-
Understanding the Duration of Membership in Predictive Audiences with Firebase/Google Analytics
Hey there! So I’ve been toying around with Firebase/Google Analytics Predictive Audiences, and honestly, some of the terms have me scratching my head. Specifically, I’m trying to figure out what they mean by Membership Duration. Apparently, you can set a certain number or just go for max limit when setting up your predictive audience.
The official docs say something like this:
“Membership duration: Enter the number of days that users stay in the audience (1 – 540). If a user does something that matches the audience criteria, the clock on their membership duration resets.”
So, what does this all mean for predictive audiences?
Let’s say I want to make a predictive audience of users likely to churn in the next week, and I set the membership duration to a month. So, does that mean a user predicted to churn will hang in that audience for the next month? What happens if I choose to max out the limit?
Plus, according to the official docs:
“Predictive metrics for each eligible model will be updated daily for each active user.”
So, what if I set the membership duration to a month, a user gets predicted to churn, and gets placed in that audience. The following day, the prediction changes and they aren’t likely to churn. Do they get booted from the audience, or do they stick around because of the month-long membership duration?
Any Firebase experts in the house that can help me out? The Google guide seems a bit thin on info on this front.
I’ve tried going through the official docs and a whole lot of blogs for more clarity, but it’s like they’re all singing from the same hymn sheet without adding much else.
-
2 Replies
-
The membership duration in Firebase’s Predictive Audiences refers to the number of days that users will remain in that particular predictive audience. For example, if you set the duration for one month, then users predicted to churn would stay in that audience for a month. If you opt for the ‘max limit’, users will stay in the audience for the maximum duration allowed by Firebase, which is 540 days. However, each day Firebase updates its predictive metrics for active users. If a user initially predicted to churn does something that changes this prediction, they will still remain in the predictive audience until the end of the set membership duration. Despite the daily updates, the user will not be removed midway through the duration. It’s a way of not losing track of users due to sporadic changes in behavior or prediction models. So even if the prediction changes the next day, the user will continue being a part of the audience for that month (given a month-long membership duration) or until the end of the duration you set.
-
In Firebase Analytics, “Membership Duration” refers to the number of days a user remains a part of a particular audience after they have fulfilled the conditions to be in that audience. When setting a “Membership Duration”, you can determine a specific timeframe for how long a user can be part of an audience. For instance, if you set a “Membership Duration” of one month, then a user who fulfills the conditions will stay in that group for a month, regardless of any activity or lack thereof during this period. If during this period the user does something that fulfills the requirements, their “Membership Duration” will reset. Setting the “Membership Duration” to the maximum limit means they will remain in the audience for as long as possible, in this case 540 days. If the predictive metrics of a user change daily and they are no longer predicted to churn, they will not be removed from the audience immediately. They will remain in that group for the duration that had been set, which in your example would be a month.
Log in to reply.