Blended data sources, or blends, are data sources where information from two or more different datasets is merged into a single dataset.
Data sources are the core of your reports, and without them, no visualisation would be possible. A few years ago, Looker Studio (then Data Studio) began offering blended data sources, referred to as blends in the tool.
For this post, I will use the Superstore Products and clients tables. As always, you can refer to the poat on how to add data sources in Looker Studio to learn more about them and how to access them.
1. What are blends?
Blends are data sources that merge information from two or more different datasets into one single dataset. These datasets can come from the same connector (such as Google Sheets) or from multiple different ones (like Google Analytics and Google Ads). You can blend up to five tables.
2. Why blend data?
Blends allow you to generate new insights that would not be possible by analysing the datasets separately. For example, by combining Google Analytics and Google Ads, you can gain insights into the correlation between user behaviour (from Google Analytics) and the performance of your campaigns (from Google Ads) in a single visualisation. Similarly, by blending data from your CRM, you can gain access to additional information that helps better understand your customers’ journey to conversion or repeat purchases.
In the CRM example, using Superstore Products, you can create a table in Looker Studio displaying the transaction ID, product name, net revenue, and customer ID associated with each transaction.

Then, using the clients data source, you can create another table with the customer ID, age, gender, profession, and annual salary.

By joining these two tables using the customer ID, you’ll have much more information about the type of customer, allowing you to visualise data such as net revenue by gender—something that would otherwise have been impossible.

3. How to add blends?
Before you can create a blend, all data sources to be combined must be added to your report.
To add data sources in Looker Studio:
- From the top menu, navigate to Resource → Manage added data sources → Add a data source.
- Or, you can simply click Add data from the menu directly above the canvas for a one-click solution.
Once this is done, there are two ways to create blends:
- From the top menu, select Manage Blends under Resource and click Add a blend.

- From the set-up column when selecting a chart. Below the data source, you’ll see the option to Blend Data. In this case, the data source used by the chart will appear as the first table in the blend, although this can be modified during the setup.

After entering the configuration, select Join another table to add another dataset. You can join up to five tables.

When selecting Join another table, a dropdown box will appear where you can choose the new data source to add.

Once the new dataset is added, there are two crucial aspects to configuring your data blend, which are set in the join configuration box between the two tables you want to merge: the join conditions and operator.

3. Join conditions
This simply means deciding which field(s) match across the two datasets and will be used to join them. In the earlier example of purchase lists and customer data, choose Customer ID as the join key.
The field may have different names in the two datasets being merged (for example, one might be Customer ID and the other Client ID). That’s not a problem as long as they represent the same variable.

Tables in a blend can be joined by one or more fields.
4. Join operator
The join operator specifies how the rows from the left and right tables are combined. There are five options available:
a. LEFT OUTER

What is it?
It returns all rows from the left table and only the matching rows from the right table.
Example
Imagine you have a table of December 2013 transactions on the left and a table of VIP customers on the right. You want the result to be a table with the transaction ID, net revenue, and the customer’s profession. The first two fields are from the transaction table, and the last one is from the customer table.
In a left outer join, the final table will show all rows from the left (with transaction ID and net revenue for every row), but it will only display the profession for customers who appear in both tables. For others, it will show null.
In this example, the left outer join results in 173 rows, corresponding to the number of rows in the December 2013 transactions table. The number of rows in the final table will be at least the number of rows in the left table.

Important
The transaction table only contains transactions from December 2013, so a filter needs to be added to the left table during configuration to show transactions from December 1 to 31, 2013.


b. RIGHT OUTER

What is it?
It returns all rows from the right table and only matching rows from the left table.
Example
Using the same example, a right outer join will return information about the profession in every row, as the right table (VIP customers) is shown in full. The transaction ID and net revenue will only be displayed in rows where the customer ID matches in both tables, with nulls for the rest.

In this case, the result will contain 111 rows, which is the size of the VIP customers table.
c. INNER

What is it?
It shows only the rows present in both the left and right tables.
Example
Continuing with the same example of December 2013 transactions and VIP customers, an inner join will display only rows where the customer IDs appear in both datasets. In this case, the final table will have 68 rows, representing transactions from customers found in both tables.

d. FULL OUTER

What is it?
It returns all rows from both tables, whether they match or not.
Example
If you join the 2013 transactions with the VIP customers, a full outer join will return all rows from both tables. The profession field will be null for customers in the transaction table that are not in the VIP table, and the transaction ID and net revenue will be null for customers that are only in the VIP table.

Don’t forget to apply the filter to the first table to only show transactions from December 2013.
e. CROSS JOIN

What is it?
It returns all possible combinations of rows from the left and right tables.
Example
Let’s use a different example for a cross join. You have a table with four foods (pizza, burger, pasta, salad) and another with four drinks (Coca-Cola, orange Fanta, pineapple juice, water). A cross join would create all possible food and drink combinations, which is ideal for building meal menus.


Cross joins don’t require a join field.
In the next post, I’ll address one of the most common issues when creating joins. Keep reading, and feel free to leave any questions in the comments!