Book summary “The Big Book of Dashboards”: summary

Introduction

You face a common challenge when building dashboards: your screens end up cluttered, confusing, and ignored by the stakeholders who need them most. You struggle to turn raw corporate data into clear visual tools that drive actual business decisions.

Visualizing data effectively requires combining solid foundational design principles with tailored real-world dashboard scenarios across diverse industries.

Steve Wexler, Jeffrey Shaffer, and Andy Cotgreave present a comprehensive guide divided into three clear parts. You will learn the core foundations of visual perception, explore twenty-nine detailed real-world dashboard examples, and master the organizational strategies needed to deploy these tools successfully.

The book answers several vital questions to improve your workflow:

• How do you select the right chart type for your specific data?
• What design mistakes cause users to misinterpret critical metrics?
• How do you balance high-level executive summaries with detailed operational views?

1. Foundations of Data Visualization

Visualizing data is essential because numerical tables and statistical properties alone can hide critical patterns and trends. British statistician Frank Anscombe created Anscombe's Quartet in 1973 in Graphs in Statistical Analysis. It contains 11 pairs of numbers in each of the four groups of Anscombe's Quartet. These groups have identical statistical properties, yet when plotted on a chart, they reveal vastly different underlying patterns.

Effective visualization relies on understanding preattentive attributes, correct data types, color schemes, and avoiding deceptive charts like pies.

Preattentive attributes are visual properties that the human brain processes in milliseconds before conscious attention is applied. It takes 250 milliseconds or less for the brain to process preattentive attributes. Attributes such as color hue, size, and length allow you to spot anomalies or specific categories instantly across large datasets. You can spot a single red 9 among 2,500 numbers in a test grid.

Data must be classified into distinct types to determine the appropriate visual encoding:

• categorical data;
• ordinal data;
• quantitative data;
• nominal data.

Categorical data represents mutually exclusive things without numerical value. Ordinal data has a clear order with arbitrary intervals. Quantitative data consists of measurable and aggregatable numbers. Brent Spiner appeared in 177 episodes of Star Trek. His name and profession serve as categorical data, his rank as ordinal data, and his height as quantitative data. His positronic brain operates at 60 trillion operations per second.

Color in data visualizations should be applied purposefully using specific schemes rather than merely for decoration. You can use a sequential color scheme to show unemployment rates by state.

Color schemes include:

• sequential color;
• diverging color;
• categorical color;
• highlight color;
• alerting color;
• categorical-sequential color scheme.

Designers must account for color vision deficiency to ensure visualizations remain readable for individuals who cannot distinguish certain colors. Color vision deficiency (CVD) affects 8 percent of males and 0.4 percent of females. You encounter a peak rate of 11 percent among Caucasian men and a 50 percent chance of CVD in a group of nine men. You can use protanopia and deuteranopia simulations to demonstrate how traditional traffic light colors or problematic palettes merge into indistinguishable browns and grays. You can rely on color-blind friendly palettes such as blue and orange.

Pie charts, donut charts, and sized circles should be avoided because the human visual system cannot accurately estimate angles, areas, or proportional slices. Length encoded in bar charts is far more efficient than area or angle because human beings are extremely good at seeing small differences in length from a common baseline. You can compare skyscraper silhouettes versus concentric donut charts, where buildings are twisted into difficult-to-compare semicircles, including radial bar charts.

2. Academic and Operational Dashboards

Operational dashboards track specific metrics like course evaluations, wage adjustments, sales trajectories, and goal pacing using zero-baseline rules and index charts.

Academic course dashboards should track class sizes, course counts, and student ratings over time while comparing individual courses to departmental and college averages. Jeffrey Shaffer's design for the University of Cincinnati tracks enrollment history, class counts, and student ratings on a scale of 1 to 8 using a dot plot that compares a specific course against department and college averages without starting the axis at zero.

