Originally published on LetsLaunch.
Every maker asks this the night before a launch, usually at 1am, usually in a
Slack DM to someone who has done it once. "How many upvotes do I actually
need?" The honest answer, for most of the boards makers ask this about, is that
nobody outside the company can tell you — and most of what gets posted as an
answer is a guess dressed up as a number.
There are really two different questions hiding inside that one. The first is
"how many upvotes get me to the top of an opaque, algorithmically-ranked
board" — and the honest answer is that nobody outside that company knows,
because the formula isn't published. The second is "how many upvotes keep my
listing up at all on a board that actually discloses its rule" — and that
question has a real, checkable answer, at least for the one board we can speak
to directly: this one.
Why the real number for Product Hunt doesn't exist
Product Hunt does not publish its ranking algorithm, and anyone quoting you an
exact vote-count threshold for the front page is guessing. What's publicly
observable, from makers comparing notes over the years rather than from any
official documentation, is roughly: raw vote count matters less than vote
velocity early in the day, comments and engagement seem to carry real weight
alongside votes, and the ranking clearly decays over the 24-hour window rather
than just summing votes at the end. None of that is official. It's pattern-
matching from the outside, and it should be treated that way — useful for
intuition, useless as a target number.
That opacity is deliberate, and for a reasonable reason: a board that
published its exact formula would hand vote-ring operators a spec sheet. Keep
the formula secret and gaming it gets expensive and unreliable. The tradeoff is
that a maker planning a launch has nothing solid to plan against — just
folklore.
The one board that actually tells you the number
We can't tell you Product Hunt's threshold. We can tell you ours, because we
set it and there is no reason to hide it: LetsLaunch requires
5 upvotes within 24 hours of launch for a listing to survive. Clear the bar
and the listing is permanent — dofollow link included, no expiry. Miss it and
the listing is soft-removed: hidden from browse, the page goes noindex, and
the outbound link switches to nofollow. One free relaunch is available
afterward. The mechanism is described in full on how it works.
We're disclosing that specific number, rather than keeping it vague the way
most boards do, because an opaque threshold optimizes for one thing: making
the platform impossible to game. A disclosed threshold optimizes for something
we think is more useful to a maker — being able to actually plan a launch day
around a real number instead of a folklore one.
One honest caveat, stated plainly rather than buried: that 5-vote rule is
currently in a cold-start grace period. Nothing gets soft-removed today — the
enforcement switch stays off until the board sustains a steady flow of real
daily launches, so a thin early board doesn't punish the makers who show up
first. The number and the window are real and exact; whether a listing is
actually being held to them, right now, is not yet — and we'd rather say that
plainly than let the disclosed rule imply an enforcement that isn't switched
on yet.
There's a second, quieter number behind that first one: votes from accounts
younger than 15 minutes don't count toward the threshold. They still show up
on the page — nothing is hidden — they just don't move you toward survival.
That's aimed at the same problem the opaque boards solve by hiding their
formula: a five-vote bar is trivial to fake with five fresh accounts unless
faking it costs something.
"Rank #1 for a day" and "survive" are answering different questions
It's worth being explicit that these two numbers aren't measuring the same
thing, because conflating them is where a lot of the folklore comes from.
Ranking on an opaque board is a relative contest — you're being compared
against everyone else who launched that day, on a curve nobody outside the
company can see, and the reward is a spike of attention concentrated into
roughly 24 hours. A disclosed survival threshold is an absolute bar — you're
being measured against a fixed number, not against whoever else happened to
launch on the same Tuesday, and the reward for clearing it is a permanent
outcome rather than a one-day spike: the listing and its dofollow link don't
expire once the 24 hours end.
That's a genuinely different thing to plan for. A maker chasing a Product
Hunt ranking is, whether they think of it this way or not, competing against
an unknown and shifting field — the same five real votes that would comfortably
clear a fixed bar might not move the needle at all against a day when three
well-networked founders happen to launch alongside you. A fixed, disclosed
threshold removes that variable entirely: the bar doesn't move because someone
else had a bigger launch the same day. It's a lower ceiling in exchange for a
number you can actually plan against.
Neither shape is wrong — a relative contest is what makes a single big spike
possible in the first place, and that spike is real and worth having. But
they answer different questions, and "how many upvotes do I need" only has a
stable answer for the second one.
What actually moves votes, on any board
Whether the board is opaque or disclosed, the things that reliably produce
real votes are the same, and none of them are secret:
Your own network, reached honestly. Not a bot list, not a
pay-for-upvotes service — the people who already know what you built and will
say something real in the comments if you ask them to look at launch day
specifically. A board's algorithm, published or not, tends to reward genuine
engagement (comments, replies, time spent) more than a bare vote, so a
network that actually shows up and talks does more than the same number of
silent clicks.
