You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
| Branches | Configure up to 4 event branches, each tied to a specific event and optional event property filters. <br> Events must share a unique identifier with the entry event if the journey allows re-entry. <br> Branches must be mutually exclusive to avoid validation errors. |
34
+
| Branches | Configure up to 9 event branches, each tied to a specific event and optional event property filters. <br> Events must share a unique identifier with the entry event if the journey allows re-entry. <br> Branches must be mutually exclusive to avoid validation errors. |
35
35
| Filters | Event properties refine the triggering conditions for a branch. |
36
36
| Maximum hold duration | The fallback branch activates after the hold period, ranging from 5 minutes to 182 days (about 6 months) |
37
37
@@ -136,38 +136,45 @@ To configure the Delay step:
136
136
137
137
## Data split
138
138
139
-
The **Data split** step sends profiles down different branches based on audience membership or profile traits. This lets you personalize how users move through a journey, like sending different messages to new users instead of returning customers, or targeting re-engagement campaigns based on inactivity.
139
+
The **Data split** step sends profiles down different branches based on event property values, audience membership, or profile traits. This lets you personalize how users move through a journey, like:
140
140
141
-
Data split is useful when you want to take different actions based on what you already know about the user, rather than waiting for a new event. For example, you might use it to separate users who haven’t purchased in 30 days from those who lapsed 90 days ago, or from users who are still actively engaged.
141
+
1. Sending different messages to new users instead of returning customers.
142
+
2. Personalizing campaigns based on user membership tier.
143
+
3. Sending different discount codes based on a user's total cart value.
144
+
145
+
Data split is useful when you want to take different actions based on what you already know about the user or event properties, rather than waiting for a new event. For example, you might use it to separate users who haven’t purchased in 30 days from those who lapsed 90 days ago, or from users who are still actively engaged.
142
146
143
147
### How Data split works
144
148
145
149
When a profile reaches a Data split step:
146
150
147
-
1. Segment checks whether the profile matches the first branch’s conditions.
151
+
1. Segment checks whether the event properties or profile data matches the first branch’s conditions.
148
152
2. If not, it checks the next branch, and so on, in the order shown in the journey.
149
-
3. The profile moves down the first branch it qualifies for. Each profile can only follow one branch.
153
+
3. The journey instance moves down the first branch it qualifies for. Each profile can only follow one branch.
154
+
4. If none of the logic branches resolve to true, it goes down the everyone else branch.
150
155
151
156
### Configuration options
152
157
153
-
You can configure up to five branches in a Data split step. Each branch can have one or more conditions:
158
+
You can configure up to nine branches in a Data split step. Each branch can have one or more conditions:
| With trait | The profile includes a specific trait and value. |
163
+
| Without trait | The profile does not include a specific trait. |
164
+
| Part of audience | The profile is a member of a selected audience at the time of evaluation. |
165
+
| Not part of audience | The profile is not a member of a selected audience. |
166
+
| With journey context | The event properties in context (from triggering or hold until events) match conditions. |
161
167
162
168
You can also give branches uniques name to differentiate them from each other on the journey canvas.
163
169
164
170
> info "Evaluation is sequential"
165
171
> Segment evaluates branches in the order they appear in the configuration side sheet. If a profile qualifies for multiple branches, Segment sends it down the first one it matches. Profiles can't qualify for more than one branch, and Segment doesn't wait for audience membership to update after the profile enters the step. You can change the evaluation order by dragging branches up or down in the configuration side sheet.
166
172
167
-
### Example: Target different customer types
173
+
### Example: Target different customer types or event properties
168
174
169
-
You can use a Data split to branch profiles based on traits or audience membership that already exist on the profile when it reaches this step. For example:
175
+
You can use a Data split to branch profiles based on event properties, traits, or audience membership that already exist on the profile when it reaches this step. For example:
170
176
177
+
- Journey instances where the triggering event had a `transaction_total` > $100 are sent specific messaging about their high-ticket purchase.
171
178
- Profiles with a known `email_subscription_status` trait get treated as existing customers.
172
179
- Profiles that belong to a `VIP` audience are routed down a separate path for high-value users.
173
180
- Profiles with a specific set of traits (like favorite color and a known name) can receive personalized messaging.
0 commit comments