Storytelling with Data: A Comprehensive Guide for Business Professionals

Cole Nussbaumer Knaflic’s “Storytelling with Data” serves as an essential guide for anyone looking to communicate effectively using data. Knaflic, drawing from her extensive experience at Google and in consulting, illuminates the critical skill gap between data analysis and compelling communication. The book’s core premise is that while technology makes data ubiquitous, the ability to transform raw numbers into impactful stories is a distinct and often underdeveloped talent. This summary will comprehensively break down every important idea, example, and insight from Knaflic’s methodology, ensuring readers grasp the nuances of turning data into information that drives better decision-making.

Introduction: Bad Graphs Are Everywhere

The introduction sets the stage by highlighting the pervasive issue of ineffective data visuals in today’s business world. Knaflic points out that while we learn language and math in school, the crucial skill of combining them—telling stories with numbers—is rarely taught. This leaves individuals ill-equipped for a task increasingly demanded in data-rich environments. The proliferation of accessible graphing tools, like Excel, paradoxically exacerbates the problem; anyone can create a graph, but few understand best practices beyond tool defaults. The book aims to bridge this gap, promising to shift readers from merely “showing data” to “storytelling with data” through practical lessons and illustrative before-and-after examples.

Chapter 1: The Importance of Context

This foundational chapter emphasizes that effective data visualization begins not with data, but with understanding the situation. Before creating any visual, clarity on the context is paramount to ensure the communication is on point and successful.

Exploratory vs. Explanatory Analysis

Knaflic distinguishes between two critical types of analysis. Exploratory analysis is the process of understanding the data yourself, like “hunting for pearls in oysters,” often involving looking at data in many ways to find insights. In contrast, explanatory analysis is about communicating those insights (the “pearls”) to an audience, having a specific story to tell. The book focuses on explanatory communication, cautioning against overwhelming the audience with all exploratory findings.

Who, What, and How: Key Contextual Questions

Before touching any data, Knaflic stresses the importance of clearly answering three core questions:

• Who is your audience? This involves understanding not just who they are, but their relationship with you, their existing knowledge, and potential biases. Communicating effectively means tailoring the message to a specific audience, potentially creating different versions for different groups. Identifying the decision-maker is often key to narrowing the focus.
• What do you want your audience to know or do? This is the call to action. Knaflic encourages presenters to confidently recommend an action, even suggesting possible next steps if a direct recommendation isn’t feasible. The goal is to elicit a productive reaction and drive action, rather than simply presenting interesting data. A list of action words is provided to inspire clarity in desired outcomes.
• How will you communicate to your audience? The communication mechanism significantly impacts detail and presenter control.
  • Live presentations offer maximum control, allowing the presenter to guide attention and respond to cues, meaning less detail is needed on slides. Knaflic advises against using slides as teleprompters and emphasizes practice.
  • Written documents or emails offer less control, as the audience dictates consumption, requiring higher detail and explicit information.
  • The “slideument,” a single document trying to serve both needs, poses challenges that require strategic design.

The tone desired for the communication also falls under “how,” influencing design choices. Only after these three questions are clearly answered should one turn to the data, which then serves as supporting evidence for the story.

Consulting for Context: Asking the Right Questions

When creating a communication at someone else’s request, crucial context might be unspoken. Knaflic suggests asking specific questions to tease out this information:

• What background information is relevant or essential?
• Who is the audience or decision-maker, and what do we know about them?
• What biases might they have?
• What data is available to strengthen the case, and is the audience familiar with it?
• What are the risks or factors that could weaken the case?
• What would a successful outcome look like?
• If you had limited time, what single sentence would you say?

These last two questions, in particular, help to boil down the core message.

The 3-Minute Story & Big Idea

To achieve conciseness, Knaflic introduces two powerful concepts:

• The 3-minute story: A concise summary of what the audience needs to know if you only had three minutes. This liberates the presenter from slide dependence and ensures clarity of message regardless of time constraints.
• The Big Idea: A single, complete sentence that captures the essence of the communication. Nancy Duarte’s definition states it must articulate a unique point of view, convey what’s at stake, and be a complete sentence.

An example of the summer learning program on science illustrates how these concepts help articulate a clear message for the budget committee.

Storyboarding

Storyboarding is highlighted as the single most important upfront step. It’s a visual outline that establishes the structure of the communication. Knaflic strongly advises starting low-tech (whiteboards, Post-it notes, paper) to avoid attachment to early digital drafts, making it easier to rearrange and refine the narrative flow. An example storyboard for the summer program demonstrates this visual planning.

The chapter concludes by reinforcing that pausing to build a robust understanding of context saves time and ensures the communication meets its intended purpose, setting the stage for effective visual content.

Chapter 2: Choosing an Effective Visual

This chapter focuses on selecting the most appropriate visual display for your data. Knaflic emphasizes that while many graph types exist, a handful will suffice for the majority of business needs, and the key is to choose the visual that best enables clear communication to your audience.

The Most Commonly Used Visuals

Knaflic identifies four primary categories of visuals she uses most frequently: points, lines, bars, and area.

Simple Text

For conveying just one or two numbers, simple text is often the most effective method. Placing numbers prominently with minimal supporting words avoids the unnecessary “oomph” tables or graphs can steal. An example transforming a graph about stay-at-home moms into a concise sentence or prominent text highlights this approach, emphasizing that the sheer presence of numbers doesn’t necessitate a graph.

Tables

Tables engage our verbal system, meaning we read them. They excel at communicating to mixed audiences who seek specific data points and are ideal for displaying multiple units of measure.

• Tables in live presentations are generally discouraged because they cause the audience to read, losing their attention. Instead, Knaflic suggests pulling out key insights for visualization or moving full tables to an appendix.
• Design consideration: Table design should fade into the background, prioritizing data over heavy borders or shading. Light borders or white space are preferred for legibility and aesthetic appeal.

Heatmap

A heatmap combines the detail of a table with visual cues. It uses color saturation in tabular format to convey the relative magnitude of numbers, making it faster to identify patterns, highs, and lows. An example shows how a heatmap makes data interpretation quicker than a standard table, stressing the importance of including a legend for proper interpretation.

Graphs

Graphs interact with our visual system, which processes information faster than our verbal system. A well-designed graph typically conveys information more quickly than a table.

Points

• Scatterplot: Useful for showing the relationship between two quantitative variables, encoding data on X and Y axes. While often seen in scientific fields, they have business applications, such as analyzing miles driven versus cost per mile for a bus fleet. A modified scatterplot can highlight specific areas of interest by de-emphasizing other points.

Lines

Line graphs are best for plotting continuous data, typically over time. The connected points imply a direct relationship, making them unsuitable for categorical data.

