The median indie launch last week got 3 upvotes

Last week I tracked every indie product launch I could find — 6,837 products across 12 platforms, July 20–26. Not a sample. Every one I could reach, checked daily.

Here's the number that stopped me: the median launch got 3 upvotes.

Not 3,000. Not 300. Three.

The full picture, from 5,639 Product Hunt launches with a reported vote count:

651 (11.5%) finished with zero votes
2,297 (40.7%) got two or fewer
5,055 (89.6%) never made it past ten
Only 120 (2.1%) cleared 100

Split into ten equal groups of 564 products: the top decile took 66.6% of all 51,042 votes cast that week. The bottom decile — 564 real products, built by real people, launched on the same day as everyone else — took zero. Combined.

That's not a long tail. That's a cliff with a crowd standing at the bottom of it.

Where everyone went

I track 12 launch platforms. 97.9% of founders posted to exactly one of them — and for 80% of everyone, that one was Product Hunt.

The day mattered too: 1,331 products launched on Monday, 661 on Sunday. Launch on a Monday and you're competing with twice the crowd for the same eyeballs.

What the copy actually says

I ran every launch description through a tagger that extracts who it's for and what it does. Then I grouped products by their full signature.

230 products' entire distinguishing signal was "for developers." Nothing else extractable. 198 were just "works with photos." 134 were just "works with video." And 305 products (4.5%) gave off no usable signal at all — the copy was too vague to say who it's for or what it touches.

If a script reading your launch can only conclude "it's for developers," a human skimming 1,331 launches on a Monday morning will conclude less than that.

It's not just who — it's what they're not telling you. Of the 6,825 products we wrote an actual summary for:

Only 31.5% state a price anywhere in the copy
Only 21.1% say what makes them different
390 products call themselves "an alternative to ___" — and don't name the ___

The part that made me go back and check my last post

Two weeks ago I posted that we'd found 787 tools for developers and 5 for nurses, and a couple people asked whether that was just an artifact of things piling up over time.

So I re-ran the same cut on this one week alone: developers 374, marketers 211, freelancers 202, job seekers 175, students 161 … realtors 13, aviation 10, nurses 5, farmers 4, security 2.

Same shape. One week. Freshly generated, no accumulation. 75 developer tools for every nurse tool — recreated from scratch every seven days.

So what

None of this means "don't launch." It means a launch-day vote count was never a measure of demand — it's a measure of who else launched next to you that day, and how legible your one-liner was to a stranger scrolling fast.

The people who'd actually use a scheduling tool for a salon, or a shift-handover tool for a nursing floor, were never on Product Hunt on a Tuesday morning to vote on it. That's not a them problem. It's a distribution problem, and it's the same one from my last post, just sharper this week: where you have distribution and where the need is unmet are almost never the same place.

I'll keep tracking and post next week's cut. If you've got a specific question you want run against 12 platforms of launch data, drop it below — happy to pull the number.