How to Join Data Sources in Tableau

Your most powerful business insights often live in the gaps between your different data sources. Your website traffic is in Google Analytics, your sales are in Shopify, and your ad spend is in Facebook Ads. To find out which ad click led to which sale, you need to connect the dots. This article will show you exactly how to join, relate, and blend data sources in Tableau to get a complete view of your business 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

Why Combine Data Sources in the First Place?

Working with a single spreadsheet or database table is clean and simple, but it rarely tells the full story. The real magic happens when you start layering and combining different datasets to answer more complex questions. The goal is to move from looking at isolated metrics to understanding the entire customer journey.

Imagine you run an e-commerce store. You're trying to figure out if your expensive Facebook ad campaigns are actually profitable. To do this, you need data from three separate places:

• Facebook Ads: To see your ad spend, clicks, and impressions by campaign.
• Google Analytics: To see who visited your site, which pages they looked at, and where they came from (the campaign source).
• Shopify: To see who actually made a purchase, what they bought, and how much revenue was generated.

By connecting these three sources, you can finally calculate your true return on ad spend (ROAS) for each specific campaign. You can stop guessing which ads are working and start making data-backed decisions to invest in the winners and cut the losers. This is just one example, but the principle applies everywhere: combining lead data from Salesforce with marketing automation data from HubSpot, or combining support tickets from Zendesk with customer usage data from your app. To get a 360-degree view, you need to bring your data together.

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

Data Relationships vs. Joins vs. Blends: The Key Differences

Before you start dragging tables around in Tableau, it's crucial to understand the three distinct ways it combines data. Getting this right from the start will save you a lot of headaches later on.

1. Relationships (The Default "Noodle" Method)

If you've used a recent version of Tableau, you've seen relationships in action. When you drag two tables onto your Data Source canvas, a line (we call it a "noodle") appears between them. This is a relationship. Think of it as introducing two smart spreadsheets to each other. You tell them they have a column in common (like Order ID), and they agree to talk to each other when needed.

The key concept: Relationships are flexible and preserve the original level of detail for each table. The tables remain separate but related. Tableau automatically figures out the right way to query them in the background based on the fields you use in your visualization. This prevents many of the data duplication and aggregation issues that older methods caused.

When to use it: Most of the time. This is Tableau’s recommended and default method for combining tables from the same data source. It’s powerful and handles most use cases well.

2. Joins (The Classic "Physical Merge")

A join is more permanent. Instead of just introducing two tables, a join physically merges them into a single, new (and usually wider) table before you even start building a viz. You define the join type (Inner, Left, Right, Full Outer) to control exactly which rows are included in the final, combined table.

The key concept: You are forcing all the data into one big table from the get-go. This can be necessary for certain types of advanced analysis but requires you to be very careful, as it can easily create duplicated rows if the relationship between your tables isn't one-to-one.

When to use it: When you need more control over how your tables are merged into a single entity, or if you're using a version of Tableau before relationships were introduced. You can still access the join interface by working within the "physical layer" of a relationship.

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

3. Data Blending (For Completely Separate Sources)

Data blending is different from both relationships and joins. It's used when you need to combine data from entirely different data sources that can't be joined at the database level (e.g., an Excel file and a published Salesforce data source).

The key concept: Blending works on a worksheet-by-worksheet basis. You have a "primary" data source that sets the foundation for your viz and a "secondary" source that adds context. Fields from the secondary source are always aggregated. You’ll know you’re blending when you see a little orange chain link icon next to a field name, meaning it's the "linking field."

When to use it: When your data lives in separate, published data sources. It is often considered a last resort because it's less flexible and performant than relationships.

How to Combine Data Using Relationships (Step-by-Step)

Let's walk through the most common scenario: establishing a relationship between tables from a single database. We'll use a sample spreadsheet with Orders and People tables.

