Build segments by events

Event-based segmentation allows you to target users based on their actual behavior and actions within your app or website. This creates highly personalized and effective campaigns by focusing on user engagement patterns.

Build segments by events

To build a segment by events:

1. Hit Event in the Add filter by section.

2. Select an event from the drop-down list.

3. Indicate conditions for the selected event: the number of times it happened and a time frame.

When specifying a time frame for your segment, you can choose from two categories: Relative dates and Specific dates.

Relative dates

Use these options for dynamically updating time windows based on the current date:

during the last N days	Includes events that occurred within the past N days
exactly N days ago	Targets users who triggered the event exactly N days ago
N or more days ago	Includes users who triggered the event at least N days ago
between N and M days ago	Defines a range of days before today
between N and M days ahead	Defines a range from N to M days from today, where 0 means today. Both N and M are included

Specific dates

Use these options to define fixed calendar-based filters:

on a specific date	Includes users who triggered the event on a selected calendar date
on or after a specific date	Includes users who triggered the event starting from the selected date onward
on or before a specific date	Includes users who triggered the event up to and including the selected date
between dates	Includes users who triggered the event between two specific calendar dates

4. (Optional) Click + Attribute to narrow the condition down to a specific event attribute. Select the attribute from the dropdown, choose an operator, and specify a value. Start typing to search suggested values based on previously recorded data.

Tip

For custom events, add attributes when you create the event. Attribute types are listed in Understanding events. Default events already include their attributes. See Default events.

For the PW_EmailSubscribed / PW_EmailUnsubscribed events, the categories attribute value dropdown shows your account’s actual email categories instead of raw recorded values, if the preference center is enabled.

See event attribute operators by type for the full list of available operators.

5. (Optional) Click + Sum when you need a total across matching events (for example, total spend), not a limit on a single event.

Choose an integer or decimal (price) attribute, an operator, and a threshold. + Sum appears only when the event has at least one integer or decimal (price) attribute. See operators for sums.

Example: Purchase at least once during the last 90 days, with sum of amount greater or equals 100, targets users who spent 100 or more in that window, in one purchase or several.

Attribute filters on the same event still apply: the sum uses only events that match them (for example, total amount where category is "shoes"). You can add more than one sum, on the same or different attributes. A user must pass every sum you add.

How are sums calculated?

• Which events count: Only events that already match this condition’s frequency, time frame, and attribute filters are included in the sum.
• Missing values: If a matching event has no value for the summed attribute, it counts as 0.
• Refunds: Negative values subtract. Track a refund with the same event and a negative amount so the total reflects net spend.
• No matching events: If the frequency is not never, users with zero matching events in the time frame do not pass the sum. For example, sum of amount less or equals 0 does not include users who never purchased. To target non-buyers, set the event to never without a sum.
• Never with a sum: If the frequency is never, the segment includes every user except those whose matching events meet the sum threshold. For example, Purchase never during the last 90 days, with sum of amount greater or equals 100, includes users who spent under 100 in that period and users with no purchases.
• Data retention: By default, Pushwoosh stores event data for one year, so a sum cannot cover a longer period or show lifetime spending.

6. If you set several events as a segment condition, select an operator you’d like to use (AND/OR).

7. Below the filter groups, enable Include all of a user’s devices to target a user across all their devices, not just the one that matched the segment condition. This ensures they receive your messages on every platform they use.

If any of a user’s devices matches the conditions, all their devices are added to the segment. If the checkbox is left unchecked, only matching devices are included.

This same checkbox appears, disabled, on a segment’s read-only detail page — it reflects the setting the segment was created with and can’t be changed there.

Event attribute operators by type

The available operators depend on the type of the event attribute you’re filtering by. Select a type below to see its operators.

String

Operator	Description	Example
is	Attribute value equals the specified value	button_name is "checkout"
is not	Attribute value does not equal the specified value	url is not "/cart"
starts with	Attribute value begins with the specified string	url starts with "https://example.com"
ends with	Attribute value ends with the specified string	url ends with ".html"
contains	Attribute value contains the specified string	url contains "/products/"
in	Attribute value matches any value in a provided list	button_name in ["submit", "cancel"]
not in	Attribute value does not match any value in a provided list	category not in ["spam", "test"]
not set	No value was recorded for the attribute	coupon_code not set
any	Attribute has any recorded value	referrer any

Integer

Operator	Description	Example
is	Attribute value equals the specified number	level_number is 5
is not	Attribute value does not equal the specified number	score is not 0
greater or equals	Attribute value is greater than or equal to the specified number	points_earned greater or equals 100
less or equals	Attribute value is less than or equal to the specified number	attempts less or equals 3
between	Attribute value is within the specified range (inclusive)	level_number between 10 and 20
in	Attribute value matches any value in the list	level_number in [5, 10, 15]
not in	Attribute value does not match any value in the list	attempts not in [0, 1]
not set	No value was recorded for the attribute	score not set
any	Attribute has any recorded value	level_number any

Boolean

Operator	Description	Example
is true	Attribute value is true	enabled is true
is false	Attribute value is false	is_trial is false
not set	No value was recorded for the attribute	enabled not set
any	Attribute has any recorded value (true or false)	is_trial any

Decimal (price)

Use this type for monetary values with decimals. In the Control Panel it appears as decimal (price) when you create or edit an event attribute.