At the Carl H. Lindner College of Business at the University of Cincinnati, the Data Visualization course BANA6037 was evaluated over time to show how student growth stemmed from increasing the number of classes from two in 2012 to six in 2016. A total of 1,097 students enrolled in five years, with 687 students enrolled in 2015 to 2016. You can measure course success by checking whether people are signing up and whether students rate the class well.

• Answer whether people are signing up and whether students rate the class well to measure course success
• Use a reference line or bullet chart to show average class size or course capacity
• Apply a dot plot to compare a specific course against departmental and college benchmarks

Charts encoding data with length, height, or area must always start their axes at zero to prevent distortion, whereas charts encoding with position can sometimes break the axis away from zero. Starting at zero is mandatory for bar charts, bullet graphs, lollipop charts, and area charts because they rely on physical dimensions to represent values. Position-based charts like dot plots, box plots, and line charts can safely break away from zero when differences within a small range matter.

• Always start the axis at zero on graphs using length, height, or area
• Allow the axis to break away from zero when encoding data with position

What-if dashboards with adjustable controls allow decision-makers to immediately visualize the financial and headcount ramifications of policy changes in real time. Matt Chambers created a dashboard featuring sliders that control proposed minimum wages and required service years, instantly updating stacked bar charts, scatterplots, departmental impact bars, and normal distribution histograms.

• Place interactive sliders in the upper-left corner to adjust variables
• Show departmental impact sorted in descending order by dollar amount
• Superimpose a normal distribution bell curve over wage histograms

Index charts align multiple time periods to a common starting point at zero on the x-axis, enabling executives to evaluate cumulative growth trajectories across quarters or years instantly. By resetting every quarter or year to zero on the x-axis and tracking cumulative progress by week, an index chart lets sales managers compare current performance against previous quarters and the same period last year without being thrown off by absolute calendar dates.

• Set the x-axis to represent the week or period of the cycle starting at zero
• Draw cumulative metrics as lines for each historical period
• Thicken and color-code the lines for the current period, previous period, and same period last year while rendering older periods as thin gray background lines

3. Performance Tracking Across Industries

Complex organizations and sports teams monitor performance by combining hierarchical filters, lollipop charts, normalized distributions, and timeline tracking.

Robert Rouse designed a master dashboard for InterWorks containing hundreds of KPIs. By placing division and hierarchy filters on the side and implementing tooltip links and drill-down panes, stakeholders can view the big picture while zooming into specific operational problems without maintaining dozens of separate dashboards.

Robert Rouse's InterWorks dashboard lets users select divisions via side filters, view totals and details, and click KPI names to open detail panes showing goals, numerators, denominators, and trend lines.

• Include side filters and parameter controls to let users focus on specific divisions and business levels.
• Provide hover tooltips with definitions and commenting links.
• Enable a click-to-open detail pane showing numerator, denominator, goal status, and trend line.

Jonathan Drummey used a lollipop chart in a healthcare KPI dashboard to track nursing and infection issues. A vertical line marks the target, a circle indicates current performance relative to the target, and a connecting stick illustrates the direction and magnitude of change from the previous period.

Jonathan Drummey's healthcare dashboard at Southern Maine Healthcare tracks occurrences like falls and Clostridium difficile cases using a lollipop chart.

• Mark the target with a vertical line.
• Place a circle to the right for meeting targets and to the left for underperforming.
• Use a stick connecting the previous period's level to the current circle to show improvement or decline.

Stas Kashepava and the authors evaluated bar charts on independent scales, common scales, and percentage scales for monitoring power plant output and found them confusing or deceptive. A dot plot showing percentage difference from a 100 percent target line, with dot size encoding the absolute kW gap, accurately reflects operational impact.

In the GTA Power plant monitoring dashboard by Stas Kashepava, a facility at 92 percent of expected output actually had a larger negative impact on output than smaller facilities at 86 or 90 percent due to its larger absolute size, which dot sizing successfully revealed.

• Plot actual versus expected output as a percentage using position from a 100 percent target line.
• Encode the absolute difference in units using the size of the dots.