• Standard Line Graph: Can display a single, two, or multiple series of data. Knaflic notes that too many series can become visually overwhelming.
• Showing Average within a Range: Line graphs can also effectively show summary statistics (like averages) alongside a range (e.g., min/max or confidence intervals), as demonstrated by passport control wait times.
• Consistent Intervals: When plotting time on the x-axis, data must be in consistent intervals to avoid misleading visual comparisons.

Slopegraph

Slopegraphs are valuable for comparing two time periods or points of comparison across various categories, quickly showing relative increases, decreases, or differences.

• They efficiently pack in information, intuitively conveying rate of change through the slope of lines.
• An example shows employee survey category changes, highlighting a decrease in “Career development.”
• If lines overlap excessively, emphasizing a single series while de-emphasizing others can still make them effective. Knaflic provides a template for setting up slopegraphs in Excel.

Bars

Bar charts are Knaflic’s go-to for categorical data. They are effective because they are familiar and easy for our eyes to read, allowing quick comparisons of endpoints.

• Zero Baseline: It is critical that bar charts always have a zero baseline. Violating this rule (as seen in a Fox News example) distorts visual comparisons and can be highly misleading. This rule does not apply to line graphs, though caution is still advised.
• Bar Width: Bars should be wider than the white space between them, but not so wide that they encourage area comparison.
• Graph Axis vs. Data Labels: When specific numerical values are important, labeling data points directly often allows for axis omission to reduce redundancy. For general trends, deemphasizing the axis might suffice.

Specific types of bar charts:

• Vertical Bar Chart (Column Chart): Standard choice, can be single, two, or multiple series. Visual grouping due to spacing means category order matters.
• Stacked Vertical Bar Chart: Allows comparison of totals across categories and subcomponent pieces. Can be absolute numbers or 100%. Hard to compare subcomponents beyond the bottom series due to inconsistent baselines.
• Waterfall Chart: Useful for showing the pieces of a stacked bar or illustrating a starting point, increases/decreases, and an ending point. An example shows employee headcount changes. Knaflic notes they can be created in Excel by making a stacked bar series invisible.
• Horizontal Bar Chart: Knaflic’s preferred bar chart for categorical data because it’s easy to read, especially with long category names. Text flows naturally from left to right, and eyes hit category names before the data, aiding comprehension.
• Logical Ordering of Categories: For any categorical data, categories should be ordered logically (e.g., numerically, alphabetically, or by decreasing/increasing value) to provide a clear construct for the audience, typically from top-left to bottom-right (Z-shape).
• Stacked Horizontal Bar Chart: Good for showing totals and subcomponents, especially useful for visualizing portions of a whole on a negative-to-positive scale (e.g., Likert scale survey data), as they provide consistent baselines at both ends.

Area

Knaflic generally avoids most area graphs because humans are poor at attributing quantitative value to two-dimensional space, making them hard to read. The square area graph is an exception, useful for visualizing numbers of vastly different magnitudes in a compact way due to its two dimensions (height and width).

Other Types of Graphs

Knaflic notes that while many other graph types exist (infographics included), mastering the basics is paramount. When choosing less familiar visuals, extra care is needed to make them accessible and understandable.

To Be Avoided

Knaflic identifies specific graph types and elements to avoid:

• Pie charts: Deemed “evil” due to the human eye’s inability to accurately compare angles and areas. A skewed 3D perspective further complicates interpretation. Horizontal bar charts are recommended as alternatives.
• Donut charts: Similar to pie charts, they ask the audience to compare arc lengths, which is difficult. Don’t use donut charts.
• 3D charts: Never use 3D unless plotting a genuine third dimension, and even then, it’s tricky. 3D skews numbers, making them difficult or impossible to interpret, and introduces unnecessary chart elements. Excel’s 3D plotting often misrepresents values.
• Secondary Y-axis: Generally a bad idea because it forces the audience to work hard to understand which data maps to which axis. Knaflic recommends alternatives: labeling data points directly or pulling graphs apart vertically with separate left-hand y-axes (while sharing a common x-axis).

The chapter concludes by stating that the best visual display is always what is easiest for your audience to read. Testing visuals with others helps assess effectiveness and identify areas for improvement.

Chapter 3: Clutter Is Your Enemy!

This chapter focuses on the crucial task of identifying and eliminating visual clutter from data communications, as every element added demands cognitive effort from the audience. The goal is to minimize this “cognitive load” to ensure the message is conveyed efficiently.

Cognitive Load

Cognitive load refers to the mental effort required to process new information. Excessive or extraneous cognitive load occurs when elements take up mental resources without aiding understanding. Knaflic emphasizes that designers should be smart about how they use the audience’s finite mental processing power. The aim is to minimize the perceived cognitive load, preventing the audience from dismissing a visual as too complicated before even trying to understand it.

The Data-Ink or Signal-to-Noise Ratio

To explain the concept of reducing cognitive load, Knaflic references:

• Edward Tufte’s data-ink ratio: Maximizing the proportion of a graphic’s ink devoted to data, ensuring that unnecessary elements are removed.
• Nancy Duarte’s signal-to-noise ratio: Where the “signal” is the information to communicate, and “noise” are elements that detract from or don’t add to the message.

Clutter

Clutter comprises visual elements that occupy space but do not enhance understanding. It makes visuals appear unnecessarily complicated, leading to audience discomfort and a risk of disengagement. The core reason to reduce clutter is that it makes visuals easier and more comfortable to consume.

Gestalt Principles of Visual Perception

Knaflic introduces six Gestalt Principles of Visual Perception (developed in the early 1900s) as tools to identify noise and enhance clarity. These principles describe how people instinctively perceive order from visual stimuli.

• Proximity: Objects physically close together are perceived as belonging to a group. This can be leveraged in table design to group rows or columns through spacing, directing the eye without explicit borders.
• Similarity: Objects with similar color, shape, size, or orientation are perceived as related. In tables, similar colors can guide the eye across rows.
• Enclosure: Objects physically enclosed together are perceived as a group. Even light background shading can create this effect, useful for visually separating data sections (e.g., forecast from actual data).
• Closure: People tend to perceive incomplete shapes as complete. This implies that unnecessary chart borders and background shading can be removed; the graph will still appear cohesive, and the data will stand out more.
• Continuity: Eyes seek the smoothest path and naturally create continuity. This allows for the removal of elements like Y-axis lines in bar charts, as the consistent alignment of bars still implies a baseline.
• Connection: Objects physically connected are perceived as a group, offering a strong associative value (often stronger than color, size, or shape). This is commonly used in line graphs to imply relationships between data points.

These principles help designers understand how the audience sees, enabling the removal of unnecessary elements and easing visual processing.

