2  Functions of Data Visualization

NoteWhat This Chapter Covers

Data visualization serves distinct purposes depending on who is looking at it and what decision it needs to support. This chapter presents four core functions of data visualization in Tableau: exploration, communication, monitoring, and storytelling. For each function, you will learn what it is, when to use it, and which Tableau features directly support it. Understanding these functions before building any visualization ensures that your design choices serve the actual purpose of the analysis. This chapter draws on the practical frameworks described in Milligan (2022), Jones (2014), Wexler, Shaffer, and Cotgreave (2017), and Kriegel (2022).

flowchart LR
    A[Data <br> Source] --> B[Exploration <br> Find patterns]
    A --> C[Communication <br> Share findings]
    A --> D[Monitoring <br> Track performance]
    A --> E[Storytelling <br> Drive decisions]
    B --> F[Tableau <br> Worksheets & EDA]
    C --> G[Tableau <br> Formatted Views]
    D --> H[Tableau <br> Dashboards]
    E --> I[Tableau <br> Stories]
    style A fill:#e3f2fd,stroke:#1976D2
    style F fill:#e8f5e9,stroke:#388E3C
    style G fill:#e8f5e9,stroke:#388E3C
    style H fill:#e8f5e9,stroke:#388E3C
    style I fill:#e8f5e9,stroke:#388E3C

2.1 Function 1: Exploratory Visualization

NoteExploring Data to Find What You Do Not Yet Know

Exploratory visualization is the first and most open-ended function. Its purpose is to interrogate the data without a predetermined hypothesis, looking for patterns, outliers, relationships, and trends that might not have been anticipated. Milligan (2022) describes this as the natural starting point for any Tableau analysis: you connect to a data source and begin asking questions of the data by dragging fields onto shelves, changing chart types, and adding filters, all in rapid succession.

In exploratory work, speed and flexibility matter more than visual polish. The audience is the analyst alone. The goal is to discover something worth communicating, not to communicate it yet. Kriegel (2022) reinforces this point by showing how Tableau’s drag-and-drop interface is designed specifically to support fast iterative questioning, allowing you to change the entire view in seconds simply by rearranging fields.

Tableau features that support exploration:

  • Show Me panel: Suggests appropriate chart types based on the fields currently selected. Clicking through Show Me options is one of the fastest ways to see many views of the same data in a short time.
  • Analytics pane: Adds trend lines, reference lines, forecasts, and cluster analysis to any view with a single drag. These analytical overlays reveal patterns that are invisible in raw data.
  • Instant filter and highlight: Right-clicking any mark gives options to Keep Only or Exclude, enabling rapid subset investigation without building a formal filter.
  • Quick Table Calculations: Right-clicking a measure on a shelf gives access to running totals, percentage of total, year-over-year difference, and other derived metrics that surface patterns in time-series data.
NoteHow To: A Typical Exploratory Workflow in Tableau
  1. Connect to your data source and open a blank worksheet.
  2. Drag the primary dimension (for example, Sub-Category) to the Rows shelf.
  3. Drag the primary measure (for example, Sales) to the Columns shelf. Tableau creates a horizontal bar chart automatically.
  4. Click Show Me and cycle through chart types to see the same data from different angles.
  5. Drag a second dimension (for example, Region) to the Colour shelf to reveal disaggregation.
  6. Open the Analytics pane and drag a Trend Line onto the view to check for directional patterns.
  7. Right-click any outlier mark and select Annotate > Mark to note what you observe before moving on.
  8. Duplicate the worksheet (right-click the tab and select Duplicate) to preserve the view before changing it.

[Insert screenshot of a Tableau worksheet showing a bar chart with a trend line from the Analytics pane and a mark annotation on an outlier bar]

TipPreserve Every Exploratory Sheet

Milligan (2022) advises keeping all worksheets created during exploration, even the ones that appear unproductive. Worksheets in Tableau are lightweight and do not duplicate data. A view that looks like a dead end today may provide critical context when you investigate an anomaly in a later session. Use a naming convention such as EDA-01-Sales-by-Category to keep exploratory sheets organised and easy to find.

2.2 Function 2: Communicative Visualization

NoteSharing a Known Finding with an Audience

Communicative visualization begins where exploration ends. Once you have found a pattern or insight worth sharing, the visualization must be redesigned for a different audience: the stakeholder, manager, client, or decision-maker who was not part of the analysis process.

Jones (2014) describes this shift as fundamental to effective data communication. A visualization built for exploration is cluttered with analytical scaffolding that the analyst needed but the audience does not. Communicative design strips away that scaffolding and replaces it with clarity. The guiding principle from Jones (2014) is that every element of a communicative visualization should answer one question for the reader: “What do I need to understand from this chart?”