Steve Wexler designed a smartphone-optimized sales dashboard for managing regional performance. Placing text labels inside the bars avoids conflicts with reference lines, and line charts accompanying the bars reveal whether performance gaps are closing or widening over time.

Steve Wexler's regional sales dashboard allows a manager visiting individual stores to press a data point on a mobile phone to see monthly activity details or jump to sales systems via hyperlink.

• Place data numbers inside the horizontal bars to save space on mobile displays.
• Use reference lines to show target achievement without crowding.
• Pair summary bars with trend lines to show whether performance gaps are narrowing.

Andy Kirk designed a player performance dashboard for an English Premier League club. By using a normal distribution scale where dots on the extreme right represent exceptional performances, metrics like total distance in kilometers and high-intensity runs under a kilometer can be evaluated together.

An anonymous Premier League player receives his physical performance dashboard on his iPhone during Sunday recovery, right before Monday morning breakfast and team debriefs to encourage data-driven discussions.

• Normalize all metric scales to a normal distribution rather than a linear numerical scale.
• Orient all measures so that marks on the right side represent better performance.
• Normalize metrics to a 90-minute scale for players who participated in 70 minutes or more of a match.

Charlie Archer designed an RBS 6 Nations Championship match performance dashboard featuring play-by-play data since 2000. Line charts track cumulative scoring progress where crossing lines indicate lead changes, while pitch maps and histograms display possession duration and specific action locations.

The dashboard reveals the dramatic 2016 France versus Italy match where the lead changed eight times, as well as the 2015 Scotland versus Italy game where Italy won with a last-minute try and conversion after missing 20 tackles.

• Plot cumulative running total scores on a line chart to visualize game momentum and lead changes.
• Map specific actions like missed tackles onto a pitch background image.
• Use histograms above and below pitch maps to display possession distribution across the field.

Jeffrey Shaffer created a web analytics dashboard connecting to Google Analytics to track pageviews, visitors, and blog posts over time. By utilizing bar charts, sparkbar-style histograms without y-axis labels, and a simple dot map of user locations, the dashboard presents clear insights without visual clutter.

Jeffrey Shaffer's dashboard reveals through its unlabelled weekday histogram that website pageviews peak heavily during workdays, specifically on Tuesdays and Thursdays.

• Track core top-level metrics like pageviews, total visitors, and content counts prominently at the top.
• Use unlabelled histograms as sparkbars to show general distribution trends across weekdays.
• Display user geographic distribution on a simplified world map without country or city text labels.

Simon Beaumont designed a patient history planning tool for the NHS Southern Health Trust to assist clinicians in planning hospital discharges. Each row represents an individual patient, where colored dots along a timeline denote healthcare worker visits, hospital admissions, and discharges over a five-month period.

In the Southern Health NHS Foundation Trust dashboard, Patient 64's continuous peach dots show intensive home care visits interrupted by a hospital stay, providing clinicians with immediate historical context during daily meetings.

• Display each patient as an individual row rather than aggregating data into averages.
• Represent healthcare visits, admissions, and discharges with distinct colored dots along a horizontal timeline.
• Include hyperlinks on each dot to jump directly to detailed intervention records.

4. Specialized Monitoring and Economic Indexes

Specialized dashboards manage hospitality RevPAR, sentiment distribution, net promoter components, server monitoring, economic indices, and hex map complaints.

David Voung designed a hospitality dashboard for Dundas Data Visualization where regional map icons filter hotels and display above- or below-index performance using colored dots and triangles. By converting raw values into a 100 index baseline for diverging bar charts showing variance, you can compare different locations on maps and bar charts regardless of local currency differences.

Jeffrey Shaffer used a double-jittered dot plot in a sentiment analysis dashboard for Twitter data about banking institutions to visualize individual customer reviews. When plotting discrete numbers like a 1 to 10 sentiment scale, points stack on top of each other. Applying a random offset along the y-axis and a slight horizontal offset on the x-axis separates the dots while keeping them clustered around the true score.

• Determine the appropriate amount of jitter based on the number of points to prevent excessive overlap or data skewing.

