Graphed LogoGraphed
Graphed LogoGraphed
How It WorksAgentsIntegrationsPricing

How to Use Group By in Tableau

Cody Schneider••8 min read

If you have any experience with SQL, the GROUP BY clause is likely second nature. It’s the essential command for summarizing rows of data into aggregate groups. When moving to a visual tool like Tableau, a common question is, "Where is the GROUP BY button?" This article will show you that while there isn't a single button, the entire tool is built around this principle. We'll walk through the straightforward ways to group and summarize your data in Tableau to create powerful visualizations.

GraphedGraphed

Build AI Agents for Marketing

Build virtual employees that run your go to market. Connect your data sources, deploy autonomous agents, and grow your company.

Watch Graphed demo video

What Does "GROUP BY" Actually Mean in Tableau?

In the world of databases, a GROUP BY statement takes countless rows of data, finds all the rows with a common value in a specific column, and combines them into a single summary row. For example, you might group millions of transaction records by ProductCategory to find the total sales for electronics, furniture, and clothing.

Tableau accomplishes this visually. There’s no syntax to write because the act of building a visualization is the act of grouping data. Every time you drag a qualitative field (a blue "Dimension" pill) onto the canvas, you are defining the criteria you want to GROUP BY. When you drag a quantitative field (a green "Measure" pill), you're defining the aggregation (like SUM, AVG, COUNT) you want to see for those groups.

For instance, dragging the Region dimension to the Columns shelf and the Sales measure to the Rows shelf is the visual equivalent of writing:

SELECT Region, SUM(Sales) FROM YourTable GROUP BY Region,

Understanding this core concept is the key to mastering data aggregation in Tableau. The rest is about learning the different methods to define your groups.

Free PDF · the crash course

AI Agents for Marketing Crash Course

Learn how to deploy AI marketing agents across your go-to-market — the best tools, prompts, and workflows to turn your data into autonomous execution without writing code.

Method 1: The Standard Drag-and-Drop GROUP BY

This is the most fundamental and common way to group data in Tableau. It’s intuitive, fast, and forms the basis of nearly every chart you'll build. Let's walk through it with a practical example using Tableau’s Sample - Superstore dataset.

Example: Visualizing Sales by Product Category

Our goal is simple: We want to see the total sales for each main product category.

  • Step 1: Get Your Data Ready Start by connecting to your data source. For this exercise, we’ll use the Sample - Superstore dataset that comes packaged with Tableau.
  • Step 2: Define Your Groups Look in the Data pane on the left side of your screen. Dimensions are the fields you typically use for grouping. Find the Category dimension.
  • Step 3: Drag the Dimension into the View Click and drag the Category dimension and drop it onto the Columns shelf. You will now see three column headers appear in your primary view: 'Furniture', 'Office Supplies', and 'Technology'. You’ve just told Tableau what to group your data by.
  • Step 4: Define Your Aggregation Next, find the Sales measure in the Data pane. Measures are the numerical values you want to analyze.
  • Step 5: Drag the Measure into the View Click and drag the Sales measure and drop it onto the Rows shelf. Tableau instantly creates a bar chart. It has automatically summed up the sales for every record belonging to each category.

You’ve just performed a GROUP BY without writing a single line of code. The view clearly shows the aggregate SUM(Sales) for each unique value within the Category dimension.

Method 2: Create Manual Groups for Dimensions

Sometimes your data doesn't have the exact groupings you need. You might want to combine several existing dimension members into a single, custom group for analysis - without having to modify the underlying data source. This is perfect for classifying territories, combining product variations, or simplifying a busy visualization.

Example: Creating a Custom "West Coast" Sales Region

Let's say our data has State information, but we want to analyze the performance of 'California', 'Oregon', and 'Washington' as a single "West Coast" unit and compare it to other individual states.

  • Step 1: Select the Dimension In the Data pane, find the State dimension. Right-click on it.
  • Step 2: Create a Group From the context menu, navigate to Create > Group... This will open the Create Group dialog box.
  • Step 3: Form Your Group You'll see a list of all the states. Select 'California', 'Oregon', and 'Washington' by holding the Ctrl key (or Cmd on Mac) while clicking. With those three selected, click the Group button.
  • Step 4: Name and Finalize the Group Tableau will bundle these three states into a new line item in the list. You can click "Rename" to give it a more descriptive name, like "West Coast". To make analysis cleaner, you can also check the "Include 'Other'" box, which will lump all non-grouped states into a single catch-all "Other" category. Click OK.