Key differences between exploratory and communicative visualization:

Dimension Exploratory Communicative
Audience The analyst Stakeholders and decision-makers
Purpose Discover patterns Confirm a specific finding
Chart title Variable name (for example, Sales) Declarative finding (for example, North Region Leads All Categories)
Annotations Analyst notes Explanations for the reader
Formatting Default Tableau formatting Cleaned, branded, minimal
Filters Many, for drilling down Few or none, pre-filtered to the relevant view
NoteHow To: Preparing a Communicative View in Tableau
  1. Open the completed exploratory worksheet and select Format > Workbook to set a consistent font for the entire workbook.
  2. Double-click the chart title and rewrite it as a declarative finding statement, not a variable name. For example, change “Sales by Region” to “North Region Outperforms in All Three Product Categories.”
  3. Right-click key data points and select Annotate > Mark to add brief explanatory text pointing to the most important finding.
  4. Remove unnecessary gridlines via Format > Lines > Grid Lines > None to reduce visual clutter.
  5. Remove default axis titles if the chart title already explains the axes.
  6. Adjust colours: use one primary colour for the main measure and reserve a contrasting colour only for the mark you want the audience to focus on.
  7. Add a data source caption via Worksheet > Show Caption so the audience can verify the data origin.

[Insert screenshot of a Tableau worksheet showing a cleaned communicative bar chart with a declarative title and one annotated mark highlighted in a contrasting colour]

WarningCommunicative Visualizations Must Be Honest

Jones (2014) cautions that the process of simplifying a visualization for communication introduces opportunities for distortion. A chart edited to “look cleaner” can inadvertently misrepresent the data if axes are truncated, scales are inconsistent, or context is removed. Before finalising any communicative visualization, verify that: the Y-axis starts at zero for bar charts, all axes share the same scale when comparing multiple charts, and any data exclusions are labelled or footnoted.

2.3 Function 3: Monitoring Visualization

NoteTracking Ongoing Performance Against Targets

Monitoring visualization is designed to be consulted repeatedly by the same audience. The purpose is not discovery or persuasion but operational awareness: is performance on track right now, and if not, where is the problem? Wexler, Shaffer, and Cotgreave (2017) describe monitoring dashboards as the most common type of business dashboard, used daily by managers across sales, operations, finance, and customer service.

Because monitoring dashboards are viewed repeatedly and quickly, their design requirements are different from both exploratory and communicative visualization. Wexler et al. (2017) identify three requirements that monitoring dashboards must meet:

  1. Immediate status recognition: The viewer must assess the overall status of the business in under five seconds. This requires prominent use of colour to encode performance against target, green for on-target, red for below-target, and a consistent layout that never changes so that viewers can locate information from memory.
  2. Benchmark comparison: Every metric in a monitoring dashboard must be shown alongside a target, prior period, or industry benchmark. A number without a comparator communicates nothing about whether performance is good or bad.
  3. Exception highlighting: The dashboard should direct attention to metrics that require action, not to metrics that are performing normally. Normal performance is unremarkable; deviations from normal are what the viewer needs to act on.

Tableau features that support monitoring:

  • KPI tiles: Single-number text marks with conditional colour formatting (green/red) signal target status instantly.
  • Bullet charts: Display a primary measure, a comparative measure, and qualitative performance ranges in a single compact chart, the most efficient chart type for KPI monitoring according to Wexler et al. (2017).
  • Sparklines: Small line charts placed alongside a current value show trend at a glance without requiring a separate chart.
  • Dashboard actions: Filter actions and highlight actions allow the viewer to click into a problem area for details without leaving the dashboard.
NoteHow To: Building a KPI Monitoring Tile in Tableau
  1. Create a new worksheet. Drag the target measure (for example, Sales) to the Text shelf in a Mark type of Text.
  2. Create a calculated field for the conditional colour signal:
Code
# Returns a colour label based on target comparison
IF SUM([Sales]) >= [Monthly Target]
THEN "On Target"
ELSE "Below Target"
END
  1. Drag the colour-label calculated field to the Colour shelf. Open Edit Colours and assign green to “On Target” and red to “Below Target.”
  2. Format the text to a large, bold font (18pt or larger) via Format > Font.
  3. Hide all axis lines, headers, and gridlines so only the number and its background colour are visible.
  4. Place this worksheet on a dashboard alongside the charts it summarises.

