How to Merge Datasets in Tableau

Bringing different datasets together is one of the most powerful things you can do in data analysis, but it's often the most challenging part of the process. This guide walks you through the three main ways to merge data in Tableau - Relationships, Joins, and Data Blending - so you can create comprehensive reports and unlock deeper insights.

Why Merge Datasets in the First Place?

Your business data rarely lives in one perfect, all-knowing file. More often, it's scattered across different systems and spreadsheets. Sales transaction data is in your CRM, website traffic data is in Google Analytics, and ad spend is in Facebook Ads Manager. Combining these sources is essential for answering critical business questions.

Imagine you want to see which marketing campaigns are driving the most profitable customers. To do this, you need to merge your advertising platform data (like campaign names and costs) with your sales data (like customer purchase history and revenue). By combining them, you can build a report showing the true ROI of your marketing efforts.

Merging data lets you:

• Get a holistic view: See the complete picture by connecting disparate parts of your business.
• Enrich your analysis: Add context to your data. For example, merge sales data with demographic data to understand who your best customers are.
• Answer complex questions: Tackle questions that can't be answered with a single dataset, like "How does social media follower growth correlate with sales in our new product category?"

Understanding Tableau's Three Methods for Merging Data

Tableau offers three different ways to combine your data, each suited for different situations. Understanding the distinctions is the key to building efficient and accurate dashboards.

Let's briefly break them down before diving into the step-by-step instructions for each.

1. Relationships (The Modern, Flexible Default)

Introduced in recent versions of Tableau, Relationships are now the recommended method for combining tables from the same data source. Instead of physically merging tables into a single giant table, Relationships tell Tableau how your tables are related to each other. Think of it like a smart contract between your tables.

The magic of Relationships is that they query data at the right level of detail as you build your visualizations. This makes your workbooks more efficient and avoids common issues like data duplication that can happen with traditional joins. You’ll use relationships for most of your work inside Tableau.

2. Joins (The Traditional, Physical Merge)

Joins are the old-school way of combining data. When you join two tables, you are creating a new, single, fixed table that contains columns from both. This happens in the "physical layer" of your data source tab, before your analysis even begins.

While still useful, joins are less flexible than relationships. Since you're creating a permanent, merged table, you can sometimes accidentally duplicate data or lose rows if you choose the wrong join type. There are four primary types of joins:

• Inner Join: Returns only the matching rows from both tables.
• Left Join: Returns all rows from the left table and the matched rows from the right table.
• Right Join: Returns all rows from the right table and the matched rows from the left table.
• Full Outer Join: Returns all rows from both tables.

3. Data Blending (For Completely Separate Data Sources)

Data Blending is what you use when your data lives in entirely different data sources that you can't join at the database level - think combining data from a Google Sheet with a report from Salesforce. Unlike joins and relationships, blending works at the sheet level, not the data source level.

It works by taking aggregated data from a "secondary" source and displaying it alongside data from the "primary" source. The key here is aggregation, it doesn't combine row-level data. It's an essential tool when a true join or relationship isn't possible.

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

Relationships are Tableau's default for a reason - they're powerful and intuitive. Let's walk through how to create one using an example of "Orders" and "Returns" data from a fictional superstore.

1. Connect to Your Data: In Tableau, connect to your data source (e.g., Excel, a SQL database). For this example, we'll use the sample "Superstore" dataset that comes with Tableau.
2. Drag Your First Table to the Canvas: From the left-hand pane, drag your main table (e.g., Orders) into the workspace area. This becomes the base of your data model. The canvas will now show a box representing your "Orders" table. This is the "logical layer" where relationships live.
3. Drag Your Second Table to the Canvas: Now, find your second table (e.g., Returns) and drag it near the Orders table on the canvas. Tableau will automatically try to create a relationship, indicated by a line (often called a "noodle") connecting the two tables.
4. Configure the Relationship: Tableau is pretty smart and often finds the common field correctly. In this case, it automatically detects that both tables have an Order ID field. If you need to change the fields, just click on the noodle. A dialog box will pop up where you can manually select the fields that link the two tables together. You can also specify performance options like cardinality, but the defaults are usually fine for getting started.
5. Start Analyzing: That's it! Now go to a new worksheet. In the Data pane on the left, you will see both tables listed. You can now drag fields from both the Orders and Returns tables into your view, and Tableau will handle the querying behind the scenes automatically.

