Would a finite return digest make social media less demanding?

I keep wondering whether “catching up” is itself a product problem. After being offline, a platform could offer one finite returning-user digest instead of dropping someone into an endless feed and making them scroll backward until they give up—or lose an evening.

My concrete version would be opt-in and generated on-device from a bounded interval. I would define important as direct replies, mentions, messages from selected people, community announcements, and bookmarked topics. The client would request only the minimum metadata needed to sync, then summarize locally. No engagement bait, read receipts, like or follower counts, popularity rankings, or recommendations unless I explicitly allow them.

That might reduce doomscrolling, but it could also create a new duty to clear the digest. Should “important” be entirely user-defined, and would this make online communities feel less demanding—or merely turn social life into another inbox?

The wording matters as much as the algorithm. “You have 17 important updates” creates a task; “Here are a few things you may have missed” preserves permission to ignore them. I’d make the digest expire quietly rather than show an unread badge forever.

Local ranking is good, but it does not automatically make the system private. The server still sees whatever metadata is needed for synchronization, and summaries can leak sensitive context if cached or backed up. I’d want a clear threat model and a deletion control, not just “on-device” branding.

User-defined importance sounds principled until setup becomes a rules-management hobby. Start with a handful of understandable switches, then let people refine them later. The strongest default may be direct human contact only; recommendations can stay in the main app for people who actually want them.

Forums already had this shape: show me replies since my last visit, then stop. The innovation may be resisting the urge to add a feed, ranking layer, streak, and celebratory confetti to it.

I’d treat reduced scrolling as a hypothesis, not a promised wellbeing benefit. A digest could help some people and burden others, depending on context and how strongly the platform frames participation as required. An experiment should measure voluntary use and abandonment, not just opens.

A protocol detail I’d add: the client should store a signed high-water mark for the bounded interval, so reopening the app cannot silently expand the digest into a backlog. “Finite” needs to be enforced technically, or the product team will eventually call more content helpful.