Net Promoter Score charts must display the underlying components of promoters, passives, and detractors rather than just aggregated scores. Showing only a final Net Promoter Score hides crucial information because identical scores can derive from completely different proportions of customer feedback.

• Ask the classic NPS question on a scale from 0 to 10.
• Classify 0 to 6 responses as detractors.
• Classify 7 or 8 responses as passives or neutrals.
• Classify 9 or 10 responses as promoters.
• Compute NPS by taking the percentage of promoters, subtracting detractors, and multiplying by 100.

Mark Jackson designed a server-monitoring dashboard for Piedmont Healthcare where daily failure rates, Gantt task durations, and reference lines for scheduled start times help diagnose overnight IT failures. Placing task labels directly on Gantt bars instead of a separate legend allows administrators to instantly identify failing processes.

• Track the number of failed tasks daily.
• Review reference lines indicating scheduled start times and average durations.
• Click failed Gantt bars to reveal task performance summaries over the previous month.
• Use embedded tooltip URL actions to jump directly to the server.

Donald Hounam and Roxana Willis designed The Economist Big Mac Index dashboard, allowing you to select a base currency and switch between raw and GDP-adjusted parity indexes. Complex economic models can be communicated effectively by visualizing a single global index through multiple coordinated views.

• Place primary filters and selectors beneath the title on the top left.
• Use a map with color encodings to display spatial distributions of over- and undervalued currencies.
• Provide callouts for tiny territories that would otherwise be invisible.
• Replace scatterplots with timelines upon selecting a specific country.

Jeffrey Shaffer created a complaints dashboard for Unifund where a hex map displays open complaints by state and acts as an interactive filter for the entire view. Hexagonal state maps treat all geographic regions equally and solve the visibility problems associated with small states in traditional choropleth maps.

• Utilize summary KPI numbers at the top of the dashboard to double as color legends.
• Position the most critical segment of a stacked bar chart on the bottom axis to ensure a common baseline for precise height comparisons.
• Keep zero-complaint states visible on the hex map in white with hexagonal borders.

5. Resources, Churn, and Risk Metrics

Managing hospital rooms, sales rankings, workers compensation claims, subscriber churn, and agency utilization requires multi-view layouts and waterfall charts.

You can track scheduling accuracy and room efficiency by combining a monthly calendar heat map with a detailed Gantt chart. Dan Benevento and Katherine S. Rowell designed a hospital operating room utilization dashboard where the top section uses a diverging color palette to show daily scheduling accuracy percentages. The lower section employs a modified Gantt chart displaying scheduled versus actual case durations.

• Track on-time case starts, resource utilization, operating room turnover times, and patient satisfaction.
• Define an accurate case as one falling within 15 minutes of the estimated time.
• Use a calendar view to quickly identify scheduling trends across weeks and months.

On September 25, 2015, OR 3 ran efficiently with six cases taking between 41 and 45 minutes each, whereas OR 5 suffered a late start causing a domino effect of delays. The layout recorded 68 percent of September OR cases accurately scheduled, 77 percent in October, and 87 percent in November, alongside a 90 percent efficient scheduling rate on Thursday with 27 out of 30 cases running to time.

You can display both the overall ranking and the magnitude of sales for top products across multiple demographic dimensions to avoid the visual spaghetti of a bump chart. The Whirlpool Upright Freezer ranks fourth overall in sales, seventh among consumers, first within corporate, sixth within small office, and does not appear in the top 20 for small business.

• Show the top 20 products sorted by sales magnitude using horizontal bars.
• Provide a drop-down menu to compare rankings across dimensions like region, customer segment, and gender.
• Add secondary charts or multiple views when comparing metrics without a common baseline.

You can keep a kitchen sink dashboard uncluttered by eliminating vertical grid lines and using subtle shading and white space to define zones. Jonathan Effgen built a workers' compensation dashboard that allows you to switch between four key performance indicators on the fly and filter all nine charts simultaneously by clicking any data point. Selecting the Midwest region on the regional analysis chart reveals a distinct spike in claims during February, a higher concentration of head injuries, and 48 percent female claimants compared to 40 percent overall across 36 different combinations yielded by 9 charts and 4 KPIs.

