[UA] Build new segments [Legacy]

You use the segment builder to configure the individual filters that together constitute a segment. You then apply segments to your reports and dashboards so you can see that specific data.

Segment-builder user interface

The segment builder lets you create the component filters of a segment based on Analytics dimensions and metrics. You choose a dimension or metric, a comparative operator, and you enter a value to set the condition for the filter. In many cases, you can also choose the scope of the data (hit, session, user). For example:

Language - exactly matches - "fr"
(dimension - operator - dimension value)

Revenue - per user - > - "100"
(metric - scope - operator - metric value)

In some cases, the operator is implied, for example, when you use the Age or Gender dimension:

Age - "18-24" (Age = 18-24)
(dimension - implied operator - dimension value)

A Segment can contain up to 20 filters.

Data scope

Each filter in a segment has one or more scopes to the data that it defines:

• Hit: Behavior confined to a single action, for example, viewing a page or starting a video.
• Session: Behavior within a single session; for example, the goals that users completed during a session, or the amount of revenue they generated during a session.
• User: Behavior across all sessions within the date range you’re using, up to 93 days; for example all the goals users completed or all the revenue they generated (across all sessions) during the date range.
• Product: Behavior associated with products. Available only when you use a product-related metric.

You can choose the scope for some filters; for other filters the scope is fixed. When you can choose the scope, you see one or more menus that offer those choices, for example:

When a filter includes multiple scopes, they work together as in the following examples:

Sessions in which the Revenue per hit > 10

Users whose Revenue per Session > 10

Users whose Revenue per hit > 10

How filters are evaluated

Multiple values within the same dimension are joined with OR logic, for example:

• Age: 18-24 OR Age: 25-34

Data is included if it meets either condition.

Multiple dimensions and values within the same category are joined with AND logic, for example:

• (Demographics/Age: 18-24) AND (Demographics/Gender: Female)

Data is included when it meets both conditions.

Multiple metric values within the same category are joined with AND logic, for example:

• Behavior: Sessions > 1 AND Behavior: Transactions per user > 1

Metric values and dimension values within the same category are joined with AND logic, for example:

• Ecommerce: Revenue per user > 10 AND Ecommerce: Product = T-shirt

Data is included when it meets both conditions.

Filters from multiple categories are joined with AND logic, for example:

• Demographics filter AND Technology filter AND Sequences filter

Data is included when it meets all conditions.

Working with Conditions and Sequences

Conditions and Sequences filters are also based on dimensions and metrics, but they offer some additional options:

• Rather than being restricted to specific categories, they let you create filters for any dimension or metric.
• They can include or exclude specific data.
• They can include AND and OR conditions.
• When you include user- and session-based rules in the same filter, those are joined with AND logic. Data is included when it meets both conditions.
• Sequences filters let you determine whether the sequence begins with the first user interaction or with any user interaction.
• When you include multiple steps in a Sequences filter, you can specify that one step can follow another at any time or that it must follow immediately. The subsequent step can occur in the same session or in a subsequent session.

Sessions vs. Count of Sessions

When you create a segment, you may encounter some confusion around using the Sessions metric in the Behavior section, and using the Count of Sessions dimension in the Conditions section.

While both Sessions and Count of Sessions provide a count of sessions at the user level, they differ in the time frames they cover:

• Sessions is the number of times users initiate sessions during the specific date range you are using for the report; for example, Sessions = 5 or Sessions > 5 from January 1 to January 15.
• Count of Sessions is the number of times users initiate sessions over a lifetime, with the final session in the count occurring during the date range you are using for the report (regardless of whether the preceding sessions occurred within that date range); for example: Count of Sessions = 5 means the users initiated their 5th sessions during the date range of the report; Count of Sessions > 5 means the users initiated their 6th or later sessions during the date range of the report.

Create new segments

If none of your existing segments address the data you want to analyze, then you can create new segments and configure them as broadly or narrowly as necessary.

To create a segment:

1. Sign in to your Analytics account.
2. Open the View whose data you want to analyze.
3. Open Reports. You can start with any report.
4. Click + Add Segment...

   

   ...to open the segments list.

   

5. Click + NEW SEGMENT...

   

   ...to open the segment builder.

   

6. Enter a name for the segment.
7. Use the options in the different categories to configure the filters you want for your segment.
   Note: Chrome has implemented updates to enhance user privacy by reducing the granularity of data available via the browser. As a result of these updates, the following fields may now have limited or no data available:

   • Mobile Device Branding
   • Mobile Device Marketing Name
   • Mobile Device Model
   • Mobile Input Selector
   • Operating System Version
   • Screen Resolution
   
   As you add filters, the Summary pane updates to give an estimate of how many users and sessions will be included in the segment. This estimate is based on sampled data, and is not intended to exactly match the number of users or sessions that will be included when you finally apply the segment to your reports.

   

8. When you have finished configuring filters, you can preview and test the segment.
   
   Click Test to see the the percentage of users and sessions that match your segment filters.
   
   Click Preview to see how the segment affects the current report. If you don’t see the information you want in the report, you can continue to edit the segment, and then preview again. (When you click Preview, the Test function is automatically invoked.)
9. When you’re satisfied that the segment includes exactly the data you want, click Save to close the segment builder and apply the segment to your reports.

Copy and modify existing segments

Existing segments can provide a good starting point for new segments.

For example, you could start with the system segment Sessions with Conversions, make a copy, and then add additional filters for things like Country/Territory and Campaign to focus on data about specific countries where sessions with conversions originated, and about specific campaigns that led to sessions with conversions.

To copy a segment:

1. Open the segments list, and locate the segment you want to copy.

   

2. Click Actions > Copy...

   

   ...to open the segment builder with the configuration for the current segment:

   

3. Give the segment a new name.
4. Edit the existing filters, and/or add new ones. For example:

   

   In its original configuration, the segment included only sessions in which Goal Completions were greater than zero. With a little modification, the segment now restricts those sessions to also include only ones that originated in the United States and that were initiated as a result of the Summer Swimwear campaign.
5. When you have finished modifying the original segment, Click Save.

Related resources

• Import segments that other users have created from the Analytics Solutions Gallery
• Manage segments