How to Create a Cluster in Tableau

Trying to make sense of a large dataset can feel like staring at a massive, shapeless cloud of dots. Clustering in Tableau helps you find patterns in that cloud, automatically grouping similar data points together to reveal hidden segments and insights. This guide will walk you through exactly how to create, interpret, and use clusters to better understand your customers, products, and performance.

Graphed

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.

Build Your First Agent

What is Clustering (and Why Should You Care)?

Cluster analysis is a technique used to group objects into clusters where objects in the same group are more similar to each other than to those in other groups. Think of it like a streaming service recommending new shows. It doesn't look at you in isolation, it puts you in a "cluster" of viewers who like similar types of movies and shows. This allows them to make much smarter, more personalized suggestions.

In a business context, clustering helps you answer questions like:

• Which customers behave alike? You can segment customers based on their purchase frequency, monetary value, and recent activity (RFM analysis) to identify your "VIPs," "At-Risk Customers," and "Newbies."
• Which products are often bought together? Uncover product bundles or marketing opportunities.
• Are there distinct performance tiers for your sales reps? Group reps based on metrics like deals closed, average deal size, and sales cycle length.

The real benefit is moving beyond simple averages and totals. Instead of looking at your "average customer," you can create data-driven personas for several types of customers and tailor your strategies for each. Tableau’s built-in clustering feature lets you do this visually, without needing to write code or understand complex statistical algorithms.

Getting Your Data Ready for Clustering

Before jumping in, a little preparation goes a long way. The quality of your clusters is directly tied to the data you feed the model. Fortunately, Tableau makes this part easy.

The main thing to know is that clustering in Tableau works with measures (quantitative, numerical data). You'll be grouping a dimension (like Customer Name, Product ID, or Region) based on the values of the measures you select (like Sales, Profit, Quantity, or Shipping Cost).

Here are a few quick tips for choosing your data:

• Choose relevant variables. The measures you include should be directly related to the question you are trying to answer. If you're creating customer segments based on purchasing behavior, variables like Total Sales, Number of Orders, and Average Order Value are great choices. Customer ID on its own would be meaningless.
• Don't worry about scaling. Sometimes, variables are on wildly different scales (e.g., Sales in tens of thousands of dollars and Quantity in single digits). In many statistical tools, this requires manual normalization. Tableau handles this for you automatically in the background, which is a massive time-saver.
• Start with a clear goal. Know what you're trying to segment. Focusing your analysis — for example, "I want to segment my customers based on profitability" — will help you choose the right measures and produce more meaningful results.

AI Agents for Marketing

Crash Course

Graphed

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.

Download the PDF

How to Create Clusters in Tableau: A Step-by-Step Guide

Let's walk through building clusters using a classic customer segmentation example. We'll use a scatter plot to visualize individual customers based on their total sales and profit.

Step 1: Build Your Initial Visualization

A scatter plot is the perfect starting point for visualizing relationships between two measures. Every mark on the chart will represent one member of the dimension we want to group (in this case, one customer).

1. Drag a measure, like Sales, to the Columns shelf.
2. Drag another measure, like Profit, to the Rows shelf.
3. Now, pull the dimension you want to cluster, such as Customer Name, onto the Detail level in the Marks Card.

You should now see a scatter plot where each point represents a unique customer, positioned based on their total sales and profit.

Step 2: Open the Analytics Pane

Next, you will be moving from the Data pane over to its neighbor, the Analytics pane. This is where Tableau stores its more advanced modeling features.

Step 3: Drag 'Cluster' Onto the View

In the Analytics pane, under the Model section, you'll see an option for Cluster. Simply click and drag it from the pane onto your visualization. When you drag it over, you’ll see a drop target appear. Release the mouse button on that target.

Step 4: Configure the Clusters

Tableau will immediately run its clustering algorithm and a "Clusters" dialog box will pop up. This is where you tell Tableau how to form the groups.

Initially, Tableau will automatically include the measures you have in your view (in our case, SUM(Profit) and SUM(Sales)) as the basis for the clusters. You can drag additional measures from the Data pane into this box if you want to include more variables in the analysis.