[Insert screenshot of a Tableau dashboard header row showing three KPI tiles, one green (on target), one red (below target), and one green, with large bold numbers and no visible axes]

TipThe Bullet Chart for Monitoring

Wexler et al. (2017) recommend bullet charts as the most space-efficient format for displaying a measure against its target in a monitoring context. A bullet chart shows the actual value as a bold bar, the target as a thin line, and qualitative performance zones (poor, acceptable, good) as grey shading behind the bar. Tableau supports bullet charts natively via the Show Me panel when you select one measure and one reference value.

2.4 Function 4: Storytelling Visualization

NoteGuiding an Audience from Data to Decision

Data storytelling is the highest-value function because it connects analytical findings directly to business decisions. Where a communicative visualization shares a single finding, a data story builds a structured case across multiple views, leading the audience from context through evidence to a recommended action.

Kriegel (2022) describes the Tableau Story feature as the dedicated tool for this function. A Story is a sequence of Story Points, individual slides containing a worksheet or dashboard, each with a short caption that advances the narrative. The viewer moves through the story point by point, following a logical progression from setup to conclusion.

Milligan (2022) describes the structure of an effective data story in three stages:

  1. Setup: Establish the business context. What is the current situation, and why does it matter to the audience? This stage typically uses a broad summary view showing overall performance.
  2. Finding: Introduce the specific insight, pattern, or anomaly that emerged from the analysis. This is the heart of the story, the point at which the audience shifts from passive reception to active engagement with the data.
  3. Recommendation: Connect the finding to an action. Data stories that end at the finding leave the audience uncertain about what to do. The recommendation stage specifies what action the data supports and what outcome that action is expected to produce.

This three-stage structure ensures that every story begins with why the audience should care and ends with what they should do, the two things decision-makers most need from a data presentation.

NoteHow To: Building a Tableau Story
  1. Build the individual worksheets and dashboards that correspond to each narrative stage.
  2. Click the New Story tab at the bottom of the workbook (the stacked-page icon).
  3. Drag a worksheet or dashboard from the Story panel on the left into the Story canvas.
  4. Click the caption box below the navigator and write a one-sentence declarative statement. For example, write “Revenue grew 8% year-over-year” rather than “Revenue chart.”
  5. Click New Blank Point to add the next story point and drag the next sheet onto the canvas.
  6. Repeat for each stage of the narrative.
  7. Navigate through all story points using the arrow keys to verify that the captions read as a coherent narrative without the charts.

[Insert screenshot of the Tableau Story workspace showing the story point navigator at the top with three captioned boxes, the canvas showing a dashboard, and the caption box highlighted below]

WarningCaptions That State the Finding, Not the Topic

Kriegel (2022) notes that the most common mistake in Tableau Stories is writing captions as topic labels rather than findings. A caption that reads “Sales by Region” tells the audience nothing. A caption that reads “West Region accounts for 42% of total revenue” gives them the conclusion before they look at the chart. Write every caption as a complete sentence stating the single most important thing to take away from that story point.

2.5 Matching Function to Tableau Feature

NoteSummary: Which Function Uses Which Tableau Feature
Visualization Function Primary Tableau Feature Supporting Features
Exploration Worksheets with drag-and-drop, Show Me Analytics pane, Quick Table Calculations
Communication Formatted worksheets, annotations Declarative titles, cleaned colour and layout
Monitoring Dashboards with KPI tiles and bullet charts Dashboard actions, sparklines, conditional colour
Storytelling Story points with declarative captions Annotated worksheets, dashboard drill-through

Choosing the correct function before opening Tableau is the single most important decision in any visualization project. The function determines the audience, the audience determines the design, and the design determines which Tableau features to use. All four functions are used across this book: exploration in the early analytical chapters, communication in the chart-type chapters, monitoring in the dashboard chapters, and storytelling in the Stories and case-study chapters.

2.6 Summary

NoteKey Concepts at a Glance
Function Core Question Primary Audience Key Tableau Feature
Exploration What is in this data? The analyst Worksheets, Show Me, Analytics pane
Communication How do I share this finding clearly? Stakeholders, decision-makers Annotations, declarative titles, clean formatting
Monitoring Is performance on track right now? Operational managers Dashboards, KPI tiles, bullet charts
Storytelling How do I move the audience to action? Executive and decision-making audiences Story points with declarative captions
TipApplying This in Practice

Before building any visualization in Tableau, write a single sentence completing this prompt: “The purpose of this visualization is to ___.” The answer will identify which of the four functions applies, and that function will tell you which Tableau features to use, how much detail to include, who your audience is, and how to measure whether the visualization has succeeded. This one habit will save more rework than any other single practice in Tableau.