Push open rates have been declining for five years. A memories feed works on a fundamentally different mechanism.
The average app's push notification open rate sits below 5%. On iOS, where users can opt out entirely, it's often lower. Growth teams spend significant effort optimising notification copy, send times, and segmentation. The returns keep shrinking.
A memories feature works on a different mechanism entirely. Understanding why helps explain why the retention lift is so consistent.
Notifications are interruptions. Memories are destinations.
When you send a push notification, you're asking a user to stop what they're doing, switch context, and open your app to see something. Even when the notification is relevant, it's friction.
A memories feed inverts this. Users open the app because they want to see something: their own captured moments, content they've created or saved. The motivation is internal.
Personalisation without the ML overhead
A memories feed is inherently personal without recommendation infrastructure. The content is the user's own moments. You don't need to predict what they want to see because you already know: they want to see what they captured.
The habit loop it creates
Apps with strong Day 30 retention tend to create a habit loop: the user opens the app, gets something worth having, and leaves with a sense of completion that builds the expectation of returning.
Users without content to return to churn at the same rate as before. Users with moments they captured and can come back to churn significantly less.
Where push still fits
This isn't an argument against push notifications. The point is that a memories feed addresses a different problem: it creates intrinsic motivation to return, rather than relying on external prompts after users have already drifted.
FyreMemories is a production-ready memories SDK for Android. One dependency, one afternoon to integrate.