18 Creating Basic Visualization and Formatting Tableau Visualization

What This Chapter Covers

Building a chart in Tableau takes seconds; building a chart that communicates clearly takes deliberate decisions about chart type, mark properties, colour, and formatting. This chapter covers the end-to-end process of creating basic but professional visualizations in Tableau Desktop: choosing the right chart type for your data, building bar charts, line charts, scatter plots, and maps from scratch, and then applying the full range of formatting options, colours, fonts, borders, tooltips, axis formatting, and reference lines, to produce publication-ready outputs. The focus is on practical, step-by-step construction combined with principled formatting decisions that improve clarity and reduce noise.

flowchart LR
    A[Data <br> in Data Pane] --> B[Choose <br> Chart Type]
    B --> C[Drag Fields <br> to Shelves]
    C --> D[Set Mark <br> Properties]
    D --> E[Format <br> Axes & Labels]
    E --> F[Apply <br> Colour & Style]
    F --> G[Polished <br> Visualization]
    style A fill:#e3f2fd,stroke:#1976D2
    style D fill:#fff9c4,stroke:#F9A825
    style G fill:#e8f5e9,stroke:#388E3C

18.1 Choosing the Right Chart Type

The Chart Type Decision Framework

Before building in Tableau, select the chart type based on the analytical question and the data structure. This decision determines every subsequent choice, shelf placement, mark type, and formatting approach.

Analytical Question	Recommended Chart	Mark Type in Tableau
How does a measure compare across categories?	Bar chart	Bar
How does a measure change over time?	Line chart	Line
What is the relationship between two measures?	Scatter plot	Circle
How does part contribute to a whole?	Treemap or pie	Square / Circle
Where do things happen geographically?	Map	Map (auto)
What is the distribution of a measure?	Histogram or box plot	Bar / Circle
How do multiple KPIs compare at a glance?	Bullet chart or text table	Bar / Text

The two most common mistakes in chart selection: using a pie chart when there are more than four categories (a bar chart is almost always better), and using a line chart for categorical data that has no natural ordering (a bar chart is correct, line charts imply continuity between points).

Tableau’s Show Me Panel as a Starting Guide

Show Me (the panel in the top-right of the Tableau workspace) suggests valid chart types based on the fields currently selected in the Data pane. It greys out chart types that require a different data structure. While Show Me is a helpful starting point, it defaults to the most statistically valid chart, not always the most communication-effective one. Use it as a suggestion, not a rule: a simple bar chart is often clearer than the dual-axis chart Show Me recommends.

18.2 Building a Bar Chart

How To: Creating a Sorted Bar Chart

A bar chart is the workhorse of business analytics, clear, precise, and universally understood. Build a category-vs.-sales bar chart:

Drag Sub-Category to the Rows shelf.
Drag Sales to the Columns shelf. Tableau creates horizontal bars automatically (when a dimension is on Rows and a measure on Columns, the default is a horizontal bar chart).
Click the Sort Descending button on the Sub-Category axis label (the sort icon that appears when you hover over the axis header) to sort bars from highest to lowest sales. This is the single most important formatting step for bar charts, sorted bars allow the reader to rank categories immediately.
To change to vertical bars: drag the fields from Rows to Columns and Columns to Rows (or use the Swap Rows and Columns toolbar button).

[Insert screenshot of a horizontal sorted bar chart: Sub-Category on Y axis, Sales on X axis, bars sorted from longest to shortest, with the sort icon visible on the axis header]

Adding Colour to Bars by Dimension

Colouring bars by a second dimension creates a stacked or grouped bar chart, allowing comparison across two categorical variables simultaneously.

Stacked bar chart: 1. Drag a second dimension (e.g., Segment) to the Colour shelf on the Marks card. 2. Each bar is subdivided by Segment with a different colour per segment. 3. Best for showing part-to-whole relationships within each category.

Side-by-side (grouped) bar chart: 1. Drag Segment to the Columns shelf alongside Sub-Category (inside it, as a second level of nesting). 2. Tableau groups bars by Sub-Category, with one bar per Segment within each group. 3. Best for comparing absolute values across both dimensions simultaneously.

Choosing between stacked and grouped: Use stacked when the total is important and the individual parts add to a meaningful whole. Use grouped when comparing specific values across the second dimension is more important than the total.

18.3 Building a Line Chart

How To: Creating a Time-Series Line Chart

Line charts are the correct choice for continuous data over time. Build a monthly sales trend line:

Drag Order Date to the Columns shelf. Tableau defaults to Year, click the pill and select Month (the continuous version, not the discrete MONTH() function). The axis shows a continuous time series.
Drag Sales to the Rows shelf. A line chart appears automatically.
To add a second measure on the same chart: drag Profit to the Rows shelf. Tableau creates two separate panels (one per measure). To merge them onto one axis, drag Profit to the existing Sales axis until the green “+” appears, then drop, this creates a dual-axis chart with two Y-axes.
Right-click the secondary Y-axis and select Synchronise Axis if you want both measures to share the same scale.