Lack of Visual Order

Thoughtful design, including visual order, should blend into the background. When design lacks order, it creates burden.

• Alignment: The biggest impact on visual order comes from left-justifying text (rather than center-aligning) to create clean vertical lines. Knaflic advises using presentation software rulers/gridlines for precise alignment. Diagonal lines and text should generally be avoided as they are messy and harder to read (e.g., rotated text is significantly slower to read).
• White Space: Blank space (white space) is crucial, like pauses in public speaking, for dramatic effect and audience comfort. Fear of white space leads to cramming too much information. White space should be preserved in margins, and visuals sized appropriately to their content, not stretched to fill space. It can be used strategically for emphasis.

Non-Strategic Use of Contrast

Clear contrast helps direct audience attention. Lack of clear contrast creates clutter, making it hard to discern important elements. Knaflic uses Colin Ware’s analogy of a hawk among pigeons: if everything is different, nothing stands out. When many elements compete for attention (e.g., too many colors), the message becomes muddled. An example of a “weighted performance index” graph, cluttered with different colored shapes for competitors, is transformed into a clearer horizontal bar chart that uses color strategically to highlight “Our Business” versus competitors, demonstrating effective contrast.

Decluttering: Step-by-Step

Knaflic provides a detailed, step-by-step example of decluttering a line graph showing IT support tickets (incoming vs. processed):

1. Remove chart border: Unnecessary, as per the closure principle.
2. Remove gridlines: If used, make them thin and light gray; ideally, remove them to enhance data contrast.
3. Remove data markers: They add cognitive load if data is already clear via lines. Use them only with purpose.
4. Clean up axis labels: Remove “trailing zeros” and abbreviate month names to avoid diagonal text.
5. Label data directly: Leverage the proximity principle to place labels next to the data, eliminating legend lookup.
6. Leverage consistent color: Use the similarity principle to color data labels the same as their corresponding lines, visually reinforcing the connection.

This detailed process demonstrates how identifying and eliminating clutter significantly reduces cognitive load and improves accessibility, leading to a much clearer “before-and-after” visual.

The chapter concludes by reiterating that clutter is an enemy, and by applying Gestalt principles, strategic contrast, alignment, and white space, visuals become more comprehensible and comfortable for the audience.

Chapter 4: Focus Your Audience’s Attention

This chapter delves into how humans perceive and process visual information, and how designers can leverage this understanding to effectively direct their audience’s attention using preattentive attributes. The goal is to make the audience see what you want them to see, quickly and intuitively.

You See with Your Brain

Knaflic explains that while light is captured by our eyes, the actual “seeing” and visual perception primarily occur in the brain.

A Brief Lesson on Memory

Understanding how memory works is crucial for effective visual design:

• Iconic Memory: This is a super-fast, unconscious visual memory that lasts for a fraction of a second. It’s highly sensitive to preattentive attributes, which are evolutionary mechanisms for quickly noticing differences in the environment (e.g., a predator’s motion).
• Short-Term Memory: Limited capacity, typically holding about four “chunks” of visual information at a time. This implies that complex graphs with many distinct elements (e.g., ten different colored lines with ten different marker shapes) overload this memory, forcing the audience to work too hard to decipher information. Solutions include direct labeling to reduce lookup effort and forming larger, coherent chunks of information.
• Long-Term Memory: Built over a lifetime, it’s vital for pattern recognition and general cognitive processing. It encompasses both verbal and visual memory. Knaflic notes that images can help recall verbal information, suggesting that combining visual and verbal elements aids the formation of lasting memories.

Preattentive Attributes Signal Where to Look

Preattentive attributes are powerful visual cues that our iconic memory picks up instantly, often before conscious thought.

• The “Count the 3s” Example: Knaflic demonstrates this power by asking the reader to quickly count the number of ‘3’s in a block of numbers (without cues), then repeating the exercise when the ‘3’s are colored differently. The latter is significantly faster and easier because the preattentive attribute of color intensity makes the ‘3’s instantly stand out.
• List of Preattentive Attributes: Knaflic lists various preattentive attributes, including:
  • Form: Orientation, shape, length, width, size.
  • Color: Hue, intensity (saturation).
  • Position: 2D position.
  • Motion: Flicker, direction of motion (though Knaflic notes motion is generally not recommended for explanatory data visualization).
• Quantitative vs. Categorical: It’s important to recognize that some attributes inherently convey quantitative value (e.g., line length, 2D position, size, intensity), while others are better for categorical differentiation (e.g., hue, shape).

Preattentive attributes, used sparingly, serve two main purposes:

1. Drawing audience attention quickly to desired focal points.
2. Creating a visual hierarchy of information, guiding the audience’s eye through the visual in a specific order.

Preattentive Attributes in Text

Applying preattentive attributes to text can transform how information is processed:

• Without cues, a block of text must be read entirely.
• With strategic use of bolding, different font sizes, or color, certain words or phrases immediately grab attention, making the text scannable and highlighting key information (e.g., “ATTENTION GRABBING,” “milder emphasis”).
• The varying strength of these attributes allows for creating a hierarchy, signaling what to read first, second, etc., and pushing less critical information to the background. This can convey the gist of the message within the crucial 3-8 seconds an audience spends deciding whether to engage with a visual.

Preattentive Attributes in Graphs

Preattentive attributes are equally effective in data visualizations:

• An initial “busy” bar chart showing car design concerns demonstrates how, without cues, all information competes for attention.
• By strategically using color and text, the focus can be directed to specific concerns (e.g., “engine noise” and “AC performance”).
• The same visual can be iterated upon to lead the audience from macro to micro insights, emphasizing different points sequentially (e.g., first showing overall trends, then zooming into specific areas). Knaflic notes that highlighting one aspect can make others harder to see, making this strategy suitable for explanatory analysis where a specific story is being told.

Knaflic then elaborates on three key preattentive attributes:

• Size: Relative size denotes relative importance. Important elements should be larger. Unintended size differences can lead to unintended focus, underscoring the need for intentional design choices.
• Color: One of the most powerful tools for drawing attention when used sparingly. Knaflic recommends designing visuals in shades of grey and using a single bold color (often blue, for its lack of colorblindness issues and good print quality) for emphasis.
  • Use color sparingly: Too many colors dilute the effect; effective color requires sufficient contrast (e.g., a heatmap using varying saturation of a single color is better than a rainbow palette).
  • Use color consistently: Changes in color signal changes in meaning; avoid novelty for novelty’s sake. Consistent color use trains the audience.
  • Design with colorblindness in mind: Avoid red-green combinations. Use additional cues like bolding, saturation, or +/- signs. Blue/orange is a good alternative for positive/negative.
  • Be thoughtful of tone: Color evokes emotion and should align with the desired tone (e.g., bold black for a “clinical” feel, bright colors for a “peppy” feel).
  • Brand colors: Can be leveraged, but ensure they provide sufficient contrast for emphasis. Sometimes, deviating from a weak brand color is necessary.
• Position on Page: Most audiences start at the top left and scan in zigzag patterns. This makes the top of the page precious real estate for the most important information (e.g., main takeaway, call to action, or key data). Designing with this natural viewing pattern (rather than against it) enhances comprehension.