• Provide a drop-down menu front and center to change the active KPI across all charts on the fly.
• Make every chart and element an interactive filter so clicking a data point subsets the rest of the dashboard.
• Use faint gray rules sparingly to separate filter sections without creating visual clutter.

You can bridge the gap between big-picture growth trends and monthly details by using a waterfall chart that displays cumulative running totals alongside monthly gains and losses. In February, a company gained 245 subscribers and lost 150 for a net of 95, which when added to January's ending total of 210 resulted in a running total of 305 active subscribers. Over the course of the year, the layout tracked 3,028 total subscribers gained, 1,544 total subscribers lost, and 1,484 net subscriber growth from an initial 210 subscribers welcomed in January.

• Start the first period with initial gains without losses.
• Add monthly net gains and subtract monthly losses from the running total of the previous period.
• Use heat maps or highlight tables instead of text tables to display numeric details while highlighting performance outliers.

You can expose overstaffing and opportunity costs by comparing actual billable hours against targets and labor costs using an agency utilization dashboard. Designed by Vanessa Edwards, the dashboard stacks billable and nonbillable percentages against target bars, using bullet charts and financial metrics against a 40-hour standard workweek baseline. The Creative Department operates above its billable utilization target, whereas the Technology and Project Management departments fall significantly below target, reflecting performance ranging from a 105 percent highest working hours percentage achieved by Account Management down to a 4 percent lowest departmental utilization percentage achieved by Operations.

• Display high-level financial KPIs and bullet charts at the top of the dashboard to show current fees versus potential fee targets.
• Stack billable and nonbillable hours on a common baseline with wide target background bars.
• Provide a drop-down menu to simulate what-if scenarios at varying target utilization rates.

6. Executive, Call Center, and Personalized Views

Healthcare provider productivity reports must standardize physician output and acuity using specialized metrics like wRVUs rather than raw patient counts alone. Southern Maine Health Care monitors primary care efficiency because organizational costs are several hundred thousand dollars per year. The dashboard evaluates scheduling efficiency, work relative value units, encounters per month, panel sizes, and billing ratios to identify whether providers are overworked, under-booked, or miscoding patient acuity.

Dr. Dolly Parkman of Androscoggin Family Practice has a 0.75 FTE status, 100 percent scheduling efficiency, but misses her work relative value unit targets because her appointments are too long and she sees fewer patients than the national median. Dr. Parkman averages 14.3 encounters per day compared to an 18.5 national median, achieving a 1.47 prior quarter average for wRVUs per encounter. Her 1,200 patient panel size target for a 0.75 FTE provider compares to provider Harlan Palmyra whose 938 panel size represents 67 percent of target.

• Monitor current FTE and scheduling efficiency excluding unavoidable cancellations
• Track monthly wRVU totals alongside daily averages and national medians
• Compare encounters per month and adjust targets based on FTE level
• Analyze the ratio of Level 4 to Level 3 encounters to check for proper acuity documentation

Executive dashboards for telecom operators should organize information into high-level strategic rows with supporting driver metrics and trailing labels to eliminate legend clutter. A telecom executive dashboard tracks three primary goals: reducing subscriber acquisition costs, increasing average revenue per user, and reducing churn. Each goal row pairs a main time-series chart with key risk indicators, financial summaries, and segment breakdowns, driven by year-to-date increases of $4 in subscriber acquisition costs and $6 in average revenue per user alongside an 18 basis points decrease in churn.

Mark Wevers from Dundas Data Visualization designed a telecom dashboard where trailing labels at the ends of lines indicate specific companies without requiring an external color legend.

