Creating Scatter Plots in Tableau

Course: Tableau Training

Topic: Creating Scatter Plots

Difficulty Level: Beginner to Intermediate

A scatter plot is a data visualization that uses individual points to represent the relationship between two numerical variables. In Tableau, scatter plots help analysts identify relationships, trends, clusters, unusual observations, and possible correlations between measures such as Sales and Profit, Advertising Cost and Revenue, or Product Price and Demand.

Course Link: JustAcademy Tableau Training

Demo Link: Register for a Tableau Course Demo

1. Learning Objectives

After completing this topic, you will be able to:

• Understand the meaning and purpose of scatter plots.
• Identify suitable data for a scatter plot.
• Create a basic scatter plot in Tableau.
• Use the Columns and Rows shelves to position numerical measures.
• Use the Marks card to customize color, size, shape, detail, and tooltips.
• Display individual data points using dimensions.
• Add and interpret trend lines.
• Identify positive, negative, and weak relationships.
• Find clusters and potential outliers.
• Create an interactive scatter plot for business analysis.

2. What Is a Scatter Plot?

A scatter plot, also called a scatter chart or scatter graph, is a chart that displays individual data points using two numerical variables. Each point has a horizontal position and a vertical position determined by the values of the two variables.

In Tableau, you normally place one numerical measure on the Columns shelf and another numerical measure on the Rows shelf. Tableau uses these measures to create the horizontal and vertical axes.

Example

Suppose a retail company wants to examine the relationship between Sales and Profit for its products.

In a scatter plot:

• The horizontal axis (X-axis) represents Sales.
• The vertical axis (Y-axis) represents Profit.
• Each point represents a product or another selected level of detail.
• The position of a point helps the analyst compare sales and profit values.

For example, a point far to the right represents a product or group with relatively high sales. A point near the bottom may represent a product or group with low or negative profit.

3. Why Are Scatter Plots Important?

Scatter plots are useful because they reveal patterns that may be difficult to notice in a table of numbers.

4. When Should You Use a Scatter Plot?

Use a scatter plot when you want to explore the relationship between two numerical measures or examine how individual observations are distributed.

Common Business Use Cases

• Sales analysis: Compare Sales with Profit for products, customers, or regions.
• Marketing analysis: Compare Advertising Spend with Revenue.
• Customer analysis: Compare Customer Age with Total Spending, where appropriate for the analysis.
• Financial analysis: Compare Risk and Return measures.
• Operations analysis: Compare Production Cost with Production Output.
• Retail analysis: Compare Product Price with Units Sold.
• Quality analysis: Compare production speed with defect rates.

When a Scatter Plot May Not Be Suitable

• When the main goal is to compare totals across categories, a bar chart may be more appropriate.
• When the goal is to show changes over time, a line chart may be clearer.
• When both variables are categorical, a different chart type may be needed.
• When too many points overlap, consider filtering, changing mark size, adding transparency where supported, or using density-oriented views.

5. Understanding the Main Parts of a Scatter Plot

6. Understanding Dimensions and Measures

Before creating a scatter plot, it is important to understand dimensions and measures in Tableau.

Dimensions

Dimensions are fields used to categorize, group, or describe data. Examples include Product Name, Category, Region, Customer Name, and Order ID.

Measures

Measures are fields that can be analyzed quantitatively. Examples include Sales, Profit, Quantity, Discount, and Cost.

How They Work in a Scatter Plot

• Measures define the numerical axes.
• Dimensions can divide the data into individual marks or groups.
• Dimensions placed on Detail can increase the number of points in the view.
• Dimensions placed on Color can assign different colors to categories.

Important: Tableau aggregates measures by default in many views. If you place SUM(Sales) on Columns and SUM(Profit) on Rows without adding a dimension to define individual marks, Tableau may display a single aggregated point instead of many points.

7. Requirements for Creating a Scatter Plot

A basic scatter plot generally requires two numerical measures.

Example dataset: A retail sales dataset containing the following fields:

You can practice using Tableau's Sample - Superstore dataset, which is commonly used in Tableau learning exercises.