[Insert screenshot of a monthly sales line chart with Order Date (Month, continuous) on X axis, SUM(Sales) on Y axis, and a reference line at the average value]

Styling Line Charts: Paths, Size, and End Labels

A raw Tableau line chart needs several styling decisions to reach publication quality:

Line weight: In the Marks card, click Size and drag the slider to increase line thickness. A weight of 2–3px is usually optimal, thin enough to show the trend clearly, thick enough to read on screen or in print.

Mark shapes on the line: Click Shape in the Marks card and select a circle or square. This adds a dot at each data point, improving readability for discrete monthly data.

End labels: Right-click the line and select Mark Labels > Always Show. This places a value label at the final data point, immediately communicating the most recent value without requiring the reader to read the axis.

Smoothing: Right-click the line > Trend Lines > Show Trend Lines to overlay a linear or polynomial regression trend line showing the long-run direction of the data.

18.4 Building a Scatter Plot

How To: Creating a Scatter Plot

A scatter plot reveals the relationship between two measures across a set of entities. Build a Sales vs. Profit scatter by Sub-Category:

Drag Sales to the Columns shelf.
Drag Profit to the Rows shelf.
Drag Sub-Category to the Detail shelf on the Marks card. Each Sub-Category becomes a separate mark (dot).
Drag Sub-Category to the Label shelf to label each dot with its name. Set label to Selected marks only to avoid clutter.
Drag Sales (again) to the Size shelf to size each dot by its sales volume, creating a bubble chart variant.

Adding a reference line: Go to Analysis > Add Reference Line. Add a line at the average of Profit (horizontal) and at the average of Sales (vertical). These two lines divide the scatter into four quadrants, a powerful analytical frame for segment analysis.

[Insert screenshot of a scatter plot with Sales on X, Profit on Y, Sub-Category labels, and average reference lines creating four quadrants, labelling quadrants as “High Sales / High Profit”, etc.]

18.5 Axis Formatting

Formatting Axes for Clarity

Axis formatting has more impact on chart readability than almost any other single formatting decision. In Tableau, right-click any axis to access the axis formatting options.

Edit Axis dialog (double-click the axis):

Option	Recommendation
Range: Fixed	Use when comparing across sheets with the same scale, prevents automatic rescaling that misleads comparisons
Range: Automatic	Use for standalone charts where the data range matters
Include Zero	Always include zero for bar charts (a bar not starting at zero is misleading); optional for line charts
Axis Title	Remove the auto-generated “SUM(Sales)” and write a meaningful title: “Total Revenue (USD)”
Tick Marks	Reduce tick density, 5–6 ticks maximum on a single axis
Scale: Reversed	Use sparingly, only when low values are desirable (e.g., a chart of error rates)

Number formatting on axes: Right-click the axis > Format > Number. For revenue: Currency (0 decimal places, thousands separator). For percentages: Percentage (1 decimal place). For large numbers: Custom with “K” or “M” suffix: $#,##0,"K".

Removing Axis Clutter

Every element on a chart should earn its place. Default Tableau charts have several elements that can be removed to improve the signal-to-noise ratio:

Grid lines: Format > Lines > Grid Lines. Change grid lines from solid to a very light grey (10–15% opacity) or remove them entirely for clean, minimal charts. Keep grid lines only if the reader needs to read precise values from the chart.

Zero line: Format > Lines > Zero Line. The zero line should remain visible on any chart where zero is a meaningful reference point (bar charts, profit charts). Remove it from charts where zero is not in the data range.

Row/column banding: Format > Shading > Row/Column Banding. The light alternating grey bands in tables can be removed for a cleaner look or kept to improve row tracking in wide text tables.

Field labels: The “Sub-Category” and “SUM(Sales)” labels above axes are redundant when the axis title is clear. Right-click them and select Hide Field Labels for Rows/Columns to remove them.

18.6 Colour Formatting

Colour Best Practices in Tableau

Colour is the most powerful visual encoding in data visualization, and the most frequently misused. Apply these principles when choosing colours in Tableau:

Use colour purposefully: Colour should encode information, not decorate. Do not colour bars by a dimension just because it looks nice, only colour by a variable that carries analytical meaning.

Sequential palettes for continuous measures: Use when colour encodes a measure (e.g., revenue intensity on a map). Tableau’s Blue and Blue-Teal sequential palettes are the safest default choices, they are perceptually uniform and print well in greyscale.

Diverging palettes for measures with a meaningful midpoint: Use for profit margin (diverging around zero), survey scores (diverging around neutral), or any measure where both the direction and magnitude of deviation from centre matter. Tableau’s Orange-Blue diverging palette is excellent, red/orange for negative, blue for positive, white at zero.