The chapter concludes by emphasizing that preattentive attributes, when used sparingly and strategically, are powerful tools for directing attention and creating visual hierarchies, ultimately making information easier and faster for the audience to consume.

Chapter 5: Think Like a Designer

This chapter challenges readers to adopt a designer’s mindset, applying principles of traditional design to data visualization. Knaflic argues that “form follows function,” meaning visuals should be crafted to enable the audience to easily perform their intended task. The chapter explores affordances, accessibility, and aesthetics, emphasizing how good design facilitates understanding and acceptance.

Affordances

Affordances are inherent design aspects that make it obvious how a product should be used (e.g., a knob affords turning). In data visualization, visual affordances guide the audience on how to interact with and interpret the visual. Knaflic discusses three specific lessons to leverage affordances:

• Highlight the important stuff: Use preattentive attributes to draw attention to key elements. Critical here is to only highlight a small fraction (Lidwell et al. suggest at most 10%) to maintain effectiveness.
  • Techniques: Bold, italics, underlining (sparingly), uppercase text (for titles/labels), color (sparingly, often with other techniques), inversing elements (sparingly), and size.
  • Example: Remaking a Pew Research Center graph on new marriages, Knaflic demonstrates how highlighting the “Bachelor’s degree or more” category with a distinct color instantly directs attention, whereas the original design obscured the key message.
• Eliminate distractions: This aligns with the “clutter is your enemy” principle. Distractions are elements that take up space but don’t add informative value, or are irrelevant given the context.
  • Strategy: Ask, “Would eliminating this change anything?” If not, remove it. Push necessary but non-message-impacting items to the background (e.g., light grey).
  • Example: Further improving the Pew Research graph by changing from a bar to a line graph (better for trends), removing redundant “All” categories, rounding decimals, simplifying italics, and aligning elements, leading to a much cleaner visual.
• Create a clear visual hierarchy of information: Use preattentive attributes to visually prioritize elements, guiding the audience’s eye through the information in a desired order.
  • Example: A scatterplot showing car issues vs. satisfaction (Figure 5.6) demonstrates effective hierarchy through font size, color, and data point emphasis. The title (bolded keywords), axis labels, “Prior Year Average,” and then specifically the “High Satisfaction, Many Issues” (red quadrant) are prioritized, making a complex visual easy to understand.
  • Super-categories: Using labels like “High Satisfaction, Few Issues” further simplifies interpretation by providing immediate context for quadrants.

Accessibility

Accessibility in design means ensuring products are usable by people of diverse abilities, extending beyond disabilities to include people with widely varying technical skills. Knaflic emphasizes that the designer is responsible for making graphs accessible, not the user.

• Poor Design: Who Is at Fault?: When users struggle to understand a visual, it’s typically a design flaw, not the user’s fault. Good design prioritizes the user’s needs.
• Don’t Overcomplicate: “If it’s hard to read, it’s hard to do” (Song and Schwarz, 2008). Overly complicated designs are perceived as more difficult and less likely to be engaged with.
  • Tips: Make it legible (consistent, easy-to-read font/size), keep it clean (leverage affordances), use straightforward language (define jargon, spell out acronyms), and remove unnecessary complexity (favor simple over complicated).
  • Knaflic recounts an anecdote of a smart speaker whose unnecessarily complicated language alienated the audience, emphasizing that sounding smart shouldn’t make the audience feel dumb.
• Text is your friend: Thoughtful use of text ensures accessibility.
  • Essential Text: Every chart needs a title, and axes need titles (unless context is overwhelmingly clear).
  • Action Titles on Slides: Use the title bar for a clear call to action or main takeaway (e.g., “Estimated 2015 spending is above budget”) rather than a descriptive title.
  • Annotations: Text can be used directly on graphs to explain nuances, highlight points, or describe external factors.
  • Example: David McCandless’s “Peak Break-up Times According to Facebook Status Updates” uses concise, witty annotations to bring the data to life, making it instantly understandable and engaging, even without explicit axis labels.
  • Revisiting the Ticket Example: Demonstrates how adding a clear action title and specific annotations transforms a decluttered graph into a compelling, accessible visual with a clear “so what.”

Aesthetics

Knaflic argues that it is necessary to make visuals “pretty” because people perceive more aesthetic designs as easier to use, regardless of actual ease. Aesthetic designs are also more readily accepted, promote creative thinking, and foster positive relationships, making audiences more tolerant of minor design flaws (e.g., Method soap’s leaky bottle accepted due to its aesthetic appeal).

• Key elements for aesthetics:
  • Be smart with color: Use sparingly and strategically.
  • Pay attention to alignment: Create clean vertical and horizontal lines for unity.
  • Leverage white space: Preserve margins, size visuals appropriately, and use white space for emphasis.
• Example: Unaesthetic vs. Aesthetic Design: An original graph showing population breakdown by customer segments (Figure 5.12) is criticized for overuse of color, poor alignment, and misuse of white space. The redesigned version (Figure 5.13) demonstrates how thoughtful color, precise alignment, and strategic white space create an organized, respectful, and more engaging visual.

Acceptance

For any design to be effective, it must be accepted by its intended audience. Knaflic addresses common resistance to change:

• Strategies for gaining acceptance:
  • Articulate benefits: Explain why the new approach is better (e.g., new insights, improved observations).
  • Show side-by-side: Demonstrate the superiority of the new design by comparing it directly to the old.
  • Provide options & seek input: Offer multiple design choices and involve stakeholders in the decision-making process.
  • Get vocal members on board: Engage influential individuals early, solicit their feedback, and incorporate it to build buy-in.
• Testing for design issues: If resistance persists, seek input from a neutral third party to identify potential flaws in the new design itself.

The chapter concludes by empowering readers to apply design concepts: use affordances, prioritize accessibility, cultivate aesthetics, and actively work to gain acceptance for their visual designs, ultimately leading to more effective data communication.

Chapter 6: Dissecting Model Visuals

This chapter offers a practical deep dive into what “good” data visualization looks like by thoroughly examining five exemplary visuals. Knaflic dissects the deliberate thought processes and design choices behind each, reinforcing the lessons covered in previous chapters regarding context, clutter, attention, and design principles. The goal is to provide concrete archetypes for readers to emulate and learn from.