8. How to Create a Basic Scatter Plot in Tableau

Follow these steps to create a scatter plot using Sales and Profit.

Step 1: Open Tableau

1. Launch Tableau Desktop or open an appropriate web-authoring environment.
2. Connect to your sales dataset.
3. Open a new worksheet.

Step 2: Add Sales to Columns

1. Locate the Sales field in the Data pane.
2. Drag Sales to the Columns shelf.
3. Tableau typically displays the measure as SUM(Sales) and creates the horizontal axis.

Step 3: Add Profit to Rows

1. Locate the Profit field in the Data pane.
2. Drag Profit to the Rows shelf.
3. Tableau typically displays SUM(Profit) and creates the vertical axis.

At this stage, Tableau may display one mark because Sales and Profit are aggregated across the data.

Step 4: Add a Dimension to Detail

1. Drag Product Name, Customer Name, Region, or another suitable dimension to Detail on the Marks card.
2. Observe how the number of marks changes.
3. Select the dimension that matches the level of analysis you want.

For example, adding Product Name to Detail can create a separate mark for each product, subject to the structure and aggregation of the data.

Step 5: Add Category to Color

1. Drag Category to Color on the Marks card.
2. Tableau assigns colors to the different category values.
3. Review the legend to understand the color assignments.

Step 6: Review the Chart

• Check that the X-axis represents Sales.
• Check that the Y-axis represents Profit.
• Confirm that the marks represent the intended products or groups.
• Review the axes, colors, title, and tooltips.

For the official walkthrough, see Tableau: Build a Scatter Plot.

9. Creating a Scatter Plot Using Show Me

Tableau's Show Me panel can help you choose a suitable chart type based on the fields you select.

Steps

1. Open a worksheet containing your data.
2. Select two numerical measures, such as Sales and Profit, from the Data pane.
3. Select an appropriate dimension if you want to distinguish individual records or groups.
4. Open the Show Me panel.
5. Select the scatter plot option if it is available for the selected fields.
6. Review the generated view and adjust the marks and dimensions as needed.

Note: Show Me may not produce the exact level of detail you need automatically. You may still need to add dimensions to Detail or change the aggregation of measures.

10. Understanding Aggregation and the Number of Marks

Aggregation is one of the most important concepts when creating scatter plots in Tableau.

Tableau commonly aggregates measures such as Sales and Profit using SUM. If the view contains only SUM(Sales) and SUM(Profit), Tableau may display one point representing the totals for the entire filtered dataset.

Example

Suppose a dataset contains 500 orders. A scatter plot with SUM(Sales) and SUM(Profit) alone may show one aggregated mark. To examine the relationship at order level, you need to define the marks using a suitable field such as Order ID, depending on the data structure.

How to Display More Marks

• Drag a dimension such as Product Name to Detail.
• Use Order ID on Detail when analyzing orders.
• Use Customer Name on Detail when comparing customers.
• Use Category on Color to distinguish categories.
• Use the Analysis menu to review whether aggregate measures are enabled, if your analysis requires a different level of detail.

Important: The dimension you choose determines what each point represents. Always confirm whether a point represents an order, product, customer, region, or aggregated group.

11. Using the Marks Card in Scatter Plots

The Marks card lets you customize the appearance and level of detail of the marks in a scatter plot.

Choosing the Correct Mark Type

Tableau may select a mark type automatically. For a scatter plot, circles are commonly used because they are easy to compare visually. You can change the mark type using the Marks card when another style is more suitable.

12. Formatting Colors in a Scatter Plot

Colors help distinguish groups or encode a meaningful measure.

Example: Color by Category

1. Drag Category to Color on the Marks card.
2. Open the Color options.
3. Choose a suitable palette or adjust individual category colors where supported.
4. Check that the legend is readable and that colors are easy to distinguish.

Example: Color by Profit

Drag Profit to Color to encode the profit value using a color scale. This can help reveal profitable and unprofitable observations.

