User Composition: View your users by common properties

The User Composition chart shows the breakdown of active users based on a single user property or group property. This grouped view provides insight into who your users are and what properties they share.

Before you begin

Events don't appear in Amplitude charts until instrumentation is complete, so finish that first. For more information, review building charts in Amplitude and the difference between user properties and event properties.

Set up a user composition analysis

Most Amplitude charts rely on the Events Module to build an analysis. The User Composition chart isn't an event-driven analysis and doesn't have an Events Module, so it works differently.

Instead of events, a user composition analysis relies on user properties or group properties. Select the user property or group property you're interested in, then define your user segments. The User Composition chart displays a breakdown of values for that property across the user segments you specified.

The example above shows a comparison of your product's users in the United States and Germany, broken out by the most recent value for the number of communities they've joined.

To build your own user composition analysis, follow these steps:

1. In the Composition By Module, select the user property or group property you're interested in.

2. In the Segment By Module, identify the user segment to include in this analysis. To import a previously saved segment, click Saved and select the one you want from the list. Otherwise, Amplitude targets all users by default.

3. To build your own user segment instead, add properties. Click + Filter by, choose the property to include, and specify the property value you're interested in.

4. To narrow the focus further to users who already performed certain actions, click + Performed, then choose the event you're interested in.

5. Click + Add Segment to add another user segment, repeating steps 2 through 4.

6. In the Measured As Module, specify the property values you're most interested in:

   • Most Recent Value: Considers your users' most recent values of that property. Amplitude draws this value from a user's most recent active event. Users can only appear in one bucket when you select Most Recent Value.
   • All Values: Includes every value your active users have had for the property during the time of your analysis. The User Composition chart only includes active users, so Amplitude doesn't return property values tied to inactive events.
   • Cross Property Values: Shows sets of properties active users have had in the time range you selected. These buckets are mutually exclusive; users can only fall into one bucket.

7. Use the date picker to specify the timezone and set the time range of your analysis.

In all cases, Amplitude breaks out the top 13 values for your property and groups other values in the Other bucket. The value in the center of the chart is the column sum value in the breakdown data table below the chart. It includes only users in the top 100 property value groups by user count.

You can also view your results as a bar graph.

You can always inspect the composition of your user groups from the Microscope.

Common questions

I ran an Identify API call for a user. Why isn't this reflected in the User Composition Chart

An identify call updates the user property for events moving forward without sending an event. The Identify call appears at the top of a user's profile, but doesn't appear in the User Composition chart results until another "active" event is sent after the Identify call.

I see the user property and value in the user profile. Why isn't it showing in User Composition

The property must be available in the most recent active event. If you updated the user property using the Identify API, the user profile shows the updated value. If the user hasn't performed an "active" event, User Composition uses the user property from the most recent active event.

Why am I seeing a high volume of 'none' values

The property wasn't available when the most recent active event was performed. To learn more about why a user is counted under '(none)', refer to this section.

The account shows the user property isn't null. Why does it appear null in User Composition in the last 30 days

Accounts is a special case. Many users can tie to an account, and if all the users perform active events, the user property values change constantly.

Why doesn't the chart display all property values

The User Composition chart displays the top 100 user property values with the highest user count, listed in the breakdown table. To export the table as a .CSV file, click Export CSV in the Breakdown Table. The .CSV file has a limit of 10,000 rows, so you can get the top 10,000 user property values by user count.