Getting Started
Segmentation is Analytics’ most popular customer analytics tool. It offers a simplified query builder to generate fast, meaningful results and create informative visualizations. Ask, answer, and act on the most important questions about your customers and their behaviors.
- Explore general patterns of events and user interactions
- Create user segments based on observed behaviors and attributes
- Build KPIs and metrics to measure customer activity over time
The Query Builder
Within the query builder, you’re able to construct a complex data search using a variety of field types. Every Segmentation query begins with an event. Start by setting whether you’d like to calculate the total count of events or the number of users who performed a particular event.
Additionally, you may categorize your results by using the Group By function or narrow your search by using a Filter Where function.
The Group By and Filter Where function are akin to SQL “group by” and “where” clauses. You can use event properties, user properties or user segments with any event you select in your segmentation.
-
Event properties are tied to a single instance of an event
- Given a scenario where a user downloads an app on iOS today and then downloads it on Android tomorrow: if the analysis looks at “Users who performed App Downloads grouped by event property Platform”, the count would be 1 for iOS and 1 for Android.
-
User properties can be chosen on any event regardless if it came in with an event since it’s part of the user’s profile.
- User properties have a chronological control and can be configured to be measured as the first value or last value. Using the example above, assume user property Platform is set to the last value. If the analysis looks at “Users who performed App Downloads grouped by user property Platform”, the count would be 1 for Android and 0 for iOS. This is because Android is the last value seen for platform for that user.
-
User Segments cross check user identities from their segment memberships.
- Special Note: User Segments using a specific time frame can be used with Filter Where to show members of a User Segment to help understand users who performed an action in two different time frames. For example, a User Segment looking at the count of users who were active yesterday can be contrasted with the count of users who were active this month, creating a DAU/MAU analysis.
Finally, you can create a combination of events by using additional For Clauses (+Did [Not] Perform).
Frequency Query
You can also run a Frequency query to group users into different segments based on the number of times they performed a particular event.
Additional Query Builder Settings
All queries within Analytics are fully customizable using Settings such as the date range and interval. Event queries may display results on a per-interval basis or as a cumulative count. There are six different chart types to choose from:
Optional settings include Annotations, which mark milestone events, and Data Labels, which communicate the numerical values at each point in the chart.
For more advanced analysis, you may create queries with multiple rows, displaying and comparing information of distinct events within the same results. You may also create calculated queries to explore, for example, proportions or percentages using the Calculator. Finally, you may compare results to a previous interval by using the Trends function.
Certain analyses run in the past can change based on what user properties are set to since user properties can change over time. A common scenario where changes are seen when running analysis at different points in time, is a direct result of looking at users who were aliased since then.
Note on Aliasing: Aliasing runs once on a daily basis and links unknown users to known users. The count of unique users triggering an event seen before the aliasing process may decrease after aliasing finishes processing. You can read more by visiting our docs on user aliasing.
Helpful Tools
A few helpful tools are located to the right of the query builder:
Was this page helpful?