Color Best Practices

• Use categorical colors for separate groups such as regions or categories.
• Use a sequential palette for an ordered numeric range.
• Use a diverging palette when values above and below a meaningful midpoint need to be distinguished.
• Keep colors consistent across related charts.
• Avoid using so many colors that the chart becomes difficult to read.
• Use labels, shapes, or other cues when color alone is insufficient.

13. Using Size and Shape

Using Size

Size can encode a third numerical variable in a scatter plot. For example, Sales may be shown on the X-axis, Profit on the Y-axis, and Quantity through mark size.

1. Locate the measure you want to represent, such as Quantity.
2. Drag the measure to Size on the Marks card.
3. Adjust the size range if needed.
4. Review the view to ensure that small marks remain visible and large marks do not obscure other points.

Using Shape

Shape can encode a category using different symbols.

1. Choose a scatter plot mark type that supports shapes.
2. Drag a dimension such as Segment to Shape.
3. Review the symbols and legend.
4. Use a small number of distinct shapes for easier interpretation.

Example: Three-Variable Scatter Plot

• X-axis: Sales.
• Y-axis: Profit.
• Color: Category.
• Size: Quantity.
• Detail: Product Name or another relevant identifier.

This configuration can display several aspects of a dataset in one view. However, avoid adding encodings that make the chart too complicated to understand.

14. Adding Labels and Tooltips

Adding Labels

Labels identify individual marks or show values directly on the chart.

1. Click Label on the Marks card.
2. Enable mark labels when appropriate.
3. Drag a field such as Product Name to Label if required.
4. Review the chart for overlapping text.

Labeling every point may create clutter when the dataset contains many observations. Use labels selectively or rely on tooltips for additional information.

Customizing Tooltips

Tooltips display details when the user hovers over a point.

1. Click Tooltip on the Marks card.
2. Edit the displayed text and fields.
3. Include information that helps explain the point.
4. Format numbers and dates consistently.
5. Hover over several marks to test the result.

Example Tooltip

Product: Product A
Category: Technology
Region: West
Sales: ₹25,000
Profit: ₹4,500
Quantity: 12

The figures above are illustrative examples only.

15. Adding Trend Lines to a Scatter Plot

A trend line is a statistical model displayed as a line that summarizes a relationship between numerical variables. It can help viewers identify the general direction of a relationship.

Steps to Add a Trend Line

1. Create a scatter plot with numerical fields on both axes.
2. Open the Analytics pane.
3. Locate Trend Line.
4. Drag Trend Line into the view.
5. Drop it on a suitable model type, such as Linear.
6. Review the trend line and its relationship to the marks.
7. Hover over the line to inspect the model information available in the tooltip.

Tableau supports several trend line model types, including Linear, Logarithmic, Exponential, Polynomial, and Power. The available options and suitability depend on the data and model assumptions.

Official guide: Tableau: Add Trend Lines to a Visualization.

Understanding Trend Line Types

Important: Do not select a complicated model simply because it fits the visible points. Choose a model that makes sense for the data and the question you are investigating.

16. Understanding Correlation in Scatter Plots

Correlation describes the strength and direction of an association between two variables. A scatter plot can provide a visual indication of a relationship, but visual inspection alone does not establish causation.

Positive Relationship

A positive relationship occurs when larger values of one variable tend to be associated with larger values of the other variable. The points often show an upward pattern from left to right.

Example: Advertising spending and sales revenue may rise together in some business settings.

Negative Relationship

A negative relationship occurs when larger values of one variable tend to be associated with smaller values of the other variable. The points often show a downward pattern from left to right.

Example: Product price and units sold may show a negative association in some circumstances.

Weak or No Clear Relationship

If points are widely dispersed without a clear direction, the chart may show a weak or unclear relationship between the variables.

Correlation Does Not Prove Causation

Even if two variables move together, the scatter plot alone does not prove that one variable causes the other to change. Other variables, sampling effects, or the way the data was collected may explain the pattern.

