What are the big design questions you’re wrestling with?
Yes, but it tends to win on a different axis. Products that avoid attention traps have to be unmistakably useful or meaningfully enjoyable in short, self-contained sessions. In my experience, when the utility is strong and the craft shows up as speed, clarity, and fit to context, people come back out of their own self-interest, not because the interface nags them. The early Facebook push was framed around daily relevance, not dopamine: make it easy to find the people and information that matter, keep it fast, and reduce friction in the core loop. The health metric was engagement, but the durable driver was whether the product felt worth returning to today.
The trap is confusing “no dark patterns” with “no opinion.” Opinionated software that sets expectations, limits scope, and tells the truth about what matters right now can feel calmer and still be sticky. Personalization helps here when it reduces noise rather than amplifies it: fewer, better notifications; context-aware surfaces that respect time and place; and graceful endings that let a session conclude cleanly. That sort of design builds trust over time, which is its own moat when the novelty of growth hacks fades.
I’ve also seen that selection happens by audience and moment. Some categories reward endless feeds; others reward getting someone to done and getting out of the way. If the product’s value compounds with identity, relationships, or workflow fit, sustainable engagement follows without theatrics.
So the useful lens isn’t “can it succeed without hooks?” but “what are people actually selecting on here: raw time spent, or outcomes they feel good repeating?”
I share that read. Once the business model selects for raw time, product decisions start orbiting whatever expands the feed. In the early era, the center of gravity felt narrower: help people find their real friends, share daily life with finite audiences, make it fast and private enough to feel safe. Engagement mattered, but mostly as a check on relevance rather than a mandate to maximize minutes. The trouble is that proxies harden into goals; what starts as a sanity metric becomes the scoreboard.
If the aim is relationship-building, the selection axis shifts. Strong signals tend to be durability of ties, reciprocity, and the ease of resuming context after a pause. That nudges the design toward finite sessions, graceful endings, fewer-but-better notifications, and identity that privileges closeness over reach. It also invites a different revenue story: when value accrues to the relationship, subscriptions or lightweight payments make more sense than ads that reward breadth. The feel is closer to a well-run group thread than a public stage.
There’s also a governance layer that’s easy to skip in v1. Small choices around defaults, visibility, and controls telegraph intent. Early Facebook reacted quickly when people felt exposed around News Feed, adding controls within days; that responsiveness built trust even as the product changed fast . A modern take would treat trust as part of the core loop, not a settings afterthought.
A useful lens here is to ask what people are actually selecting on when they return: proximity and meaning with specific others, or ambient novelty from the crowd. If it’s the former, limiting scope isn’t a handicap; it becomes the promise.
I do. The desire shift you’re describing is real: as life gets fuller, the selection axis tilts from novelty and reach to proximity and ease. In the early Facebook rhythm, the thing that worked wasn’t infinite time on site; it was that one could reliably find the right people and feel safe sharing ordinary life without performance pressure . That promise still resonates for Millennials who don’t want to maintain a public stage every day.
Two caveats feel material. Audience size and network shape will look different than teen-centric platforms; depth and reciprocity beat breadth, so growth likely rides on small clusters and latent graphs that already exist. And the business model needs to rhyme with the intent. When the value accrues to the relationship, lightweight subs or paid utilities tend to keep incentives clean, whereas ads tug the product back toward scale and spectacle.
If this is worth trying, it’s because there’s room for software that ends sessions gracefully and helps people resume context with the few who matter. The lens I keep coming back to: are returns driven by closeness with specific others, and can that feel good to repeat without asking for more time than life wants to give?