Event segmentation

Event Segmentation lets you target audiences based on lifecycle actions or Custom Events that have occurred.

To Segment based on event activity, include an Activity Object in your audience, and an optional Where Object to filter on specific event properties.

Activity audience object

The activity object defines the target audience based on lifecycle actions or Custom Event activity, using comparison operators on activity count and recency. Optionally include a where object, which filters for specific activity values.

In the example to the right, the target audience is users who have opened your app from a notification from the “neowise” campaign at least twice, 3 days ago.

Jump to examples ↓

OBJECT PROPERTIES

activity stringREQUIRED
The target event activity, e.g., app_open.
Default event values are enumerated below. If you create a custom or predefined event that you wish to segment users on, you must register the event in the Airship dashboard. See Manage events.
Possible values: app_open, message_received, message_center_read, message_center_delivered, message_center_deleted, in_app_message_display, in_app_message_resolution, in_app_message_expiration, screen_view, web_session, web_click, short_link_click, first_seen, sms_aborted, sms_delivered, sms_dispatched, sms_expired, sms_failed, sms_rejected, sms_undeliverable, sms_unknown, email_bounce, email_click, email_delay, email_delivered, email_injection, email_open, email_unsubscribe, scene_displayed, scene_completed, scene_incomplete, survey_displayed, survey_submitted, survey_not_submitted
after object
A date value for an absolute comparison or an integer value for a relative comparison.
One of
- string
  A date-time, for example: "after": "2020-08-01T12:00:00".
before object
A date value for an absolute comparison or an integer value for a relative comparison.
One of
- string
  A date-time, for example: "before": "2020-08-01T12:00:00".

metric string

The metric you are targeting for the event.

When using the metric count or total_value, operator and value are required.

When using either first or last, operator and value are not allowed. Use one of the following combinations instead:

before/precision
after/precision
on/on_precision

Metric	Description
`count`	Evaluate based on the number of event occurrences.
`total_value`	Evaluate based on the cumulative event `value` associated with the target user.
`last`	Evaluate based on the last occurrence of an event for a given user/time window.
`first`	Evaluate based on the first occurrence of an event for a given user/time window.

Possible values: count, total_value, last, first

on string
Use on in conjunction with the first or last metric to indicate either a specific date or period with precision. Use either:
- A date in ISO 8601 format, e.g., 2020-08-10T17:28:34+0000 or
- An integer that, when combined with the on_precision property, specifies the time period. For example, use "on": 12 with "on_precision: "month" to target events that occurred in December.
One of
- string
  A date-time.
on_precision string
When using the first or last metric with the on property to target events on. For example, for the 5th of any month, or only in March, use on_precision.
Possible values: day, month, month_day, year_month, year_month_day
operator stringREQUIRED
The comparison operator used to evaluate the activity expression. Use these operators (greater/less than, equal to, equal or greater/equal or less) when the metric is either count or total_value.
Possible values: greater, less, equals, greater_eq, less_eq
precision string
If using a relative number (integer) for the before/after/on property, the precision value will be interpreted as the number of time units ago, e.g., 7 days ago. Defaults to days.
Possible values: days, weeks, months, years
value numberREQUIRED
The value that the operator uses to evaluate the target audience.
where object
The Where Object.

Used in:

Examples

{
   "audience": {
      "activity": "app_open",
      "operator": "greater",
      "value": 2,
      "after": 3,
      "precision": "days",
      "where": {
         "property": "/_triggering_push/campaigns/categories",
         "operator": "equals",
         "value": "neowise"
      }
   }
}

Where object

where is an object that filters on the existence of prior user activity, specified by event properties that you provide.

Jump to examples ↓

OBJECT PROPERTIES

compare_as string
Selects the property type for comparison.
Possible values: text, number, date, boolean
operator stringREQUIRED
The operator in question
Possible values: greater, less, equals, greater_eq, less_eq, contains, present, range, before, after
precision string
Used only for date values.
Possible values: minutes, days, months, years
property stringREQUIRED
The Custom Event property on which to filter for inclusion in the audience. See Event Segmentation Properties for properties reference.
relative_to string
Used only for date values.
Possible values: future, past
value objectREQUIRED
The value of the property you are filtering for in the where object. Use compare_as to select the type for comparison.
One of
- string
  A date-time, e.g., 2020-08-10T17:28:34+0000.

Examples

{
   "audience": {
      "activity": "app_open",
      "metric": "count",
      "operator": "greater",
      "value": 0,
      "where": {
         "property": "/_triggering_push/push_id",
         "operator": "equals",
         "compare_as": "text",
         "value": "636abb88-5642-4035-998d-a04c93c499ad"
      }
   },
   "device_types": [
      "ios", "android"
   ],
   "notification": {
      "alert": "Did you get that thing I sent you?"
   }
}