You’ll also need to specify the Number of Clusters. You can either type in a number or have Tableau automatically determine the best number of groups. For your first attempt, it's often best to leave this blank and see what Tableau suggests. You can always change it later.

Graphed

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.

Build Your First Agent

Step 5: See the Results!

Once you close the dialog box, Tableau does two things instantly:

1. It color-codes the marks on your scatter plot, with each color representing a different cluster.
2. It creates a new field in your Data pane called "Clusters." This is a dimension group, which means you can now use it in your visualizations just like any other field!

Congratulations, you’ve just created your first set of clusters in Tableau.

Interpreting and Using Your New Clusters

Creating the clusters is the easy part. The real value comes from understanding what each cluster represents and using that information to make decisions.

Understand the Profile of Each Cluster

Tableau makes it simple to see what defines each cluster. You can right-click the "Clusters" field on the Marks Card (or in the Data pane) and select Describe Clusters. This opens a window that summarizes the characteristics of each cluster.

The "Summary" tab gives you the number of items (customers) in each cluster and the average values for the variables used to create them. The "Models" tab provides more detailed statistical information (like the F-statistic and p-value) for those who want to dig deeper.

Looking at our example's summary, we can immediately start to form an idea about these groups. For a four-cluster model, you might find something like this:

• Cluster 1: Very high sales and very high profit. (High-Value Champions)
• Cluster 2: Low sales and negative profit. (Costly & Problematic)
• Cluster 3: Moderate sales but high profit. (Highly Profitable & Loyal)
• Cluster 4: High sales but low profit. (High-Volume Discount Seekers)

Give Your Clusters Meaningful Names

Don't stick with the default names like "Cluster 1" and "Cluster 2." They are abstract and hard to remember. Right-click on your "Clusters" pill in the Data pane and select Edit Group. Here, you can rename each cluster to a more descriptive name, like the ones we came up with above.

Use Your Clusters in Further Analysis

Remember, the cluster group you created is now a full-fledged dimension in your dataset. This means you can use it to slice and dice your other data. You can:

• Create a bar chart showing which product categories are most popular within your "High-Value Champions" cluster.
• Build a map to see if your "Costly & Problematic" cluster is concentrated in certain geographic regions.
• Track how the size of these clusters changes over time by adding a date filter to your analysis.

Refining Your Approach

Your first pass at clustering might not be your last. Finding the most useful segments often requires a bit of experimentation.

AI Agents for Marketing

Crash Course

Graphed

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.

Download the PDF

Experiment With the Number of Clusters

Right-click the "Clusters" field in the Data pane and select Edit Clusters. Try inputting a different number to see how the groups change. Too few clusters may lump distinct groups together, while too many may create tiny, insignificant segments that are hard to act upon. You're looking for a balance that delivers clear, differentiated, and actionable segments.

Try Different Variables

The variables you choose will fundamentally change the outcome. What happens if you add Quantity to our Sales and Profit cluster analysis? Maybe a new group emerges: customers who buy a ton of stuff but at very low margins. Each new variable adds another dimension to the analysis, providing a new lens through which to segment your data.

Don't be afraid to try different combinations to see what reveals the most interesting stories hidden in your data.

Final Thoughts

Clustering transforms messy raw data into organized, actionable segments. With Tableau's user-friendly drag-and-drop interface, this powerful statistical technique is accessible to anyone, empowering you to uncover deep insights about your customers, products, and overall business performance without requiring a degree in data science.

The ultimate goal of any analysis is to get clear answers that help you make better decisions faster. For marketing and sales teams, constantly jumping between platforms like Google Analytics, Shopify, and Salesforce to manually create segments can be slow and frustrating. We built Graphed to solve this by ditching the complex setup in favor of natural language. After connecting your sources in a few clicks, you can simply ask for the segments you need - like, "show me a breakdown of new vs. returning Shopify customers by traffic source" or "create a pipeline report from Salesforce showing conversion rates by sales rep" - and get a live, automated dashboard in seconds.