Timing that matches when your audience is awake. A launch that goes live
at 3am in your own timezone loses its best hours to nobody being online. On a
fixed-window rule specifically, this matters twice over: the clock only runs
for a set number of hours, so hours spent with your own audience asleep are
hours the window doesn't get back. Match the launch window to when the people
you're trying to reach are actually looking at their phone, not to whatever
time zone the platform defaults to.
A comment thread worth reading. On boards where comments carry weight —
and most do, disclosed or not — a maker who answers every question in the
first few hours, honestly, including the skeptical ones, gets more real
engagement than one who posts and disappears.
Sequencing rather than a single shot. Nothing says a product gets exactly
one launch-day attempt across every board that exists. A quieter board first,
to work out the pitch and catch obvious feedback, then the bigger attempt with
a sharper story — see our Product Hunt alternatives roundup
for the actual list of what else exists and how each one's mechanic differs.
What reliably backfires, on every board we're aware of, disclosed or opaque:
vote-buying services and vote rings. Boards that hide their formula are
usually hiding it specifically to make these harder to profit from, and boards
that disclose a threshold — like the account-age grace period above — tend to
build in a specific, narrow defense against exactly this. Either way, the
juice-to-squeeze ratio on a bought vote is bad: real accounts age past 15
minutes and get discounted for nothing, and a comments section full of vote-
ring accounts with no real engagement is itself a signal, on any board that
weighs comments.
What a disclosed rule actually looks like in practice
It's worth walking through the whole mechanic once, because "5 votes in 24
hours" is only the headline number, and the parts around it matter more than
the number itself for anyone actually planning a launch.
The clock doesn't start at submission. A listing published through
LetsLaunch first needs the reciprocal badge found on the maker's own site —
there's a 72-hour grace window to install it — and the 24-hour survival
window only opens once the badge is verified and the listing actually goes
live. That separation matters: a maker who submits on a Friday and gets the
badge live on Monday isn't burning survival hours while they're still setting
up the badge. The clock and the badge are two different gates, not one.
Once the window is running, there's a checkpoint at the halfway mark: 12
hours in, a listing that's short on votes gets a warning email rather than
finding out at the deadline that it missed. That's a small thing, but it's the
difference between "here's a heads-up, you have half a day left to ask a few
more people" and silence until the removal notice — and it's only possible
because the threshold is a real number rather than an opaque score nobody can
check against.
Missing the bar isn't the end of the listing, either. A soft-removed page
stays reachable, just hidden from browse and search, with the outbound link
switched to nofollow — and there's one free relaunch on the table, open for
14 days after the miss, specifically so a maker who mistimed a launch (wrong
day, wrong hour, sick that week) gets a real second attempt rather than a
permanent penalty for bad timing. None of that changes the headline number.
It's the difference between a threshold that exists to punish and one that
exists to be genuinely checkable — you can see exactly where you stand at
every stage, not just at the very end.
Frequently asked questions
Does Product Hunt publish an exact vote threshold to reach the front page?
No. There is no official number, and there has never been one published by the
company. Anyone quoting you a specific vote count as "the" threshold is
extrapolating from pattern-watching, not citing documentation — treat it as
folklore, not as a target.
Why does LetsLaunch disclose its exact upvote threshold when other boards don't?
Because hiding a threshold mainly protects the platform against gaming, and we
think a maker being able to plan a real launch day against a real number is
worth more than that protection. The tradeoff is real — a disclosed number is
easier to reverse-engineer around — which is part of why the 15-minute
account-age grace period exists alongside it.
Is LetsLaunch's 5-upvote survival rule actually being enforced right now?
Not yet. The threshold and the 24-hour window are exact and real, but
enforcement is paused during a cold-start grace period so an early, thin board
doesn't punish the first makers who launch on it. Nothing is being
soft-removed today under this rule — a listing that would technically miss the
bar right now stays exactly where it is, dofollow link included, until the
switch is turned on once the board sustains a steady flow of daily launches.
What happens if a listing misses the 5-vote bar once enforcement is on?
The page doesn't disappear. It gets hidden from browse and search, its
noindex tag goes up, and the outbound link switches from dofollow to
nofollow — but the listing itself stays reachable by direct link, and there's
one free relaunch available for 14 days afterward for a maker who thinks the
timing was the real problem rather than the product.
Do comments matter as much as votes on a launch board?
On boards where the algorithm rewards engagement rather than raw vote count —
which is most of them, based on what's publicly observable — a comment
thread with real back-and-forth appears to carry real weight, not just the
tally at the bottom of the page. Treat the comments section as part of the
pitch, not an afterthought.
Do vote-buying services actually work?
Rarely, and the risk usually outweighs the benefit. Boards that discount
fresh accounts or weigh engagement over raw count are specifically built to
blunt exactly this, and a comments section with no real engagement under a
suspicious vote spike is itself a signal on any board that looks at more than
the final number.
If you'd rather launch somewhere the rule is written down instead of guessed
at, submit your product — the full mechanic, badge included, is on
how it works, and the wider directory index
covers everywhere else worth trying too.
Top comments (0)