17. Understanding Outliers

An outlier is a point that lies unusually far from the overall distribution or pattern of the other points. Outliers can represent genuine unusual cases, data errors, or special circumstances that require investigation.

Example

Imagine most products have moderate sales and positive profit, but one product has very high sales and a large loss. That point may be worth investigating because of possible discounts, high costs, returns, or pricing issues.

How to Investigate Outliers in Tableau

1. Create a scatter plot with the relevant numerical measures.
2. Add a suitable dimension to Detail so individual observations can be identified.
3. Use Color or Size to highlight meaningful categories or values.
4. Hover over unusual points to review their tooltips.
5. Filter by region, category, or another relevant dimension to investigate the observation.
6. Verify unusual results against the source data before drawing conclusions.

Do not automatically remove outliers. First determine whether the points are valid observations and whether they are important to the business question.

18. Identifying Clusters in Scatter Plots

A cluster is a group of points that appear close together or share similar characteristics. Clusters can suggest that certain products, customers, or regions behave similarly.

Example: Customer Segmentation

A company can compare customer spending on the X-axis with order frequency on the Y-axis.

• High spending and high order frequency may indicate valuable repeat customers.
• Low spending and low order frequency may indicate occasional customers.
• High spending and low order frequency may indicate customers who make large, infrequent purchases.
• Low spending and high order frequency may indicate customers who place frequent, small orders.

These are illustrative interpretations. Actual customer segments should be defined using the organization's data and business context.

Finding Clusters

• Start with a scatter plot that clearly displays the observations.
• Use color to distinguish known groups.
• Investigate whether points form natural groupings.
• Where appropriate, explore Tableau's clustering capabilities from the Analytics pane.
• Validate the groups before using them for business decisions.

19. Creating a Scatter Plot Matrix

A scatter plot matrix displays relationships among several numerical variables in a set of small scatter plots. It helps analysts compare multiple pairs of measures.

Example

A retail dataset may contain Sales, Profit, Quantity, and Discount. A scatter plot matrix can help examine relationships between pairs of these variables.

How to Explore a Matrix

1. Select multiple numerical measures and relevant dimensions in the Data pane.
2. Open the Show Me panel.
3. Select the scatter plot matrix option if available for the selected fields and Tableau version.
4. Review each small chart and identify patterns or unusual observations.
5. Use individual scatter plots for closer examination of important relationships.

A matrix can be useful for exploratory analysis, but it can become difficult to read when too many variables are included.

20. Filtering and Interacting with Scatter Plots

Filters allow users to focus on a subset of data. They are especially useful when a scatter plot contains many marks.

Adding a Filter

1. Drag a field such as Region, Category, or Order Date to the Filters shelf.
2. Choose the values or range to include.
3. Show the filter control if viewers need to interact with it.
4. Test the filter to confirm that the chart updates as expected.

Useful Filters

• Region filter to compare geographical areas.
• Category filter to examine specific product groups.
• Date filter to focus on a reporting period.
• Sales or Profit filter to investigate selected ranges.

Dashboard Interactions

When a scatter plot is placed on a dashboard, you can configure it to interact with other charts. For example, selecting a point or group can filter a sales summary or a product detail table, depending on the dashboard actions configured.

21. Creating a Scatter Plot with Quadrants

A quadrant scatter plot divides the view into four sections using reference lines. This can help compare observations against two benchmarks, such as average Sales and average Profit.

Example

Suppose the X-axis shows Sales and the Y-axis shows Profit. Add a vertical reference line for a selected Sales benchmark and a horizontal reference line for a selected Profit benchmark.

Steps

1. Create a Sales-versus-Profit scatter plot.
2. Open the Analytics pane.
3. Drag a Reference Line into the view and configure the vertical benchmark.
4. Add a second reference line for the horizontal benchmark.
5. Choose appropriate values, such as an average or business target.
6. Label the reference lines clearly and review the four sections.

Note: The meaning of each quadrant depends on the benchmarks you choose. A value above an average is not automatically good or bad without business context.