Qualitative palettes for categorical dimensions: Use for distinguishing unordered categories (Region, Segment, Category). Tableau’s Tableau 10 and Tableau 20 palettes are carefully designed for perceptual distinctiveness. Limit to 6–8 colours maximum, more than 8 categories makes colour unusable as an encoding.

Editing palettes: Double-click the Colour legend to open the Edit Colours dialog. Assign specific hex colours to specific values (e.g., brand colours for company segments).

Colour Accessibility: Designing for Colour Blindness

Approximately 8% of men have colour vision deficiency (primarily red-green). Design charts that remain readable without relying on red-green distinction:

Use Tableau’s Color Blind palette (accessible from the Edit Colours dialog) for categorical data.
For sequential palettes, use Blue, Orange, or Viridis, not Red-Green.
Supplement colour with a second encoding when critical (e.g., both colour and shape on a scatter plot, both colour and a label on a map).
Test your workbook using the Simulate Colour Blindness option: View > Simulate Colour Blindness > select a type.

[Insert screenshot of the Edit Colours dialog showing the Color Blind palette selected and category values mapped to accessible colours]

18.7 Tooltip Formatting

Building Informative Tooltips

Tooltips appear when a user hovers over a mark. A well-designed tooltip provides the precise data values that the chart cannot show at the mark level, it is the chart’s on-demand detail panel. A poorly designed tooltip is a wall of cryptic field names and raw numbers.

Editing a tooltip: 1. In the Marks card, click the Tooltip tile. 2. The tooltip editor opens. Remove all auto-generated field names that are redundant (e.g., Category: <Category>). 3. Write natural-language tooltip text: “In <ATTR([Region])>, <SUM([Sub-Category])> generated <SUM([Sales])> in sales with a profit margin of <AGG([Profit Margin])>.” 4. Apply bold to key values, reduce font size for secondary information. 5. Check the Show tooltips checkbox and test by hovering over marks in the view.

Viz in Tooltip: For dashboards requiring drill-down, replace text tooltips with a mini visualisation. Click Insert > Sheets in the tooltip editor and select another worksheet. The embedded sheet renders a contextual chart, for example, hovering over a Region bar shows a product category breakdown within that region.

[Insert screenshot of the Tooltip editor showing a formatted natural-language tooltip with bold measure values and a clean layout, no redundant field name labels]

18.8 Reference Lines, Bands, and Distributions

Adding Analytical Context with Reference Lines

Reference lines add statistical or business-context benchmarks to a chart without adding new data. They answer the implicit question every chart raises: “Is this good or bad relative to what?”

Adding a reference line: 1. Go to Analysis > Add Reference Line. 2. In the Add Reference Line dialog, choose the scope: Per Cell (one line per panel), Per Pane, or Entire Table (one line across all panels). 3. Set the value: a Constant (e.g., a quarterly target: 50,000), the Average of the measure, the Median, or a Custom calculation. 4. Set the label: choose Value, Custom (type a descriptive label like “Q2 Target”), or None. 5. Style the line: a dashed grey line at 80% opacity is usually optimal, visible but subordinate to the data.

Reference bands: Use bands to highlight a range (e.g., the acceptable range of profit margins: 15–25%). In the Add Reference Line dialog, switch from Line to Band and set the lower and upper bounds.

Distribution lines: Add lines at standard deviations or percentiles from the mean, useful for identifying statistical outliers on scatter plots or histograms.

[Insert screenshot of a bar chart with a dashed reference line at the average value, labelled “Category Average”, and two bars highlighted above and below the average]

18.9 Summary

Key Concepts at a Glance

Topic	Key Technique	Access Point
Chart type selection	Match chart to analytical question and data type	Show Me panel or manual shelf placement
Sorted bar chart	Sort dimension axis descending by measure	Sort icon on axis header
Stacked vs. grouped bars	Colour shelf (stacked) vs. nested dimension (grouped)	Marks card > Colour / Rows shelf
Dual-axis line chart	Drag second measure to existing axis	Drop on axis until green “+” appears
Axis formatting	Title, range, include zero, number format	Double-click axis
Grid line reduction	Remove or lighten for cleaner charts	Format > Lines > Grid Lines
Sequential colour palette	Continuous measure colour encoding	Marks card > Colour > Edit Colours
Diverging colour palette	Measures with meaningful midpoint (zero)	Edit Colours > diverging palette
Tooltip formatting	Natural language, bold key values	Marks card > Tooltip
Reference lines	Target, average, percentile benchmarks	Analysis > Add Reference Line

Applying This in Practice

The most impactful formatting habit in Tableau is to complete the content first, then format in a single dedicated pass at the end. Building the chart and formatting simultaneously leads to over-formatted charts, adding colour and annotations before you know which marks actually need emphasis. Build the chart to answer the question, verify it is analytically correct, then ask: “What is the single most important insight this chart should communicate?” Format to amplify that insight and reduce everything else. This habit produces charts that are both analytically correct and visually clear, the combination that builds trust with executive audiences.