Use funnels to view all events after an initial event
You can use a funnel to perform an open-ended path analysis and view all events that a group of actors or a particular actor took after an initial event. For example, one use case we've solved using this strategy is understanding all of the events a new user generates within their first week using a service.
Create a derived column that always returns 1
Use the following code to create a derived column that always returns 1.
long always1() { return 1; }
Build the funnel for the dataset and actor you want to understand
Create a funnel for the dataset and actor combination you're trying to understand.
- Set the entry conditions to use to start collecting events. For example, if you wanted to analyze the events a user takes after creating an account, set the filter to
event_type is one of account_create
. Or in the illustration, start when the user goes to the Register page. - Set the second step of the funnel to filter to Always1 equals 0. Since Always1 is always 1, this step will never be fulfilled.
- For a typical funnel, the funnel is considered complete when the requirements for all steps are fulfilled. To perform this particular path analysis, we want the condition to remain unfulfilled.
- In the Funnel Settings, set All events happen within to the amount of time you want to collect events. Don't set this to be too long of a duration, because queries can get very expensive.
- Save the funnel.
Use the funnel to explore your data
In the Explorer, use the funnel to explore your data by setting the filter to <funnel_name>_current_state is equal to 1
to filter to all events that happened during the funnel period.
Experiment with additional filters to understand the actions of certain groups of your users within the period you have specified.