1. Connect to Your Data: In Tableau, select your data connector (e.g., Microsoft Excel). Navigate to and select your file.
2. Drag Your First Table: From the left-hand pane on the Data Source page, drag your main table (like Orders) onto the canvas. This will be your primary table.
3. Drag Your Second Table: Now, drag your second table (e.g., People) nearby on the canvas. Tableau will automatically detect common field names and create a relationship–that orange noodle will appear.
4. Configure the Relationship: Click on the noodle to open the "Edit Relationship" settings. Tableau is pretty smart at guessing the joining fields, but you can change them here if needed. For our Orders and People tables, it would correctly identify the Region field as the common link.

That's it! Your tables are now related. When you go to a worksheet, you can drag fields from both the Orders and People and Tableau will handle the aggregations and connections behind the scenes.

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 a Join in Tableau (Step-by-Step)

What if you need to perform a classic physical join? You can access the join canvas by diving into the "physical layer" of your data source.

1. Go to the Physical Layer: From the relationship canvas we used above, double-click on your first table (Orders). This will open up the physical layer view, which might look more familiar to users of older Tableau versions.
2. Add a Table to Join: Now, drag a second table (e.g., Returns) into this view. Tableau will immediately prompt you with the Join configuration window, signified by a Venn diagram icon.
3. Select the Join Type: Click the Venn diagram icon. You'll see four options:

• Inner: Includes only rows that have a match in both tables. (e.g., Only orders that were actually returned).
• Left: Includes all rows from the left table (Orders) and any matching rows from the right table (Returns). Unmatched rows will have null values for the fields from the right table.
• Right: Includes all rows from the right table and matching rows from the left. The reverse of a Left Join.
• Full Outer: Includes all rows from both tables, whether they have a match or not.

1. Set the Join Clause: Below the join type, specify which field(s) connect the two tables. For Orders and Returns, this would be the Order ID.

Once you close this window, Tableau will have created a single, flat table for you to work with. Remember to double-check your row counts after a join to make sure you didn’t accidentally duplicate data.

Tips for Better and Cleaner Data Combinations

Whether you're using relationships, joins, or blends, a successful outcome depends heavily on the quality of your source data. Here are a few tips to keep in mind.

• Clean Your Data First: The most common reason joins fail is inconsistent data. Ensure your linking keys are the same data type (e.g., both Numeric) and format. Watch out for extra spaces, different capitalization, or slight naming variations (USA vs. United States).
• Understand Granularity: Be aware of the level of detail in each table. Are you joining daily summary data to individual transaction data? Doing so can create a many-to-many relationship that massively inflates your numbers. Try to join tables that exist at a similar level of detail.
• Start Small and Verify: Don't try to join ten tables at once. Add one at a time and spot-check the results. A simple way to do this is to create a quick table showing the "Number of Records" measure. If that number jumps unexpectedly after a join, something is likely wrong with the join logic.
• Rename for Clarity: After you combine several tables, you might end up with confusing or redundant field names. Take a moment to rename fields on the Data Source page or in the worksheet to make your analysis intuitive (e.g., rename customerid to Customer ID).

Final Thoughts

Combining data is one of the most fundamental skills in analysis, transforming siloed metrics into a connected story. In Tableau, using relationships should be your default for its flexibility, while physical joins offer precise control when you need to construct a specific dataset, and blends provide a solution for disparate systems. Knowing which one to use and when is the key to creating comprehensive and accurate dashboards.

While mastering Tableau joins is a powerful skill, manually connecting various marketing and sales platforms and modeling the data is often where teams get stuck. The process can be time-consuming and requires a deep understanding of data structures. At Graphed, we automate the hard part. You can connect sources like Google Analytics, Shopify, and Facebook Ads in seconds, and then use simple natural language to generate cross-platform reports. Instead of worrying about join types, you can just ask, “Show me my revenue versus ad spend by campaign for last month," and we instantly build the dashboard for you with live, real-time data.