22. Practical Example: Sales vs Profit Analysis

Scenario: A retail manager wants to determine which products generate high sales and which products generate strong profits.

Dataset Fields

• Product Name
• Category
• Region
• Sales
• Profit
• Quantity

Build the Visualization

1. Connect Tableau to the retail dataset.
2. Open a new worksheet.
3. Drag Sales to Columns.
4. Drag Profit to Rows.
5. Drag Product Name to Detail to identify products, if that is the desired level of analysis.
6. Drag Category to Color.
7. Optionally drag Quantity to Size.
8. Format the axes and add a descriptive title.
9. Customize the tooltip to show Product Name, Category, Sales, Profit, and Quantity.
10. Add trend lines or reference lines if they help answer the analytical question.

Possible Questions to Investigate

• Which products have both high sales and high profit?
• Which products have high sales but low or negative profit?
• Do products in different categories show different patterns?
• Are there unusual products that require further investigation?
• Does the relationship appear positive, negative, or unclear?

Business Interpretation

A scatter plot can help identify products that deserve further analysis. For example, a product with high sales but negative profit may require an investigation into discounts, costs, shipping expenses, or returns. The chart identifies a pattern to examine; it does not explain the cause by itself.

23. Practical Project: Customer Spending Analysis

Project Title: Customer Spending and Order Frequency Analysis

Objective: Use a scatter plot to explore customer purchasing behavior.

Required Fields

Project Steps

1. Connect to a customer orders dataset.
2. Prepare or calculate the required customer-level measures.
3. Place Total Customer Sales on Columns.
4. Place Order Count on Rows.
5. Place Customer ID on Detail.
6. Place Customer Segment on Color if the field exists and is meaningful.
7. Add a useful tooltip with customer-level values.
8. Format axes, labels, and colors.
9. Explore unusual customers and clusters.
10. Place the scatter plot on a dashboard alongside supporting summary views.

Expected Outcome

The scatter plot helps identify customers with high or low spending and order frequency. These patterns can support further analysis of customer engagement and purchasing behavior.

24. Common Problems and Solutions

25. Best Practices for Scatter Plots

• Choose two numerical measures that relate to the question you want to investigate.
• Use the Columns shelf for the horizontal measure and the Rows shelf for the vertical measure in a standard scatter plot.
• Choose the correct level of detail so each point represents the intended observation or group.
• Use clear axis titles and suitable number formats.
• Use color, size, and shape only when they add meaningful information.
• Use labels sparingly and provide useful tooltips.
• Investigate outliers rather than automatically removing them.
• Add trend lines only when they are useful for the analysis.
• Remember that correlation does not prove causation.
• Consider filtering or alternative views when marks overlap heavily.
• Keep the chart title descriptive and the overall design simple.
• Verify conclusions against the data and relevant business context.

26. Interview Questions and Answers

Q1. What is a scatter plot in Tableau?

Answer: A scatter plot displays individual marks based on two numerical measures. It helps explore relationships, patterns, clusters, and outliers.

Q2. Which shelves are used to create a basic scatter plot?

Answer: The Columns shelf usually contains the measure for the X-axis, while the Rows shelf usually contains the measure for the Y-axis.

Q3. Why does Tableau sometimes display only one mark?

Answer: Tableau may aggregate the measures across the entire dataset. Adding a suitable dimension to Detail can divide the data into more marks.

Q4. What is the purpose of Detail on the Marks card?

Answer: Detail adds information to the level of detail of the view. A dimension on Detail can create separate marks for distinct members or combinations of members.

Q5. How can you color a scatter plot by category?

Answer: Drag the Category dimension to Color on the Marks card.

Q6. How can a scatter plot represent a third measure?

Answer: A third measure can be encoded through mark size or color, depending on the data and the intended meaning.

Q7. What is a trend line?

Answer: A trend line is a statistical model displayed on a chart to summarize a relationship between numerical variables.

Q8. What does a positive relationship mean?

Answer: It means that larger values of one variable tend to occur with larger values of the other variable.

