We are working on a requirement which includes creating an "Admin Panel" to send announcements/push notifications to different user populations of a mobile application for things such as: Application Updates, System Outages, Service Announcements, etc, based on certain user parameters such as roles and other properties. We are currently sending as this user properties to our Firebase project from Android and iOS versions of a mobile application.
The solution of sending push notifications to different user segments and specially the flexibility of segmenting users with relational operators over the user properties is a fundamental requirement for which we though of using Analitycs user segments to target specific notifications dynamically. See source: https://firebase.googleblog.com/2017/03/updates-to-firebase-notifications-reach.html
The problem: Some of the user properties we send to Firebase and we would like to leverage to segment the push notifications are not static and could change in a short time period which may not be reflected in the mobile application immediately, however we did not find a way of updating user properties from the server side (i.e: from a process driven from the data source systems) and relying on users using the mobile app to push/update their user properties updates to Firebase will not suffice to keep the segments up to date and audiences accurate (users may take some time to come back and open the mobile app long after a user property changed).
The questions: Is this limitation of user properties not being manageable through REST endpoints or any other way from the server intentional? Is there any other product or alternative that could help achieving the objective described above? Any recommendations out there?
Notes: We discarded using Firebase topics due to the dynamic nature of the filtering requirements for segments. Maintaining topics up to date with the right population of users would have performance costs and serious scalability issues. AWS Pinpoint was evaluated too although the integration with Firebase doesn't seem to be working very well hence we discarded it.
