Template Structure
An Event Segmentation template consists of five modules and they are as follows:
Let's see how each module is configured in the template.
Measure Module
The Measure module in the Event Segmentation template contains a selector that lets you choose a measure to calculate for the selected events and segments.
Running an Event Segmentation analysis helps you measure the following:
MEASURE | DESCRIPTION |
Count of Events | The count of events during a specified time frame. |
Count of unique actors that performed event | The count of unique users triggering events. |
Frequency | The distribution of actors by the number of times they performed an event. |
Intervals Engaged | The count of time buckets for each actor, in which that actor met the engaged event criteria. |
Active Actor % | The count of users who performed any active event. |
Aggregate over property of an event | Custom aggregate for actors which did at least one of the events. |
Metric | A metric from the catalog or a custom block. |
Inside the Measures module, you can also enable period-over-period comparison for each measure by clicking on the over time drop-down next to the Measure drop-down.
Metric is a custom measure of the same event dataset being used in the event segmentation (for example, ProductInteractions). This means that the metric must be aggregating the same event dataset selected in the event segmentation. Let us look at this in detail with an example.
Say, we have a Metric called Freemium to Enterprise Ratio. This metric is simply the total number of events by Freemium tier users, divided by the total number of events by Enterprise tier users. Now, let us consider that we created an event segmentation with this metric (screenshot below). This event segmentation analysis shows that for all events, there are typically 4-6 freemium events for every enterprise event. However for only Add Data Source events, that ratio drops to 1-2. The takeaway is perhaps that Add Data Source is an enterprise-related feature.

The measurement of Count of unique actors that performed events, Frequency, Intervals Engaged, Active Actor %, Average event count per actor, and Metric require data analysis centered around actors such as users. To facilitate this, when these measures are selected, an additional Catalog picker becomes available, allowing you to choose objects specifically constructed for actors. In order to be able to define Active Actor %, you should have an active event annotation on your event stream.

Creating a custom measure over multiple event streams
If the pre-canned measures in Analytics do not fit your requirements, you can also create and use a custom measure inside your analysis. To do this, you need to add a custom formula block and then choose that metric in the Measure module.
Say, you want to create a measure to determine the number of users who have made a purchase on the website but you only want to see users who have performed some content interaction.
- Navigate to Explorations and create a new Event Segmentation analysis.
- Add the events - choose Make Purchase for event A and for event B, let us choose all events that are related to content interaction - Pause Content, Play Content, Rate Content, and Share Content. Click Apply.
- As the next step, create the formula block. Navigate to the shelf on the left and under Blocks, click + and choose Formula.
- The formula to determine this measure is as follows: count(@A) / uniques(@B.user_id) where count(@A) will be the number of purchases and uniques(@B.user_id) will be the number of unique users who performed any kind of content interaction. Click Apply.
- Now, go to the measure module and choose Metric and select the formula block as the value. This will create the custom measure.
- Run the exploration and save your changes.
Events Module
The Events module in the Event Segmentation template contains an event selector that enables you to choose one or more events to analyze.
The event picker features a list of Event Streams that include their respective set of event types for analysis. In Analytics, Event Streams refer to datasets configured specifically in the platform to model stream of events. However, not all datasets are Event Streams as some datasets created on the platform may not include Event Types, such as the Users dataset that includes a list of users rather than events.
The event selector in Analytics is designed to display only Event Types when a single dataset is present on the platform. In this scenario, the Event Streams selector is automatically hidden, and you will only see the Event Type dropdown list when adding an event for analysis. Additionally, the event selector has a filter that enables you to select events based on specified conditions.
Analytics also allows users to choose JSON columns in this module. When you click on a JSON column, it expands to display all the keys that are available for that particular column. You can choose a key and click Apply. Once this is done, the path of the selected field will be used as the display name for that column.
Segmentation Module
The Segmentation module enables you to choose a cohort of actors, such as users, or one or more attributes to include in the analysis. It has two sub-sections: Performed by and Grouped by that allow users to add cohorts and attributes respectively. You can choose to create a cohort either by choosing an existing cohort from the drop-down or use the + New Cohort option to create a behavioural cohort block in one click.
Filters Module
Filters can be used to narrow down data in a visualization. Filters make it easier for the user to answer exploratory questions, for example if the user wants to see results for a specific region, they can define a region filter and see the narrowed down data.
Analytics also allows users to choose JSON columns in this module. When you click on a JSON column, it expands to display all the keys that are available for that particular column. You can choose a key and click Apply. Once this is done, the end key that is selected will be chosen as the display name for that column.
Visualization Module
The Visualization module in the Event Segmentation template enables you to run and view the analysis in an Event Segmentation chart. It also provides the ability to edit visualization, examine the underlying SQL query, and add the chart to a dashboard.
There are six options available for visualization - Line Chart, Bar Chart, Pie Chart, Horizontal Bar Chart, Single Value, and Table. When you choose Pie chart, you will be able to see the distribution of events by segmentation attribute across the selected time range.
The following features are available in this module:
SECTION | DESCRIPTION |
# and % | Allow you to switch between numbers and percentage values on your chart. |
Top 'N' segments | Allows you to choose the number of top segments you want to display in the visualization. |
Time Range and Time Grain | Allows you to configure time range and time grain for the analysis. |
Sampling | Allows you to configure sampling and responses for the analysis. |
Edit visualization (Other actions) | Allows you to inspect a query, edit visualization, add an analysis to a dashboard, and download as CSV. |
Switching between numbers and percentage
These options allow users to switch the values on the Y-axis from absolute values to percentage and vice versa in one click. This feature is particularly useful when dealing with multiple segments, as using absolute values allows users to see a clearer breakdown of how many individuals are in each segment.
Top 'N' Segments
To choose the number of top segments you want to display in your event segmentation, you can click the Show top N drop-down and enter the required number. You can also use the + and - symbols to enter your input. The 'top' attribute values are defined by total count of selected events. You can also set the number of Top 'N' segments when you create beelines.
Time Range and Time Grain section
- At the top of the Visualization window, users will be able to configure the time range and time grain for the analysis. Time Range refers to the complete period of time during which events will be taken into account for the analysis. Examples include the last 2 years or the time range between two specific dates. Time Grain refers to the granularity of analysis, such as Daily (1 day), Weekly (7 days), etc.
- You can set the time range using a simple drop-down or even choose from the quick options and quickly iterate through different choices of time range and time grain without leaving the chart. It is also possible to set a lag by clicking Offset and setting the Ending.
Sampling section
In the visualization module, you can configure sampling modes for your exploration by clicking on the lightning icon on top of the exploration. The drop-down that appears will display 2 sampling modes - Enabled Faster Response or Enabled Higher Precision.
Edit visualization section
Clicking the three dots next to the lightning icon displays a drop-down with 4 options - Inspect Query, Add to Dashboard, Edit visualization, and Download CSV.
When a pie chart is used along with segmentation over multiple cohorts, results may be double counted. This is because a given event or actor may be include within multiple cohorts, whereas a pie chart is designed to show breakdown of the whole into multiple, non-overlapping segments.
Please sign in to leave a comment.