could support specific narrative goals or highlight particular comparisons.
The successful application of these principles requires constant practice and iteration. Testing visualizations with target audiences, gathering feedback, and refining based on real-world usage helps ensure that the theoretical principles translate effectively into practical communication tools.
9장
Continuing Your Journey: Building Expertise
Data visualization exists at the intersection of science and art. While there are best practices and guidelines to follow, there's also significant room for creativity and individual expression. Different people will approach visualization challenges in various ways, and there's rarely a single "right" answer-rather, multiple effective paths often exist.
Applying these lessons requires practice. Look for opportunities in your work to implement these principles, whether through incremental improvements to existing work or by applying the entire storytelling process from start to finish. Don't let overambitious goals hinder progress-consider making gradual changes, such as treating existing reports as appendices while adding focused story elements at the beginning.
While the principles of effective data visualization can be applied in any tool, mastering your chosen software will prevent technical limitations from constraining your communication. The best way to learn is through hands-on use and persistence when facing challenges. Fancy tools aren't necessary-all examples in the book were created with Microsoft Excel-but options range from free Google spreadsheets to specialized tools like Tableau, programming languages like R and D3, and design software like Adobe Illustrator.
Though presented linearly, the storytelling with data process requires iteration. When unsure how to visualize data, start with paper sketches to brainstorm without technical constraints. The "optometrist approach" is useful for refining visuals: create version A, make a copy (B) with one change, then compare to see which works better. Continue this process, preserving each "best" version.
Effective data storytelling requires significant time investment at every stage-understanding context, identifying audience motivations, crafting the story, exploring different data views, decluttering, drawing attention, iterating, and building a cohesive narrative. Without consciously budgeting adequate time, the communication step often gets shortchanged after the analytical process.
Imitation is not only flattery but an essential learning tool. When you encounter effective data visualizations, study what makes them work, save examples in a visual library for inspiration, and adapt their approaches. Despite data's analytical nature, visualization offers space for creativity. Data can be made beautiful, and developing your own personal style happens naturally over time.
Organizations can build competency through three main strategies: upskilling everyone, investing in internal experts, or outsourcing. The most successful organizations use a combined approach-providing foundational training for everyone while supporting internal experts and occasionally bringing in specialists. They recognize storytelling with data as a valuable competency worth investing in.
You're now equipped to help rid the world of ineffective graphs by making the stories in your data clear to your audience. Rather than simply showing data, you'll create thoughtfully designed visualizations that impart information and inspire action.