Q9. What does a negative relationship mean?

Answer: It means that larger values of one variable tend to occur with smaller values of the other variable.

Q10. What is an outlier?

Answer: An outlier is an observation that differs noticeably from the general distribution or pattern of the other observations.

Q11. Can a scatter plot prove causation?

Answer: No. A scatter plot can reveal an association, but additional evidence and analysis are required to establish causation.

Q12. What is a scatter plot matrix?

Answer: A scatter plot matrix displays multiple pairwise relationships among numerical variables in a collection of small scatter plots.

Q13. Why are tooltips useful in scatter plots?

Answer: Tooltips provide detailed information about individual points without requiring all values to be displayed as labels.

Q14. What is a quadrant scatter plot?

Answer: It is a scatter plot divided into four sections by reference lines, often representing benchmarks for the two variables.

Q15. When should you avoid a scatter plot?

Answer: Avoid it when the question is better answered by a different chart, such as a bar chart for category totals or a line chart for trends over time.

27. Practice Exercises

Complete these exercises using Tableau's Sample - Superstore data or another suitable dataset.

1. Create a scatter plot comparing Sales and Profit.
2. Add Product Name to Detail and observe how the number of marks changes.
3. Color the marks by Category.
4. Use Quantity on Size and examine the result.
5. Customize the tooltip to show Sales, Profit, Category, and Region.
6. Add a trend line and examine the relationship it represents.
7. Create a scatter plot comparing Sales and Discount.
8. Find products with high sales but low or negative profit.
9. Add reference lines to create a quadrant analysis.
10. Build a dashboard containing the scatter plot and a supporting sales summary.

Challenge Questions

• What happens when you remove the dimension from Detail?
• How does the view change when you put a dimension on Color instead of Detail?
• What is the effect of using SUM(Sales) versus a different aggregation?
• How can you make a scatter plot easier to read when many points overlap?
• Why should a trend line be interpreted cautiously?

28. Quick Revision Notes

• A scatter plot displays the relationship between two numerical measures.
• The Columns shelf usually defines the X-axis.
• The Rows shelf usually defines the Y-axis.
• Dimensions on Detail can create more individual marks.
• Color can distinguish categories or represent numerical values.
• Size can represent an additional quantitative measure.
• Labels display information directly on marks.
• Tooltips show extra details when users hover over marks.
• Trend lines summarize modeled relationships between numerical variables.
• Positive relationships tend to move upward; negative relationships tend to move downward.
• Outliers are observations that differ from the general pattern.
• Clusters are groups of observations with similar characteristics.
• Correlation does not necessarily imply causation.
• Good scatter plots use clear axes, meaningful encodings, and the right level of detail.

29. Final Checklist

• ☐ The chart uses two appropriate numerical measures.
• ☐ The X-axis and Y-axis are clearly labeled.
• ☐ The marks represent the correct level of detail.
• ☐ Color, size, and shape have clear meanings.
• ☐ The legend is readable and useful.
• ☐ Tooltips contain relevant details.
• ☐ Labels do not create unnecessary clutter.
• ☐ Outliers and clusters are investigated carefully.
• ☐ Any trend line is appropriate for the analysis.
• ☐ The chart does not imply causation without supporting evidence.
• ☐ The final visualization is clear, accurate, and easy to interpret.

30. Conclusion

Creating scatter plots is an important Tableau skill for exploring relationships between numerical variables. By placing measures on the Columns and Rows shelves and using the Marks card to control detail, color, size, shape, labels, and tooltips, you can build informative visualizations for business analysis.

Trend lines, reference lines, filters, and clustering can add further analytical value when used appropriately. The key is to select the right level of detail, keep the visualization readable, and interpret the results in the context of the data rather than assuming that every visible relationship represents causation.

Learn more: JustAcademy Tableau Training | Register for a Tableau Course Demo

Official Tableau References

• Build a Scatter Plot
• Add Trend Lines to a Visualization
• Spot Trends in Tableau
• Choose the Right Visualization