Model Visual #1: Line Graph (Company Giving Campaign)

This visual shows the progress of an annual month-long giving campaign.

• Words are used appropriately: Clear titles (graph, axes) and direct labeling of lines eliminate ambiguity, making the visual accessible.
• Attention focus: The “Progress to date” trend is emphasized using color, thicker line, a data marker, and larger text for the final point, guiding the audience’s eye.
• Context with de-emphasis: The “$50,000 goal” and “Last year’s giving” are included for reference but are pushed to the background with thin grey lines and lighter blue, preventing visual competition.
• Axis decisions: Y-axis labels retain full numeric values (e.g., $50,000) rather than being scaled down to thousands, as thinking in thousands isn’t always intuitive. X-axis labels are simplified (e.g., every 5th day) since overall trend is more important than daily specifics, reducing clutter without losing context.

Model Visual #2: Annotated Line Graph with Forecast (Annual Sales)

This visual displays actual and forecast annual sales.

• Distinguishing Actual vs. Forecast: Actual data is a solid, bold line, while forecast data is a thinner, dotted line (implying less certainty). Labels beneath the X-axis reinforce this distinction, with the forecast section subtly set apart by light background shading.
• Visual Hierarchy: Everything is pushed to the background (grey font/elements) except the graph title, blue dates in text boxes, and the data line itself, along with select data markers and labels. This ensures the eye follows the narrative flow.
• Strategic Labeling: Data markers are only included for points referenced in the annotations, clearly linking text to data. The final actual data point (e.g., $108) is bolded as it’s the anchor for the forecast. Forecast points are explicitly labeled with values for clarity.
• Text Size: Graph title is larger, axis titles are slightly larger (especially if rotated), and footnotes are smaller and greyed out to be present but not attention-grabbing.

Model Visual #3: 100% Stacked Bars (Consulting Project Progress)

This chart shows the percentage of consulting projects in “Miss,” “Meet,” or “Exceed” categories over time.

• Alignment: Graph title, legend, and Y-axis title are aligned in the upper-left, guiding the audience on how to read the graph. Text at the top right is right-justified and aligned with the relevant data bar.
• Attention focus: Red is the single attention-grabbing color (a burnt red, not primary), used for the “Miss” category. Data labels on these points are large, white, and bold for emphasis. Other categories are in different shades of grey.
• Category Ordering: Categories are ordered “Miss” to “Exceed” from bottom to top, with “Miss” closest to the X-axis for easy comparison of change over time. “Exceed” is at the top for easy comparison. The “Meet” category is harder to compare over time, but this is acceptable as it’s a lower priority.
• Words and Footnotes: Clear titles and labels, plus words at the top right, reinforce the key message. A footnote provides the total number of projects, which is crucial context for 100% stacked bars.

Model Visual #4: Leveraging Positive and Negative Stacked Bars (Director Talent Needs)

This visual displays expected needs for senior talent, including current directors, attrition, additions, and unmet need.

• Visual Path: The eye is drawn to the big, bold, black “Unmet need (gap)” numbers, then follows them right to explanatory text. Subsequent eye movement follows data series (attrition, additions) and their associated text.
• Color Use: “Today’s directors” are in a standard blue. “Attrition” is a less saturated blue, indicating a decrease by falling below the axis. “Additions” (promotions/acquisitions) are in green (positive connotation). “Unmet need” is an outline only, visually reinforcing a gap. Text labels match data series colors.
• Ordering of Series: “Today’s directors” is the baseline. Negative “Attrition” falls below. Positive “Promotions” and “Acquisitions” are above. “Unmet need” is at the very top, where the eye lands first.
• Axis and Labels: Y-axis is present but greyed out for background context. Only “Unmet need” points are directly labeled numerically and are larger/bolder.
• Text Hierarchy: Graph title is larger. Axis title is slightly larger. “Unmet need (gap)” text and numbers are largest and boldest. Footnote is smaller and grey, deemphasized.

Model Visual #5: Horizontal Stacked Bars (National Priorities Survey)

This chart presents results of survey questions on relative priorities in a developing nation.

• Stacked Bars: Appropriate for showing proportions of a whole, ordered from “top priority” (darkest shade) to “3rd priority” (lightest shade) of the same color.
• Horizontal Orientation: Ideal for long category names on the Y-axis, allowing for easy reading.
• Category Ordering: Categories are ordered vertically by decreasing “Total %”, placing the most important categories at the top.
• Data Labels and Axis: Numeric data labels are included within the bars (left-aligned for neatness, in lighter shades of blue/grey) while the X-axis is eliminated, forcing focus on the values within the bars.
• Words and Emphasis: Clear titles and labels are present. The top three priorities are explicitly emphasized through matching color for the category name, total percentage, and data bars. A legend is placed immediately above the first bar.

The chapter concludes by stressing that while different people may make different design choices, the core is to make intentional decisions based on understanding the audience and the story to be told. These examples are “haute cuisine of charts,” providing models for readers to adapt and learn from.

Chapter 7: Lessons in Storytelling

This chapter highlights the transformative power of storytelling in data communication, arguing that stories resonate and stick in ways data alone cannot. It draws lessons from plays, cinema, and written word to construct compelling narratives with data.

The Magic of Story

Knaflic opens with the “Red Riding Hood” exercise, demonstrating how stories, through their plot-twists-ending structure and repetition, are easily recalled and retold. She asserts that stories evoke emotional responses and can powerfully engage audiences in business communication.

Storytelling in Plays: The Three-Act Structure

Borrowing from Aristotle, Knaflic introduces the three-act structure of plays (setup, conflict, resolution) as a model for communication:

• Act I (Setup): Introduces the context, main character (the audience), and an incident that leads to the first turning point. This establishes the dramatic question—the call to action for the protagonist (your audience).
• Act II (Conflict): The bulk of the story, showing the protagonist’s attempts to resolve the problem. Often involves a character arc, where the protagonist develops new skills or awareness through struggle.
• Act III (Resolution): Resolves the story, including a climax (peak tension) and answers the dramatic question, leaving characters with a new understanding.

Lessons learned: This structure provides a clear framework for communication, and conflict/tension are integral to story.

