Filters give report creators the option to limit the data they want to display in the report in Looker Studio. In this post, we will see how to create and apply them.
Let’s remember that the data in our reports in Looker Studio can be filtered in two ways: with components called controls, or by creating filters. Using controls, the end user of your report can filter the information in the charts without having editing privileges. However, to create filters, you must be able to edit the report.
In this post, we will focus on filters, but you can visit the post dedicated to controls in Looker Studio to learn more.
In this case, we will use the Spotify data source. Visit the post on how to add data sources in Looker Studio, if you need a refresher.
1. Why use filters?
Filters are very powerful because they allow us to:
- Segment our data into more manageable parts. By focusing on a subset, we can conduct more targeted and detailed analyses.
- Narrow our focus and eliminate noise.
- View the data in a personalised way that better meets our needs.
Thanks to all of the above, working with filters facilitates the decision-making process based on data.
2. How do we create filters in Looker Studio?
There are several paths that lead us to create filters. The most common are:
- In the top menu, under Resource, select Manage Filters. There, you will see a link to Add a Filter and, if there are any created filters, a list of all available ones. This path will be used whether we want to create a new filter or view or modify an existing one.
- By selecting a chart, within the Set-up column at the bottom, you’ll find Add a Filter. If there are filters created, you’ll see all available filters and also the option to Create a Filter. If there are no filters created, it will take you directly to the screen where you create the filter.
Filters can be added at three different levels:
- At the chart level, where it will only affect the specific chart and is added within the chart’s set-up, as we just saw.
- At the page level, where it will affect all charts and controls on that page.
- At the report level, in which case it will affect all components of the report.
3. Steps to create filters in Looker Studio
To create a new filter, we need to follow these steps:
Step 1
The first thing you will see on the screen where the filter is created is a blank field to enter the name. Filters can have the same names as existing ones, so be careful, as it can be confusing to have two with the same name.
Step 2
To the right of the name is the data source for which the filter is being created. If you are creating it directly from a chart, this cannot be changed and will show the source that the chart has. If you are creating it from Manage Filters, in Resource, you can select any other source you have added to the report.
Step 3
Here is where you can start adding conditions. The first you will encounter is Include or Exclude, as the dimension or metric you choose can be included or excluded.
Suppose you want to create a filter for all the songs sung by Taylor Swift. In that case, you would select the dimension artist(s)_name, set the condition to Equals =, and type Taylor Swift in the text field.
Up to this point, everything looks good because you will see that applying the filter returns data in the charts. But what about those songs where Taylor Swift collaborates with other singers?
It’s an easy fix; you just have to change the condition Contains or RegExp Contains instead of Equals =. This way, it will look for all values containing Taylor Swift.
What if you want to see all the songs sung by Taylor Swift or Ed Sheeran? For this, you will have to add two conditions connected by OR:
- artist(s)_name Contains Taylor Swift
- OR artist(s)_name Contains Ed Sheeran
By using OR, you are stating that the filter must meet either condition.
With the AND condition, we can be more specific. If you wanted to filter those in which Taylor Swift and Ed Sheeran sing together, it would be:
- artist(s)_name Contains Taylor Swift
- AND artist(s)_name Contains Ed Sheeran
Filters distinguish between uppercase and lowercase, so if you type taylor swift and ed sheeran, it will not return results, since in this dataset, the first and last names start with uppercase letters.
To address this, I always include the source written in different ways. You can do this as I explained above by adding conditions with OR, or by using regular expressions. They are super useful to make the work easier (and faster!).
4. Troubleshooting
What happens when the filter you created does not match any data?
The chart will show No data.
What if you apply a filter that uses a different data source than the one the chart has?
In this case, it will give you an error and mark the filter in red.
How Can You Verify That Your Filter Has Been Created Correctly?
Let me give you an example of what I have done in the past to verify a filter.
First, I will create a table with the dimension artist(s)_name and the metric Record count, and I will add the filter we created for Taylor Swift (without Ed Sheeran).
Next, I add the fixed-size list control to the report and use the search box to look for Taylor Swift. In this type of control, we don’t have the problem of it being case sensitive, so it will show me the different ways it is written, even if I type everything in lowercase.
By comparing the table and the fixed control, I can see if both have the same data, thus confirming that my filter is capturing all the names it should.
So far, it seems easy, right? There shouldn’t be hundreds of artists containing Taylor Swift, so it should be straightforward to verify.
What happens if there are other ways in which Taylor Swift is written that do not include her full name?
It may happen that Taylor Swift is written as T. Swift. In this case, it’s worth applying a filter that captures all artist(s)_name containing Swift, so we can see all the different names for Taylor Swift. However, we should first check that there are no other artists with Swift in their name.
Fortunately, in all cases of this dataset, she is referred to by her full name.
If you need to edit the filter you applied to a chart, Looker Studio makes it easy. If you click on the chart, in the set-up column where you can see its filters, you will see a pencil next to its name that allows you to edit it without having to go to Resource → Manage Filters. However, any changes you make to the filter there will change it in all charts where it exists.
Filters can also be applied at the page level, meaning they will exist in all visualisations on the page without the need to add them one by one to each chart. I explain this in another post about pages actions in Looker Studio.
Feel free to leave me a comment if you have any questions. As I always say, comments and questions are welcome! 🙂