• Organize the dashboard by high-level strategic goals in rows with breakdown drivers alongside
• Place line labels at the ends of data series to eliminate separate color legends
• Use color banding on sparklines to pack multiple quarters of data into compact spaces

Economic dashboards for macro indicators should adapt to mobile viewing through responsive vertical card stacking and incorporate nowcasting data to bridge publication lags. The Financial Times builds macroeconomic dashboards for different countries that combine journalist commentary with simple charts and source links, updating underlying economic data every 15 minutes and text commentary once a week. To handle smaller screens, cards automatically reflow from horizontal desktop layouts to vertical mobile scrolling streams without arbitrary interactivity.

The Financial Times uses nowcasting data provided by Now-Casting Economics to calculate evolving GDP changes in real time before official figures are officially published.

• Break economies into major sections with a main journalist commentary card followed by metric cards
• Stack cards horizontally on desktop browsers and reflow them vertically on narrow mobile screens
• Provide direct links to source data and relevant articles to build audience trust
• Always start bar chart axes from zero while line charts display the true minimum to maximum range

Call center dashboards displayed on large wall-mounted screens should combine automatic frequent refreshes with service-level agreement tracking and agent recognition to drive positive behavior. Call center operations require real-time visibility into queue lengths, abandoned rates, speed of answer, and closure percentages, such as a 34 percent daily inbound close rate compared to a 30 percent target, a 2 percent daily outbound close rate compared to a 15 percent target, and 481 transferred calls month-to-date. Personalizing the display by featuring top-performing agents fosters healthy competition and accountability across the floor.

Jeff Hainsworth designed a call center dashboard for Dundas Data Visualization that refreshes every 60 seconds from a database and features a scrolling ticker marquee at the bottom for employee notifications.

• Display top-performing agents with their pictures and ranks to provide recognition and motivation
• Set automatic background refreshes every 60 seconds and display the time remaining until the next update
• Use bullet graphs to compare actual performance metrics against service-level agreement target thresholds
• Include a scrolling marquee banner at the bottom for procedure reminders and goal announcements

Users engage with dashboards when data is made personally meaningful rather than through flashy, complex graphical elements like packed bubbles and treemaps.

Clients often request visually arresting dashboards with bubbles and pie charts to make boring data look cool, but these complex elements hinder quick comprehension. The true driver of user engagement is allowing viewers to filter, select parameters, and see how aggregate statistics directly pertain to them.

Chad Skelton created an interactive dashboard at the Tapestry conference allowing Canadians to compare their own age and gender against national census distributions using sliders and filters, yielding 27 people falling into an identical demographic pool based on user selections.

• Collect viewer parameters or settings to highlight personal data across all dashboard sections
• Allow viewers to click specific demographic bars to filter populations matching their own background
• Avoid adding packed bubbles, pictograms, or random interactivity purely for visual flash

Achieving professional dashboard adoption requires extensive refinement of formatting details, appropriate use of white space, and choosing well-formatted tables when data ranges are extremely wide. Eighty percent of layout work takes twenty percent of the time, while the remaining time goes into micro-tweaks like axis ranges, gray shading, and alignment. Tables outperform charts when data ranges span multiple orders of magnitude or when quarterly observations eliminate the need for continuous trendlines.

Southern Maine Health Care uses a table instead of a chart for its new versus established billed patient section because ratios range from a fraction of a percent to over ten percent across different providers, such as 4 new patients added to a panel of 424 existing patients yielding a 0.9 percent ratio.

• Use subtle gray shading for chart title areas and containers following the Gestalt principle of closure
• Extend y-axes dynamically beyond the data range to provide white space and breathing room
• Use well-formatted tables instead of charts when dealing with large data ranges or sparse quarterly points
• Apply a consistent font weight hierarchy using dark titles, medium labels, and light data source text

7. Time Dimensions and Design Best Practices

Choosing the right level of detail for time series data depends entirely on the analytical objective, as continuous and ordinal time aggregations reveal completely different stories. William Playfair drew the first statistical timeline in 1786, and 23 of 28 dashboard scenarios in the book contain a chart showing time.