Storytelling and the Cinema: Robert McKee

Knaflic references Robert McKee, a renowned screenwriting lecturer, on persuasion through storytelling:

• Conventional Rhetoric vs. Story: McKee argues that conventional rhetoric (bulleted facts, statistics) is intellectual but problematic, as audiences often mentally argue. It lacks emotional engagement. Knaflic illustrates this by reducing “Red Riding Hood” to bullet points, showing how dull it becomes.
• Story Unites Idea with Emotion: Stories arouse attention and energy by expressing how and why life changes. They start with balance, which is disrupted by an event (“subjective expectation meets cruel reality”), leading to struggle and suspense.
• Key Story Questions (McKee’s Framework):
  • What does my protagonist want (to restore balance)?
  • What is the core need?
  • What is keeping the protagonist from achieving their desire (antagonistic forces)?
  • How would the protagonist act to achieve their desire?
  • Self-reflection: “Do I believe this? Is this an honest telling?”

The meta-lesson from McKee is to use stories to emotionally engage the audience, going beyond mere facts.

Storytelling and the Written Word: Kurt Vonnegut

Knaflic draws on Kurt Vonnegut’s tips for writing:

• Find a subject you care about: Genuine caring is compelling.
• Do not ramble, though.
• Keep it simple: Profound ideas are often best expressed simply.
• Have the guts to cut: Ruthlessly edit out anything that doesn’t add new, useful illumination.
• Sound like yourself: Be authentic.
• Say what you meant to say: Be clear to be understood.
• Pity the readers: Be sympathetic, patient, and willing to simplify.

These tips reinforce simplicity, ruthless editing, authenticity, and audience-centric communication. The story is for the audience, not the presenter.

Constructing the Story

Stories provide a framework for information, tying everything together.

• The Beginning: Introduce the plot, building context. This is Act I.
  • Cliff Atkinson’s Questions (for setup): When and where? Who is driving the action (the audience)? Why is change necessary (imbalance)? What’s the desired outcome (balance)? How to achieve change (solution)?
  • Problem-Solution Frame: Frame the story around the audience’s problem and your recommended solution. Conflict and tension (the gap between “what is” and “what could be”) are crucial for engagement.
• The Middle: The core of the communication (Act II), developing “what could be” and convincing the audience of the need for action.
  • Content Ideas: Background, external context, examples, data demonstrating the problem, consequences of inaction, potential options, benefits of the solution, and making it clear why the audience is uniquely positioned to act.
  • Motivation: Frame the story in terms of what motivates your audience (e.g., making money, gaining market share).
  • “Write the headlines first”: A strategy for structuring the flow, where each headline on a Post-it note becomes a slide title, creating clear horizontal logic.
• The End: Conclude with a clear call to action. Tie back to the beginning by recapping the problem and reiterating urgency, sending the audience off ready to act.

The Narrative Structure

The narrative (written, spoken, or both) is central to successful communication, giving the communication a clear flow.

• Narrative Flow: The Order of Your Story:
  • Chronological: Taking the audience through the story as you experienced it (e.g., problem -> data gathering -> analysis -> solution). Good for establishing credibility or when process matters.
  • Lead with the Ending: Start with the call to action, then provide supporting evidence. Good when trust is established, or the audience wants the “so what” upfront. This immediately clarifies the audience’s role.
  • Clarity: The narrative path must be clear to the presenter first, then to the audience.
• The Spoken and Written Narrative:
  • Live Presentation: Spoken words reinforce visuals. Allows for dynamic responses and clarification. Challenge: avoid dense slides that distract from the speaker. It’s crucial to upfront state presentation structure and audience role (e.g., “Hold questions until the end”).
  • Written Report/Handout (Slideument): Must stand alone without the presenter. Requires more explicit written narrative and clear “so what” for each section. Getting feedback from someone unfamiliar with the topic helps ensure clarity.

The Power of Repetition

Repetition helps transfer information from short-term to long-term memory.

• Bing, Bang, Bongo: A strategy where you “tell ’em what you’re gonna tell ’em” (Bing – introduction/executive summary), “tell ’em” (Bang – main content), and “tell ’em what you told ’em” (Bongo – conclusion/summary). This feels helpful, not redundant, to the audience and cements the message.
• Repeatable Sound Bites: Nancy Duarte suggests succinct, clear, and repeatable phrases to make messages easy to recall and transfer.

Tactics to Help Ensure That Your Story Is Clear

These tactics apply mainly to presentation decks:

• Horizontal Logic: Slide titles, read sequentially, tell the overarching story. Requires action titles.
• Vertical Logic: All information on a given slide (title, words, visuals) is self-reinforcing and free of extraneous information.
• Reverse Storyboarding: After creating the final communication, flip through it and write down the main point of each page. Compare this list to your intended storyboard to identify structural gaps or areas for reorganization.
• A Fresh Perspective: Have a friend or colleague (ideally without context) review your communication. Ask them what they pay attention to, what they think is important, and their questions. This helps identify if the intended story is truly coming across.

The chapter concludes by emphasizing that the audience is the protagonist in every story communicated with data. Framing the data to be relevant to them makes it pivotal, transforming mere data display into compelling storytelling.

Chapter 8: Pulling It All Together

This chapter serves as a comprehensive demonstration of the entire storytelling with data process, applying all six lessons learned throughout the book to a single real-world example: visualizing average retail price over time for five consumer products.

Lesson 1: Understand the Context

• Scenario: Working for a startup, needing to price a new product based on competitor pricing trends.
• Who: VP of Product (primary decision-maker).
• What: Understand competitor pricing trends and recommend a price range.
• How: Show average retail price over time for Products A, B, C, D, and E.
• Big Idea: “Based on analysis of pricing in the market over time, to be competitive, we recommend introducing our product at a retail price in the range ABC–ABC–ABC– XYZ.”
• Initial Observation: “Price has declined for all products on the market since the launch of Product C in 2010” (an initial hypothesis from Figure 8.1 to be validated).

Lesson 2: Choose an Appropriate Display

• Initial Visual (Figure 8.1): A bar chart with varied colors, making trend comparison difficult.
• Remove Color Variance (Figure 8.2): Step one, removing the distracting colors, makes the bars gray.
• Emphasize 2010 Forward (Figure 8.3): Temporarily highlighting the data from 2010 onward reveals that the price decline isn’t true for all products launched later, requiring a headline modification.
• Change to Line Graph (Figure 8.4): Recognizing the primary interest is “trend over time,” switching to a line graph is more appropriate than bars, removing the artificial stairstep.
• Single Line Graph for All Products (Figure 8.5): Combining all lines on one graph eliminates redundant year labels and allows for easier comparison between products at a given point in time. This is the starting point for decluttering.

Lesson 3: Eliminate Clutter

Applying principles from Chapter 3 to Figure 8.5:

• De-emphasize chart title: Make it less bold/black.
• Remove chart border and gridlines: Reduce visual noise.
• Push axis lines and labels to background: Make them gray so they don’t compete with data. Align x-axis tick marks with data points.
• Remove variance in colors: Prepare for strategic use of color.
• Label lines directly: Eliminate the need for a separate legend, leveraging proximity.

