Blends are one of the most powerful features of Looker Studio, but they’re a little tricky to get your head around. In this blog and the video below, I’ll review some of the different types of blends that you can create in Looker Studio and the gotchas with blends that you’ll likely run into. I’ll cover the three most common blend types used in Looker Studio: left outer joins, inner joins, and full outer joins. Finally, I’ll share several visualizations that use blends that we’ve used in our marketing dashboards.
You can use blends to bring together all sorts of marketing data. You can consolidate data from different sources, for example:
You can also use blends to combine different types of data based on the same source. This is useful when you want to display a list from a data source (like a list of web pages) with some calculated data from the same data source. Examples include:
Scroll to the end of the post for more examples with links to step-by-step instructions of reports that leverage blends.
The blog below shares the steps to work with blends, and I walk through the same steps in this video.
In this video:
00:00 Introduction to blends in Looker Studio
01:31 The sample data
03:11 Left outer & right outer joins
10:35 Inner joins
12:06 Full outer joins
Follow along with the sample Looker Studio dashboard and the Google Sheet with sample data. You’ll need to make copies of both and reconnect your data sources to your Google Sheets copy to create your own editable dashboard.
We’ll start with a sample data set in Google Sheets that looks like this.
I’ve changed the color of the text on the left table to red for the fruit types that are not on the right table and color coded the fruits on the right table that are not in the left table to blue, just to make it a little easier to see which fruits are in common between both sides.
Let’s start by connecting Looker Studio to your data source, a Google Sheet.
Let’s look at how left and right outer blends work. We’re going to start by creating a blend.
You could create a blend by going up to the Resource menu and selecting “Manage blends”. We’ll do it another way:
The default chart shows Quantity as the column labels, weirdly, so change the Dimension to “Fruit”, and include both “Quantity” and “Quality (1-5)” under Metric.
Let’s look at the configuration of the blend.
In the Chart properties column, under Data source, click the pencil icon next to “Blended Data”. You’ll see the data we set up on the left and the data we set up on the right.
Click on the “1 Condition” box. This window shows the different ways that we can combine that data in a Looker Studio blend. The word “condition” here refers to the join condition. A join condition is a term that comes from SQL, and describes how you want the data to be combined when you’re combining two tables of data.
The default selection is “Left Outer”, which Google describes, “returns matching rows from the right table plus non matching rows from the left table”
What does that mean? In a left join, the blend will return matching and non-matching rows from the left table (meaning it’s going to return all rows from the left table) and only matching rows from the right table.
A right outer join is exactly the same thing, but it will “return matching rows from the left table plus non matching rows from the right table”. So it’s the inverse of what a left outer join does.
Now let’s have a look at the data in the left outer join.
The left table shows we have cherries, grapefruit, and oranges. But blueberries, bananas and apples are only in the right table. The blended data table has a quantity for all fruits, because there are values for quantity in all rows in the left table. But it only got quality for the fruits that matched in the right table. It returned null values for rows in the left table that don’t have a corresponding quality value in the right table.
Now, let’s play around with this data a little bit. Right now we don’t have to deal with situations where there might be two rows of the same data–two rows of the same fruit. So let’s change that.
Now what did Looker Studio do? It took cherries, and it did what we want it to do, which is that it doubled the quantity. In the case of grapefruit, it doubled the quality. But that’s not actually what we want there.
Let’s fix how the data is aggregating on the right table in the blend.
This is one common gotcha. Pay attention to how Looker Studio aggregates values in your blend to make sure you get the values that you want.
One other circumstance I want to look at is when you have more than one dimension in either your left or your right table.
In the sample data set, on the second tab named “more sample data”, you’ll see each table has a column added for color.
In the second tab of the sample dashboard named “Blend – duplicate rows”, there is a Data Source that connects to each of these.
Let’s create a blend from these two tables
Here’s another gotcha. There’s a problem with the data that’s common when creating a blend where rows have multiple dimensions. In our source data, the red cherries have a quality of 5, while the yellow cherries have a quality of 3. In the table we just created for the blend, those values are being averaged to 4.
Because the right table wasn’t showing the color dimensions (even though the underlying data does include the color column), Looker Studio created a blend and joined using only the fruit column/dimension.
Let’s adjust that.
This is such a common gotcha in a circumstance where you’ve got dimensions on your left, dimensions on your right, and your metrics aren’t aggregating the way that you want them to. It is often because you need to have the dimensions that make a row unique in both tables that you’re blending together, and then make sure that those dimensions are used in the blend’s join conditions.
Let’s return to our first set of sample data.
In the left outer join blend, because blueberries, bananas and apples don’t exist in the right side data, those rows show null values for Quality.
What happens if we change this to an inner join? Google says an inner join “returns only matching rows from the left and right table.” You can see how this works by changing the blend configuration for this table.
Now all it’s showing us are the rows of fruit values that exist in both tables. You would choose an inner join over a left outer join in cases where you only want to see data with values in both tables, for example you are only interested in the fruit that you know both the quantity and the quality for.
The gotchas I described in the previous section apply here as well. If you have multiple dimension columns that really define the uniqueness for a row of data, you need to make sure to include those dimensions in your join condition, exactly the way you would with a left blend.
A full outer join “returns all rows from the left tables, and right tables, whether they match or not.” Change your table so that it’s using a full outer join.
What happened to limes, lemons and mangoes? Weirdly, the quality value for lime (5), lemon (4), and mango (2) is summed up under this “null” line.
If you look under the Chart column on the right, you’ll see the Dimension listed is Fruit (Table 1), meaning it’s taking the dimension from the left table–it’s treating fruit in one table as a separate dimension from fruit in the second table.
So how are we going to deal with this? We’re actually going to create a custom field to solve this problem. The custom field will blend the Fruit dimension from both table 1 and 2.
Coalesce (Fruit (Table 1)),(Fruit (Table 2))
The Coalesce() function returns the first value that is not null from a list of values.
This looks better. In cases where the fruit exists in both tables, we see there’s values for both quantity and quality. The fruits that exist on the left and not the right, we get quantity for but not quality, and vice versa: the fruits that exist on the right and not the left we’re getting these values for.
Cross join “returns every possible combination of rows from the left and right tables”. It’s just pandemonium! I have never needed to use a cross join in a Looker Studio dashboard. If you come up with a reason why it’s useful, I’d love to hear.
With blends, you could join yet another table of data…and keep going. You can add up to five tables to join together.
There are so many ways blends are useful in Looker Studio, and I’ve recorded a lot of videos that leverage blends for marketing data visualizations. Take a look at the examples below, which all contain step-by-step instructions on how to create charts and graphs leveraging blended data.
At Two Octobers, we love working with GA4 and other forms of marketing data to create terrific data visualizations. We offer GA4 and Looker Studio training, as well as analytics & dashboard services for businesses of all kinds.
Nico loves marketing analytics, running, and analytics about running. He's Two Octobers' Head of Analytics, and loves teaching. Learn more about Nico or read more blogs he has written.
Report on traffic from people clicking through from AI services like ChatGPT. Build an exploration…
The latest developments in digital marketing include Meta feature updates, Google downgrading ranking on vastly…
ChatGPT traffic in the GA house! Plus new features in GA4 and understanding GTM first-party…