Build-in-public advice usually comes down to one word: more. Post more. Ship in public. Tweet every commit. Show the work.
I believed it for a while. Then I watched my own feed and everyone else's turn into noise. A founder posts "just fixed a typo" with the same fanfare as "we hit 10,000 users." The reader can't tell which one matters, so they stop reading both. The problem with "share everything" isn't effort. It's that it destroys the signal you were trying to send.
The actual skill in building in public isn't posting. It's judgment — knowing which of the forty things you did this week is the one worth telling the world about. That judgment is hard, it's tiring, and most founders are too close to their own work to have it. Your biggest launch and your smallest bugfix both feel enormous at 1am when you shipped them.
So we built the judgment into the product. It's called FoxSense, and it's the part of FoxPlug I'm proudest of.
FoxSense is a moment classifier. It watches what actually happens in your work — a feature ships, a new person joins the team, someone writes about you, your backer count crosses a round number, you rewrite your landing page — and for each one it answers a single question: is this worth telling the world, and if so, how loud?
The important words there are "how loud." Most tools treat everything as a binary: post it or don't. Real momentum has volume. A genuine milestone deserves a video and a proper post across your channels. A solid feature deserves a short post. A minor internal change deserves a line in your changelog and nothing more. Getting the volume wrong in either direction hurts you — shout about a typo and you look desperate; whisper about your best month and you've wasted it.
Here's the part that took the longest to get right: FoxSense scores everything relative to you, not to some global idea of "big." Your fortieth shipped feature is not news. Your first paying customer is. A thousand-backer milestone is enormous for a new project and routine for an established one. A generic threshold — "post when you hit 100 of anything" — is worse than useless, because it fires for the wrong people at the wrong time. So FoxSense learns your baseline first. It sits quiet through a calibration period, watches what normal looks like for your project specifically, and only then starts telling you when something breaks out of your normal.
That means the same event produces different verdicts for different founders, which is exactly right. The system isn't asking "is this objectively impressive?" It's asking "is this a big moment for you?" — and that's the only question that matches how your audience actually experiences your story.
It also means the quiet answer is a first-class answer. Most of what FoxSense sees, it tells you to sit on. That sounds like a strange thing for a marketing tool to do. It's the whole point. A tool that only ever says "post this!" is a tool you learn to ignore. A tool that mostly stays quiet and then taps you on the shoulder — this one, this is the week's real story — is a tool you trust. Trust is the entire product.
And when it does tap you, it shows its work. Every moment comes with a plain reason: worth a video and a post because it's your biggest launch in three months. Or: a short post is plenty — good, but you've shipped four of these this month. You can see the reasoning, and you can overrule it. If FoxSense says "small" and you know this one matters because a specific customer has been waiting for it, you promote it in one click and it stays promoted. The machine has judgment; you have the final say. That's the correct division of labor between a founder and a tool, and most tools get it backwards — they either do nothing or they take the wheel.
I built FoxSense because I was the target user. I ship constantly and I'm terrible at knowing what's worth saying out loud, because it all feels urgent from the inside. I wanted something that stood a little further back than I could, saw my work against my own history, and told me the truth: this one's big, that one's noise, save your voice for the moments that earn it.
Build-in-public was never supposed to mean narrate-everything. It was supposed to mean show the real work, told well. Knowing which work, and how well, is the hard part. Now it's the part the software does.
Top comments (0)