The result is a much cleaner visual (Figure 8.6).

Lesson 4: Focus Attention Where You Want Your Audience to Focus

Using preattentive attributes on the decluttered graph (Figure 8.6) to tell different parts of the story:

• Observation 1: “After the launch of Product C in 2010, the average retail price of existing products declined.” (Figure 8.7)
  • Strategy: Highlight Product C’s launch with a data marker, and color the subsequent decline of Products A and B.
• Observation 2: “With the launch of a new product in this space, it is typical to see an initial average retail price increase, followed by a decline.” (Figure 8.8)
  • Strategy: Highlight the initial rise and fall of each product’s price post-launch.
• Observation 3: “As of 2014, retail prices have converged across products, with an average retail price of $223, ranging from a low of $180 (Product C) to a high of $260 (Product A).” (Figure 8.9)
  • Strategy: Highlight the 2014-2015 data points for each product with color and data markers.

This demonstrates how different strategic uses of color and data markers can draw attention to different aspects of the same data, allowing for a nuanced story.

Lesson 5: Think Like a Designer

Refining the visual (Figure 8.9) based on design principles:

• Accessibility with Text:
  • Simplify graph title (“Average Retail Price Trend”) and capitalize only the first word.
  • Add clear axis titles (“Average Retail Price ($)” and “Year”).
• Improve Aesthetics with Alignment:
  • Upper-left align the graph title.
  • Align y-axis title with the uppermost label and x-axis title with the leftmost label, creating clean lines.

The result is a polished, professional-looking graph (Figure 8.10).

Lesson 6: Tell a Story

Using the perfected visual (Figure 8.10) to create a compelling narrative for a live presentation (Figures 8.11-8.19):

• Beginning: Start by setting the agenda and introducing the competitive landscape.
• Middle (Plot & Twists):
  • Introduce Product A and B’s historical pricing (Figures 8.12-8.13).
  • Introduce Product C’s launch and its impact on A and B (Figure 8.14).
  • Introduce Product D’s launch and its initial pricing trajectory (Figure 8.15).
  • Introduce Product E’s launch and its initial pricing trajectory (Figure 8.16).
  • Highlight the common pattern of initial price increase then decline for new product launches (Figure 8.17).
  • Show the convergence of prices by 2014, emphasizing the current market state (Figure 8.18).
• End (Call to Action): Conclude with the recommended pricing range for the new product (Figure 8.19).

This progression shows how sequential highlighting and an accompanying narrative guide the audience through the story, from broad context to specific insights and a clear call to action. The process transforms a static graph into a dynamic, persuasive communication.

The chapter concludes with a powerful before-and-after visual (Figure 8.20), clearly illustrating the dramatic improvement achieved by applying all six lessons, moving from merely showing data to effectively storytelling with it.

Chapter 9: Case Studies

This chapter provides specific applications of the book’s lessons to common data visualization challenges through several case studies. Each case study focuses on a particular design problem and offers strategic solutions, emphasizing the iterative nature of design and the absence of a single “right” answer.

CASE STUDY 1: Color Considerations with a Dark Background

• Challenge: The default white background is often ideal for readability, but organizational brand templates sometimes mandate dark backgrounds (e.g., black or blue). Dark backgrounds can feel “heavy” and pull the eye away from data.
• Solution: If a dark background is necessary, the normal rules of contrast are reversed: lighter colors stand out more against dark.
  • Normal Approach (White Background): Darker colors (e.g., black) create strong contrast; grey is de-emphasized.
  • Dark Background Approach: White creates strong contrast; grey is de-emphasized. Some colors normally avoided on white (e.g., yellow) can be attention-grabbing on black.
  • Example: An employee survey feedback visual is initially made over for a white background, then re-made for a dark background, demonstrating how the color palette and contrast strategies must be adapted to align with the overall tone and brand.

CASE STUDY 2: Leveraging Animation in the Visuals You Present

• Challenge: A single visual often needs to serve both live presentations (where you control the narrative flow) and circulated documents (where it must stand alone). Using the same visual for both can be either too dense for presentation or not detailed enough for self-consumption.
• Solution: Use animation in live presentations to reveal data incrementally, guiding the audience’s attention, while providing a fully annotated version for circulation.
  • Presentation Strategy:
    1. Start with a blank graph (axes only) to build anticipation and ensure the audience focuses on structural elements first.
    2. Incrementally reveal data points and annotations as you speak, forcing the audience’s attention to the specific part of the story you’re telling (e.g., game user growth over time).
    3. Use PowerPoint’s simple “Appear” or “Disappear” animations (avoid flashy ones) to control visibility on a single slide.
  • Circulated Document Strategy: Provide a single, comprehensive visual with all salient points annotated directly on the graph (e.g., Figure 9.11). This allows the visual to stand alone without the presenter.
  • Benefits: Maintains presenter control during live sessions, ensures comprehensive detail for offline consumption, and allows using the same deck for both. Requires the presenter to know the story well.

CASE STUDY 3: Logic in Order

• Challenge: Categorical data, especially in bar charts, needs a logical order to facilitate audience understanding. Often, default ordering or a haphazard arrangement obscures insights.
• Solution: Always impose a logical order on categorical data unless an intrinsic order (like age ranges) already exists.
  • Impact of Ordering: The initial “User Satisfaction” horizontal stacked bar chart (Figure 9.12) has a subtle order, but its effectiveness is limited by color choices and lack of clear narrative.
  • Highlighting Specific Stories:
    • Positive Story: Reorder (or emphasize via color) to show categories with highest “Completely satisfied” + “Very satisfied” segments at the top (Figure 9.13).
    • Dissatisfaction: Reorder (or emphasize via color) to highlight categories with highest “Not satisfied at all” segments (Figure 9.14).
    • Unused Features: Reorder (or emphasize via color) to show categories with highest “Have not used” segments (Figure 9.15).
  • Maintaining Consistency for Multiple Stories: If telling multiple stories, it’s better to establish one consistent base order (Figure 9.16) and then use strategic color and annotations to highlight different aspects on successive slides (Figures 9.17-9.19). This prevents mental “tax” from re-familiarizing the audience with rearranged data.
  • Comprehensive Visual for Circulation: For a single document, combine all insights into one dense but strategically colored and annotated visual (Figure 9.20), where different colors guide the eye to different narrative points.
• Bottom Line: Always have a clear, intentional logic to the order of your data display, especially for categorical data.