Operator	Description	Example
is	Attribute value equals the specified price	total_price is 9.99
is not	Attribute value does not equal the specified price	total_price is not 0
greater or equals	Attribute value is greater than or equal to a price	total_price greater or equals 50.00
less or equals	Attribute value is less than or equal to a price	total_price less or equals 100.00
between	Attribute value falls within the specified price range (inclusive)	total_price between 10.00 and 50.00
in	Attribute value matches any value in a list	total_price in [9.99, 14.99, 19.99]
not in	Attribute value does not match any value in a list	total_price not in [0, 5]
not set	No value was recorded for the attribute	total_price not set
any	Attribute has any recorded value	total_price any

List

List-type attributes can hold multiple values at once, such as product IDs added to a cart.

Operator	Description	Example
in	Attribute contains at least one of the specified values	product_ids in ["prod-123", "prod-456"]
not in	Attribute does not contain any of the specified values	product_ids not in ["prod-789"]
not set	No value was recorded for the attribute	product_ids not set
any	Attribute has any recorded value	product_ids any

Date

Date-type attributes support the same relative and specific date comparisons used for the event’s own time frame.

Operators for sums

A sum condition compares the total of an attribute across the matched events, so it supports fewer operators than an attribute condition.

Operator	Description	Example
is	The total equals the specified number	total_price is 100
is not	The total does not equal the specified number	total_price is not 0
greater or equals	The total is greater than or equal to the specified number	total_price greater or equals 100
less or equals	The total is less than or equal to the specified number	total_price less or equals 500
between	The total is within the specified range (inclusive)	total_price between 100 and 500

Tip

Sums keep the sign of each value. To account for refunds, track them with the same event and a negative amount, so that the sum reflects the net amount spent.

Event-based segmentation examples

Below are several practical scenarios illustrating how you can use event filters to create meaningful audience segments.

Frequent purchasers

Goal: Target highly engaged users who have made multiple purchases recently.

How to set up:

• Event: Purchase_made
• Condition: exactly or more
• Times: 3
• Time frame: during the last N days
• Time frame value: 30

Use case: Send special offers or loyalty rewards to users who demonstrate frequent buying activity.

High-value single purchase

Goal: Target users who made a large purchase, regardless of how often they buy.

How to set up:

• Event: Purchase_made
• Condition: exactly or more
• Times: 1
• Time frame: during the last N days
• Time frame value: 30
• Attribute: total_price
• Operator: greater or equals
• Attribute value: 50

Use case: Invite big spenders to a loyalty or VIP program right after a qualifying purchase.

Total spend over a period

Goal: Target users whose combined purchases reach a spending threshold, even if no single purchase does.

How to set up:

• Event: Purchase_made
• Condition: exactly or more
• Times: 1
• Time frame: during the last N days
• Time frame value: 90
• Sum: total_price
• Operator: greater or equals
• Sum value: 200

Use case: Move customers into a higher loyalty tier once their spending over the last 90 days reaches the threshold.

Refunds exceed purchases

Goal: Target users who got back more than they spent over a period. This works when refunds are tracked with the same event and a negative amount.

How to set up:

• Event: Purchase_made
• Condition: exactly or more
• Times: 1
• Time frame: during the last N days
• Time frame value: 90
• Sum: total_price
• Operator: less or equals
• Sum value: -1

Use case: Route these users to a support or win-back flow instead of promotional campaigns.

Viewed products but no purchase

Goal: Identify users who have shown interest but have not converted.

How to set up:

• Event 1: Product_Viewed
  • Condition: exactly or more
  • Times: 1
  • Time frame: during the last N days
  • Time frame value: 7
• Event 2: Purchase_made
  • Condition: never
  • Time frame: during the last N days
  • Time frame value: 7
• Operator: AND

Use case: Send reminders, special offers, or personalized messages to users who browsed products but haven’t made a purchase yet.

Users engaged during a campaign

Goal: Target users who participated during a specific promotional period.

How to set up:

• Event: Promotion_Click
• Condition: exactly or more
• Times: 1
• Time frame: between dates
• Time frame value: "2024-01-01" and "2024-01-31"

Use case: Follow up with users who interacted with a past campaign to promote a related offer or gather feedback.

High-value actions over time

Goal: Identify users who performed a valuable event a certain number of times within a custom date range.

How to set up:

• Event: Subscription_Renewal
• Condition: exactly or more
• Times: 2
• Time frame: between dates
• Time frame value: "2024-01-01" and "2024-12-31"

Use case: Offer special recognition or advanced features to loyal subscribers who regularly renew their subscriptions.

Tips for event-based segmentation

• Use relative dates for ongoing campaigns that should stay updated with current user activity.
• Narrow an event condition down with an event attribute filter (e.g., purchase amount, product category) instead of relying on event occurrence alone.
• Combine event filters with other criteria such as tags (e.g., user demographics, subscription status) or existing segments for more precise targeting. For example, you can filter users based on both event occurrence and user properties stored in tags to tailor your messaging even further.
• Periodically review and update your segment filters to adapt to changing user behaviors or business goals.