How to Create Joins in Tableau (Step-by-Step)

Sometimes you need more direct control or are working with data sources that require joins. Here’s how you'd combine the same "Orders" and "Returns" tables using a traditional join instead.

1. Connect to Your Data: Connect to your data source as you did before.
2. Open the Physical Layer: Drag your first table (Orders) onto the canvas. Now, to create a join instead of a relationship, double-click the Orders box. This will open the "physical layer" where you build joins.
3. Drag in the Second Table: From the left pane, drag the Returns table onto the physical layer canvas, to the right of the Orders table. A Venn diagram icon will appear between them, indicating a join is being formed.
4. Configure the Join: Click the Venn diagram icon. This opens the join configuration menu. Here you can:
5. View Your Merged Data: After configuring the join, you can see a preview of the new, single table in the data grid below. All rows are now physically combined. You can now go to a worksheet and start your analysis using fields from this newly created table.

How to Use Data Blending in Tableau (Step-by-Step)

Data blending is your go-to when your datasets are completely separate. Let’s imagine we have sales data in our main "Superstore" file, and our monthly sales targets are stored in a separate Google Sheet.

1. Connect to Your Primary Data Source: First, connect to your main data source (the Superstore file) and navigate to a new worksheet. This becomes your primary source.
2. Build an Initial View: Create a simple chart from this primary source. For example, drag Order Date to Columns and Sales to Rows to create a line chart of sales over time.
3. Add a Secondary Data Source: Go to Data > New Data Source and connect to your second source (the Google Sheet with sales targets). It might have columns like Month and Sales Target.
4. Link the Fields: Once connected, Tableau automatically returns you to your worksheet. You will now see both data sources in your Data pane. The original Superstore source will have a blue checkmark, designating it as the primary source for this sheet. The new targets source will have an orange checkmark. For blending to work, there must be a common dimension between the two sources. In this case, it’s the date. Look in the Data pane under your secondary (orange) source. Tableau will look for fields with the same name. If it finds one, a small gray or orange link icon will appear next to the dimension. Here, Month(Order Date) would be the linking field. If the link icon is gray, click it to activate it (it will turn orange).
5. Add Fields from the Secondary Source: Now, drag the Sales Target measure from your secondary Google Sheet source onto your view. You just blended your data! The sales targets will be aggregated to the level of the linking dimension (in this case, by month) and displayed alongside your actual sales.

Relationships vs. Joins vs. Blending: Which Should You Use?

Feeling a bit confused about when to use which method? Here's a simple cheat sheet:

• Use Relationships for almost everything. If your tables are in the same data source, start here. It’s the modern, flexible, and performant standard in Tableau.
• Use Joins only when you must. If you need to create a single table for a specific reason (like advanced database functions) or if your data model is very simple, a join is a good option. Think of it as a specific tool for a specific job.
• Use Data Blending when your data is in different places. If you need to bring in data from a completely unrelated source (e.g., Salesforce and Google Analytics) and can't use a more robust solution like a data warehouse, blending is the way to go.

Final Thoughts

Mastering how to merge data in Tableau fundamentally changes your ability to produce high-value reports. By understanding Relationships, Joins, and Data Blending, you can effectively combine scattered information into a single source of truth and uncover powerful insights that would otherwise remain hidden.

Of course, prepping and connecting all this data manually can be a significant time sink, especially when you’re pulling from multiple marketing and sales platforms. That's why we built Graphed to help. We simplify the entire process by providing one-click integrations to sources like Google Analytics, Shopify, and Facebook Ads. You can just ask a question in plain English, and we generate the live, multi-source dashboard for you, completely bypassing the manual setup of relationships or blends.