CASE STUDY 4: Strategies for Avoiding the Spaghetti Graph

• Challenge: Line graphs with many overlapping lines (“spaghetti graphs”) are difficult to read and extract information from (Figure 9.21).
• Solutions:
  • Emphasize One Line at a Time: In a live presentation, de-emphasize all lines except the one being discussed using color, line thickness, and data markers/labels (Figures 9.22-9.23). This works well with a narrative voiceover.
  • Separate Spatially: Untangle lines by dividing them into multiple smaller graphs:
    • Vertical Separation: Create multiple small line graphs, all sharing the same X-axis (e.g., years) but stacked vertically (Figure 9.24). This is good for seeing individual trends (like Edward Tufte’s “sparklines”). Y-axis mins/maxes must be consistent across all sub-graphs.
    • Horizontal Separation: Create multiple small graphs, all sharing the same Y-axis (e.g., percent) but arranged horizontally (Figure 9.25). This is good for comparing relative heights of values across categories at specific points.
  • Combined Approach: Separate spatially AND emphasize one line at a time. This can be done with vertical separation (Figure 9.26) or horizontal separation (“small multiples,” Figure 9.27), where the highlighted line is bolded while others are faint. This is often better for circulated reports due to information density.
• Meta-Lesson: Don’t settle for a spaghetti graph. Understand what information is most important and choose a strategy that makes it clear. Sometimes this also means showing less data overall.

CASE STUDY 5: Alternatives to Pies

• Challenge: Pie charts are ineffective for visual comparison due to the difficulty in judging angles and areas, as discussed in Chapter 2. (Figure 9.28 shows survey data in pie charts).
• Solutions:
  • Alternative #1: Show the Numbers Directly: For a single, impactful message (e.g., “70% of kids expressed interest after the program”), simply display the numbers prominently with text (Figure 9.29).
  • Alternative #2: Simple Bar Graph: For comparing two categories (Before/After), use a simple bar graph with a common baseline, aligning values for easy comparison. This allows for clear text annotations (Figure 9.30).
  • Alternative #3: 100% Stacked Horizontal Bar Graph: Ideal when the “part-to-whole” concept is important. It provides consistent baselines at both the left and right ends, making it easy to compare segments (e.g., positive vs. negative sentiment changes in survey data) (Figure 9.31). Knaflic often retains the x-axis labels with this type.
  • Alternative #4: Slopegraph: Useful for showing percentage change between two points for multiple categories via the slope of the lines (Figure 9.32). It orders categories by their respective values. However, it doesn’t clearly show part-to-whole and doesn’t allow dictated category ordering if the values don’t align with a desired ordinal scale.
• Conclusion: There are many effective alternatives to pie charts. The best choice depends on the specific situation, how the audience should interact with the information, and the key points of emphasis.

The chapter reinforces that critical thinking is vital for solving data visualization challenges. When unsure, focusing on the audience’s needs and experimenting with different views (and seeking feedback) is the best approach.

Chapter 10: Final Thoughts

This concluding chapter reinforces the idea that data visualization is a blend of science (best practices) and art (creativity). Knaflic empowers readers to forge their own path, using artistic license to make information easier for their audience to understand.

Where to Go From Here

Knaflic emphasizes that mastery comes through practice, practice, and more practice. Readers should look for opportunities to apply the lessons incrementally in their work.

• Avoid Overhaul Overwhelm: Instead of overhauling entire reports, start with incremental improvements (e.g., treat the main report as an appendix, add a few upfront summary slides with key stories).
• Five Final Tips:
  1. Learn Your Tools Well: Don’t let tools limit your communication. Pick one (e.g., Excel, Tableau, R, D3, Illustrator) and master it. Knaflic notes all book examples were in Excel and she primarily uses PowerPoint for organizing/presenting.
  2. Iterate and Seek Feedback: The process is iterative. Start with paper sketches to brainstorm without tool constraints. Use the “optometrist approach” (A vs. B) for refinement. Seek feedback from a “fresh set of eyes” (friends, colleagues) to ensure the message lands as intended.
  3. Devote Time to Storytelling with Data: This is the only part of the analytical process the audience sees. It takes significant time to do well (understanding context, choosing visuals, decluttering, focusing attention, designing, crafting narrative). Budget for it, don’t rush.
  4. Seek Inspiration Through Good Examples: Imitate effective visualizations. Build a visual library. Learn from renowned blogs and resources (e.g., Eager Eyes, FiveThirtyEight, Flowing Data, The Functional Art, The Guardian Data Blog, HelpMeViz, Junk Charts, Perceptual Edge, Visualising Data, VizWiz, storytellingwithdata.com). Also, learn from poor examples (e.g., WTF Visualizations) by analyzing why they fail and how to improve them.
  5. Have Fun and Find Your Style: Embrace creativity in data visualization. Experiment. Develop a personal style (e.g., Knaflic’s minimalist grey and blue palette). Ensure stylistic elements enhance understanding, not hinder it.

Building Storytelling with Data Competency in Your Team or Organization

Knaflic offers strategies for fostering data communication skills organization-wide:

• Upskill Everyone: Provide foundational knowledge.
  • Ideas: Storytelling with Data book club, DIY workshops (using team’s examples), “Makeover Monday” challenges, establishing feedback loops, monthly/quarterly contests for best visualizations.
• Invest in an Internal Expert or Two: Identify interested individuals with natural aptitude, provide them with resources (books, tools, coaching), and make them in-house consultants.
• Outsource: For specific needs or when internal constraints are too high, consider external data visualization or presentation consultants. However, actively learn from the consultant during the process.
• A Combined Approach: The most successful teams combine all three strategies: universal foundational training, internal experts, and external expertise when necessary.

Knaflic encourages readers to use the book’s language to frame feedback and help others improve.

Recap: A Quick Look at All We’ve Learned

A concise review of the six core lessons:

• Understand the context: Who, what, how; Big Idea, 3-minute story, storyboarding.
• Choose an appropriate visual display: Simple text, line charts (continuous), bar charts (categorical, zero baseline); avoid pies, donuts, 3D, secondary y-axes.
• Eliminate clutter: Remove non-informative elements (Gestalt principles); strategic contrast, alignment, white space.
• Focus attention where you want it: Preattentive attributes (color, size, position); create visual hierarchy.
• Think like a designer: Affordances, accessibility, aesthetics; gain acceptance.
• Tell a story: Beginning, middle, end; conflict, narrative flow, repetition; horizontal/vertical logic, reverse storyboarding, fresh perspective.

In Closing

Knaflic concludes by reiterating that readers now possess a solid foundation and a “discerning eye” for data visualization. They are “ruined” in the best sense, ready to apply their new perspective. The ultimate message is that there is a story in your data, and the reader now has the tools to make it clear, drive better decision-making, and incite action. They are empowered to create thoughtfully designed visualizations that transform raw data into powerful narratives.

---