A new field called State (group) now appears in your Data pane. You can drag this new dimension into your view, paired with a measure like SUM(Profit), to instantly visualize the performance of your new "West Coast" region against everything else.

GraphedGraphed

Build AI Agents for Marketing

Build virtual employees that run your go to market. Connect your data sources, deploy autonomous agents, and grow your company.

Watch Graphed demo video

Method 3: Grouping with Calculated Fields

For the most flexible and powerful grouping, you need to use calculated fields. This method allows you to group data based on your own custom logic and conditions. It's the equivalent of using a CASE WHEN statement in SQL to create custom categories on the fly.

Example: Grouping Orders by Size

Imagine we want to categorize every order as "Small," "Medium," or "Large" based on its sales value. This isn't a column that exists in our data, so we need to create it.

  • Step 1: Create a Calculated Field At the top of the Data pane, click the little dropdown arrow and select Create Calculated Field...
  • Step 2: Name Your Calculation and Write the Logic Name the field "Order Size." In the formula box, we’ll use an IF/THEN/ELSE statement to define our logic. This calculation will run on every single row in our dataset.
IF [Sales] <= 100 THEN "Small"  
ELSEIF [Sales] > 100 AND [Sales] <= 500 THEN "Medium"  
ELSE "Large"  
END

This formula checks the Sales value for each row. If it's 100 or less, it labels the row "Small." If it's between 100 and 500, it's "Medium." Otherwise, it's labeled "Large."

  • Step 3: Use Your New Grouping Dimension Click OK. You now have a brand new dimension called Order Size in your data pane. You can use this exactly like any other dimension. Drag Order Size to Columns and a measure like COUNTD(Order ID) (Count Distinct of Order ID) to Rows.

Immediately, you have a bar chart that totals up the number of unique orders that fall into each of your custom-defined size buckets. This is a powerful way to add business context and segmentation to your raw data.

Advanced Tips for Effective Grouping

Understanding Granularity and Level of Detail (LOD)

The grouping in your view is always determined by the dimensions you add to it - this is called the "Level of Detail." Sometimes, however, you need to calculate an aggregation at a different level of detail. For example, you might want to find the average sales per customer within each region. For this, you would use an LOD expression, which goes beyond a basic GROUP BY.

A FIXED LOD expression looks like this:

{ FIXED [Region], [Customer ID] : SUM([Sales]) }

This creates an aggregation at the customer and region level, which can then be used in other calculations, regardless of what's in your current view. It’s an advanced topic, but one to be aware of as your needs get more complex.

Free PDF · the crash course

AI Agents for Marketing Crash Course

Learn how to deploy AI marketing agents across your go-to-market — the best tools, prompts, and workflows to turn your data into autonomous execution without writing code.

Use Hierarchies for Drill-Down Grouping

If you have dimensions that are naturally related, like Country, State, and City, you can group them into a hierarchy. Just drag one dimension on top of the other in the Data pane. When you use a hierarchy in a view, a small + icon appears on the pill, allowing you to "drill down" to the next level, effectively changing your GROUP BY clause with a single click - from GROUP BY Country to GROUP BY Country, State.

Final Thoughts

While Tableau may not have a dedicated GROUP BY command written out, a deep understanding of grouping is essential to using it effectively. Whether you're doing a simple drag-and-drop aggregation, combining dimension members into manual groups, or writing complex conditional logic in a calculated field, you are using the core concepts of GROUP BY to summarize your data and uncover new insights.

Manually building these views in a tool like Tableau saves a lot of time compared to writing SQL queries or wrangling spreadsheets. Still, we found ourselves spending hours fine-tuning calculations and views to answer follow-up questions. At a certain point, a natural language interface is simply faster. For our own reporting, we connect all our data sources to Graphed and ask questions directly. Answering a simple business question like, "Show me the count of large orders per region this quarter," becomes a single prompt, not a multi-step process. It feels like the next logical evolution in making data accessible to everyone.

Related Articles