Using Citi Bike data in New York, a standard timeline shows macro seasonal trends, but a view by hour of the day clearly reveals commuter peaks around 9 a.m. and 5 p.m.

• Use continuous time to track long-term trends along an x-axis.
• Use ordinal time aggregations, such as hours of the day or days of the week, to unmask cyclical patterns.
• Apply slope charts to compare values between the start and end of a time series by removing intermediate noise.
• Employ cycle plots to display cyclical patterns and sub-periods across a larger timeline.
• Use highlight tables and heat maps with color intensity across a tabular calendar grid to expose hidden seasonal trends.
• Deploy index charts to normalize different start dates onto a single anchor point to compare growth rates.
• Build jump plots to visualize sequential event workflows and identify process bottlenecks.

Dashboards become obsolete and turn into zombie tools if they do not evolve alongside changing business needs. Andy gave up using his Fitbit dashboard after a year because it stopped answering his evolving questions about long-term trends.

• Review your KPIs to ensure you are still measuring the right things and retiring metrics that no longer fail.
• Track usage with data by pulling metrics on how many times a dashboard is viewed over time.
• Speak to your users by sitting down in a room and watching them interact with the dashboard.
• Build starter dashboards that give users a basic core framework to personalize for their own specific job requirements.

Since approximately 8 percent of men and 1 percent of women suffer from color vision deficiency, relying solely on standard red and green hues makes data unreadable. Maureen Stone's Tableau color-blind-friendly palette ensures that color comparisons remain distinguishable under protanopia and deuteranopia simulations.

• Use a color-blind-friendly palette when appropriate, such as blue/orange or blue/brown combinations designed by Maureen Stone.
• Offer alternate methods of distinguishing data by adding icons, directional arrows, or explicit text labels alongside color encoding.
• Provide an option to swap the colors using checkboxes or drop-down selections for a user-switchable palette.
• Use light and dark colors to create good contrast between red and green by pairing a light green and a very dark red.

When stakeholders mandate suboptimal chart types like pies, donuts, word clouds, or bubble charts, you should redundantly encode the information alongside length-based charts to preserve analytical integrity. Facing a 17-category sales pie chart, the authors redesigned it into a single highlighted slice accompanied by a companion bar chart that clearly displayed the 85.6 percent versus 14.4 percent split.

• Take a deep breath and search YouTube for needs more cowbell when a client insists on donuts.
• Redundantly encode information by pairing mandated donut charts or series of donuts with a clean bar chart or dot plot on top.
• Use lollipop charts as a compromise for public-facing graphics to blend the analytical rigor of bar charts with the aesthetic appeal of circles.

10 Key Ideas

1. Frank Anscombe created Anscombe's Quartet in 1973 to demonstrate that statistical properties alone can hide critical patterns that are revealed only when data is visualized.
2. Preattentive attributes like color hue, size, and length allow the human brain to process and spot anomalies in milliseconds before conscious attention is applied.
3. Charts encoding data with length, height, or area must always start their axes at zero to prevent distortion, whereas position-based charts can safely break away from zero.
4. Index charts align multiple time periods to a common starting point at zero on the x-axis, enabling executives to evaluate cumulative growth trajectories instantly.
5. Applying a random offset along the y-axis and a slight horizontal offset on the x-axis in a jittered dot plot separates clustered points while keeping them aligned with the true score.
6. Net Promoter Score charts must display the underlying components of promoters, passives, and detractors rather than just aggregated scores to avoid hiding crucial feedback proportions.
7. Displaying each patient as an individual row with colored dots along a horizontal timeline provides clinicians with immediate historical context during daily care planning.
8. A waterfall chart bridges the gap between big-picture growth trends and monthly details by displaying cumulative running totals alongside monthly subscriber gains and losses.
9. Users engage with dashboards when data is made personally meaningful through filters and parameters rather than through flashy, complex graphical elements like packed bubbles.
10. The squiggle of visual exploration emphasizes that effective data visualization requires combining solid foundational design principles with tailored real-world dashboard scenarios across diverse industries.