A platform flagged my account as "not a real person." It was right, and it taught me something about marketing.
Something happened this month that stung more than any bad launch.
A platform I'd been active on started rejecting every post I made. No links, no promotion, just genuine questions rejected anyway. I emailed support expecting a bug. The answer was worse: my account had stopped reading as a human. It read as a channel.
Not banned for anything specific. Just quietly reclassified. And when I went back and counted, I couldn't argue. Over three months I'd posted about my product ~30 times there. Every post individually sincere. All of them, underneath, the same move me talking about my thing, to people I'd never once talked with.
I thought I was building in public. I was broadcasting in public. From the inside those feel identical, which is exactly why I couldn't see it. Every single post felt genuine, so I never questioned the pattern the posts added up to.
The uncomfortable lesson: I optimized every individual post and never once checked the ratio. How many times I showed up for someone else's thing vs. showed up with my own. Mine was indefensible.
What gets me is that a dumb automated filter spotted a pattern in my behavior that I was far too close to see myself. Which is the exact thing I keep claiming tools should do for founders.
So, genuinely asking the room: for those of you who post in communities as part of building what's your actual show-up-for-others vs. post-about-yourself ratio? Do you track it at all, or did you also only find out the hard way?
This hits harder than the platform's decision to flag you. That decision was just the symptom.
The real insight is that algorithms catch patterns humans convince themselves aren't there. Your ratio problem wasn't hidden - it was invisible to you because you were optimizing honestly, locally (each post), and couldn't see the shape of the forest.
The uncomfortable truth is that your accounts read authenticity at the aggregate level, not the post level. One genuine-but-promotional post? That's a person. Thirty? That's a signal. And signal always beats post quality.
Most founders who "build in public" get this backwards. They optimize for individual post authenticity ("this is genuinely my journey") but ignore the rate. The platform wasn't saying "you're fake." It was saying "your behavior pattern is indistinguishable from someone with different intentions."
The show-up ratio is a cheat code most people don't even look at. My guess: most successful IH builders who stay credible aim for something like 70% comments-on-others, 30% on their own thing. Not because it's moral, but because it's the signal that separates "someone building" from "someone promoting who learned the language."
The fact that a filter caught what you couldn't is less an indictment of you and more proof that communities need better visibility into these patterns.
@omri_ben_shoham "authenticity at the aggregate level, not the post level" is exactly it, and that's the part i couldn't see. every post passed on its own so i never zoomed out.
This is a great, honest title — curious what specifically triggered the flag, and what you changed afterward that ended up being a useful marketing lesson?
@Yuki_Code1 no single post triggered it, that's the thing. it was the pattern across three months. what changed after: i stopped posting for a while and just commented on other people's stuff instead.
The three-month pattern point is the part that would have taken me a long time to figure out on my own. It means there's nothing to look at post by post — each one passes on its own, and the thing being judged only exists at a scale you can't see from inside your own feed.
What I'm curious about is which half of the fix actually did the work. Pausing lets the old pattern age out, but commenting on other people's posts adds a genuinely different signal — replies you didn't control, timing you didn't choose, conversations that go somewhere you didn't plan. Those look nothing alike to a classifier. Did it feel like the recovery tracked the pause, or did it turn once the conversations started?
honestly the commenting did more, but only because the pause forced it. couldn't have gotten there without stopping first.
That ordering makes sense, and it's the part that would be easy to get wrong from the outside. Reading it as "comment more" alone would just mean adding activity on top of the same pattern — the pause is what made room for the commenting to register as different rather than as more of the same.
Thanks for spelling it out. This is a more useful answer than the version where one thing turns out to be the fix.
yeah, "comment more" on its own would've just been more noise from the same account. the pause is what made it read differently.
That's the sentence I'll remember — same account, more noise. Makes it obvious why the ordering isn't optional. Thanks for the honest breakdown of it.
The ratio is a better diagnostic than whether each post felt sincere. I made the same mistake, honestly, by treating every useful product mention as something I had to defend on its own, while missing how repetitive the account looked over time. What's helped me is looking at the last 20 interactions like a stranger would, replies on other people's work, questions, and posts about my own product. I wouldn't chase some rigid number, but if the profile still reads like a funnel, the ratio is off.
@ryanshrott "read like a funnel" is a brutal way to put it and yeah, mine did. reading your last 20 interactions like a stranger is a good gut check, i'm stealing that.
The 30 posts about your product in 3 months vs how many comments you left on others' posts tells the story the algorithm saw. Credibility is built through reciprocity, not sincerity. You can be completely genuine while broadcasting - but without the counterbalance of engaging in others' work first, you read as a distribution channel regardless of intent. The algorithm did the math: show-up-for-others ratio = zero. That's the pattern that breaks trust, not the pitch itself.
"credibility is built through reciprocity, not sincerity" hits. i kept defending my intent when intent was never what was being measured.
I’m new to these communities and started with almost no network, so most of what I’ve done so far is respond to other people’s work, with only occasional posts about my own project. I haven’t tracked a fixed ratio, but your point about the visible pattern mattering more than the intention makes me think I should. Did the platform say whether changing that pattern could reverse the classification?
@onlybillboard they didn't say directly, just that the profile needed to read like a person. starting the way you are is honestly the right order, i did it backwards and had to unlearn it.
That makes sense. I’ve just made my first proper project post after spending time responding to other people, so I’ll keep that balance and make sure the account doesn’t turn into a stream of promotion. Thanks — hearing that you had to unlearn the opposite pattern makes the lesson much clearer.
yeah, unlearning it was way harder than just doing it right the first time. you're starting in the better order.
I like that you separated the intention behind each post from the pattern those posts created over time.
I'll be interested to see whether changing that ratio also changes the kinds of conversations you end up having. Those patterns will probably reveal more about community trust than engagement metrics ever could.
same, that's the part i'm curious about too. engagement numbers never told me anything, but who actually replies might.
Appreciate the context.
Those conversations are usually where the stronger signals appear. It’ll be interesting to see what patterns emerge over time.
same, curious whether the replies get less polite once people actually know you.
I'd be interested in comparing notes once you've had a bit more time to observe those patterns. What's the best email to reach you on?
happy to keep comparing notes here honestly, this thread's been more useful than most dms. curious what you end up seeing on your side too.
That's fair. I can respect that. I've enjoyed the conversation either way.