Building a quieter social app on purpose — what we refuse to optimize for

The useful test for a calm product is not whether people stay longer, but whether they can tell why they came back. A mood or intention prompt can be good friction, but it needs a clear escape hatch when someone only wants to read, otherwise the product starts performing its values at the user. I’d also watch for the empty-room problem: a quiet feed feels intentional only when there is enough signal to make leaving feel like a choice rather than a lack of inventory.

Refusing to optimize for time-on-site sounds clean until you realise it removes the signal you'd normally rank with, and then you have to answer "ranked by what?" — which is the actual architecture question here, not a values question.

Most feeds rank with a proxy for engagement (dwell, taps, replies). If you throw those out, the honest remaining signal is explicit intent — which is exactly what your mood + intention step captures at compose time. So the friction isn't just a filter for quality, it's the input tensor for the Sacred Feed. Two consequences worth deciding early: (1) intent at post time is a cold-start signal, so a new arrival with no history can't be ranked by it — what does discovery show that first session? (2) intent decays, so a post written in a "grief" mood probably shouldn't rank the same 3 days later; is the ranking keyed on the post's intent or the reader's current mood?

The refusals are the easy part to state and the hard part to hold: the moment calm reads as empty, someone will ask to "just add a little recency boost and a popularity sort", and each of those quietly reintroduces the engagement loop you refused.

Founder disclosure: I work on Piramyd, which is why the "don't optimize for attention" line is the part I read with a stake.

Genuine question: is the Sacred Feed currently ranked by post intent, by the reader's declared mood, or by a blend — and which one did early users actually respond to?

Making the first share take a minute on purpose is a brave call. I'd just make the "why" visible in that moment (one line like "this slows things down so the feed stays calm") so it reads as a feature, not friction. Building a small calm toy app myself, I've found a clear "done" moment beats any streak mechanic — giving people a natural place to stop and feel finished might be the calm-app version of retention.

The anti-engagement stance is a clear position, but it creates a business-model tension you should name before someone else does. If you don't optimize for time spent, your paid tier has to sell depth (private circles, guided reflection, teacher-led spaces), not more access. The mood-and-intention step before posting is a good filter, so keep it. But measure how many people drop off at it, and whether those who get through post again within a week.

3 questions an investor will push on:

Without engagement metrics, which number proves it's working: weekly return rate, paid conversion?
Who is the core user, and where do they gather today?
How do you moderate "spiritual" content without becoming a wellness-misinformation feed?
If useful, I'm the founder of a tool that simulates investor questioning — happy to share the full breakdown (free beta)

Intention-led first share is a clear bar. One place onboarding still feels like regular social is asking for a phone or email before someone has shared anything. I'd leave contact fields until after the first intention post, and store phones in E.164 if you collect them later.

The empty-room problem is worse for calm apps than noisy ones, and that's the part I'd watch closest. A noisy app can fill a dead feed with algorithmic filler; you can't, without betraying the whole point. So early on, "quiet" reads as "dead" to a new arrival, and no amount of intention-setting fixes that first impression.

The question I'd want answered before anything else: what brings someone back tomorrow? If the answer is the reflection prompt, measure return rate for lurkers separately from posters. Calm communities don't die because people stop posting — they die because lurkers open the app, find three posts from Tuesday, and never build the habit.

“Leave without feeling drained” is a better north star than session length, but it needs its own signal.....

refusing to optimize is a strategy most founders can't afford psychologically - every dashboard screams for more. we made a similar call on swapfile.live: no accounts, no tracking on tool pages, files never leave the browser. growth is slower but every user arrives trusting you. what did you refuse that hurt the most?

Have early users behaved differently because of the intentional friction—more thoughtful posts, shorter sessions, or stronger return behavior—or is that still unvalidated?

Nice, this makes a lot of sense. What's been the most surprising part of it so far?