What’s cardinality, and how to avoid the (other) row in GA4? 

Cardinality is one of the terms used in Google Analytics 4 to describe the number of distinct values assigned to the dimension ( or custom dimension). 

Because you are reading this article, you may find (other) rows in your GA4 reports and want to understand how to resolve this issue, or you want to learn more about how not to reach this situation in future. You are on the right page; this article will cover them all. 

What’s a cardinality in Google Analytics 4?

As mentioned above, cardinality is the number of unique (distinct) values assigned to the GA4 dimension. 

For instance, “gender” has only two assigned values: “male” or “female”. The “device category” dimension has four distinct values: “desktop”, “mobile”, “tablet”, and “smartTV”. 

If the dimension has more than 500 unique values assigned, this will be called a high-cardinality dimension. 

Examples of high-cardinality dimensions in Google Analytics 4 are page_location, item_name, product_name, etc. 

It doesn’t mean that these dimensions will be high-cardinal for your case, but they usually have more unique values than others. Therefore, they are more frequent cases.

Suppose your GA4 setup wasn’t correctly completed, and you have URL query parameters that generate unique page URLs for almost every user (or for every page view). In that case, page_location will become a high cardinal dimension in a few days. 

What’s the sequence of high cardinality?

So, high-cardinality dimensions can negatively impact your GA4 reports. Mainly, you will stop seeing the entire website or app data in some GA4 reports because GA4 starts to aggregate it under the “(other)” row.

The standard GA4 properties reports usually have at most 50,000 rows daily, so you are in trouble if you have 300,000 page views and each generates a unique page location.

As a result, your analysis will present only some of your sample of your users, and the insights will be misleading. 

Therefore, avoiding high cardinality dimensions in Google Analytics 4 is crucial. Luckily, there are concrete ways how to do it. 

How to avoid high cardinality and the (other) row in GA4 reports

Although there are many ways how to fight high cardinality, all of them can be split into two categories:

• Leave the high-cardinality dimension as it is and try to overcome (other) rows in GA4 reports using other reports and methods (GA4 Explore, standard GA4 reports, migrate to GA4 360, etc.)

• Change your Google Analytics 4 setup to eliminate high cardinality dimensions or decrease the number of distinct values assigned to it. 

The best approach is to change your GA4 setup and decrease the cardinality of the dimensions you use or not register custom dimensions with high cardinality. Let’s look at each of them individually. 

1.1 Use the standard Google Analytics 4 reports or GA4 Explore

Google Analytics 4 includes many reports; some can show more rows while others less. It’s based on the way how Google Analytics 4 aggregates data. 

For instance, standard reports usually keep the data pre-aggregated, and therefore they require less time to load and can show more rows.

Another way is to build reports in Google Analytics 4 Explore. It allows to have access to more rows as well.

So if your report shows (other) row, try to use the standard report in GA4 or prepare the same data in GA4 Explore.

1.2 Migrate to GA4 360

If you have a budget, consider migrating to GA4 360 because it allows access to the lower limit of rows in the reports and see raw data. So you can move from 50k rows per day to 75k.

However, this option is not for everyone. At the moment of writing, GA4 360 costs around $50,000 per year. 

It’s the primary reason why the most recommended way to fight cardinality is to decrease dimension cardinality or remove it from GA4.

2.1 Remove URL query parameters 

One of the most powerful ways is to look at your dimension that includes so many unique values and ask yourself whether you need all of them. 

For instance, your page location can store many unique URL query parameters you don’t use. If this is the case, consider removing them by following this guide. 

2.2 Use in-built dimensions instead of custom dimension

When you set up Google Analytics 4, you should always follow automatically tracked and recommended events. 

These events have in-built event parameters allowing you to pass critical data. Usually such dimensions are not subject to high cardinality and can allow you to see raw data in Google Analytics 4.

Besides that, don’t create custom dimensions for such unique user attributes as “user_id”, “client_id”, use “user_id” GA4 feature instead. 

2.3 Consider using BigQuery 

If none of the above methods helped you and the information is critical in your analytics, you should consider using BigQuery instead. 

In this case, you send data into GA4 as the event parameter but don’t register it as a custom dimension in GA4. When you need to make ad-hoc analysis using this dimension or build a report, you can SQL this data in BigQuery and use Looker Studio to visualise it. 

Although this method requires knowledge of SQL, it’s one of the best ways to avoid migrating to GA4 360 or having “other” in all your reports. 

Final Word 

Cardinality is the number of unique values a particular GA4 dimension has. If the dimension has more than 500 distinct values, it’s considered a high-cardinality dimension and can cause some reports in Google Analytics 4 to show (other) rows. 

As a result, some of your insights won’t represent the entire sample of your users and will be misleading. 

It’s best to look at your GA4 setup to resolve it. You should find the dimension that caused it and decrease the number of values by grouping some values, removing URL query parameters or using recommended dimensions and GA4 native features. 

If nothing helps and this dimension is precious for your business, consider using BigQuery instead and visualise the data using Looker Studio. 

Please comment below if you need help with cardinality in your GA4 property. It’s FREE! 

Frequently Asked Questions