A Product Hunt Launch, Hour by Hour
What actually decides your rank on a Product Hunt launch day, hour by hour, and the moves that change where you land.
Somewhere around 3am Pacific, a founder in Austin or Manila or Lisbon is refreshing a leaderboard that resets every midnight, watching a number tick up by ones.
A Product Hunt launch is one calendar day, Pacific time, where a product gets submitted once, ranked live against everything else that submitted the same day, and judged almost entirely on what happens in the next twenty hours.
That number decides more than it should. A good product with a quiet comment thread finishes forty-second. A rougher one with a coordinated morning finishes fourth, picks up a badge, and inherits a few thousand visitors who never would have found it otherwise.
Treat the day like a single tweet posted at 9am and it reads that way: forgettable. Treat it like an operation with a schedule and a script, and the same twenty hours can produce a badge and a spike of signups. The backlink outlives the day itself.
What Decides Rank on a Product Hunt Launch Day
Product Hunt doesn't rank by raw upvote count anymore. It ranks by points, calculated from upvotes and comments, weighted by signals the company doesn't fully disclose, specifically so the system is harder to game, per its own help center. A product sitting on 300 upvotes can outrank one sitting on 340 if the lower number came with real comment activity behind it.
The company is blunt about why the formula stays private: "we want the homepage to remain an authentic place for makers," its official launch guide says, and it declines to publish the exact weighting.
The clock itself has a quirk worth planning around. For roughly the first four hours after the midnight Pacific reset, the homepage sorts close to randomly rather than by points, giving every submission that day a shot at early visibility regardless of who hit publish first. Real ranking kicks in after that and stays in effect for the rest of the day.
A more recent shift matters just as much. Point models are now believed to weight average engagement per hour across the full window, not just the first couple of hours.
A launch that dumps its entire supporter list in one spike at minute fourteen, then goes quiet for the next eighteen hours, ranks worse than one that spreads the same number of upvotes evenly, according to a 2026 analysis of leaderboard thresholds. Front-loading was the old advice. Not anymore. Pacing is the current one.
The action here is scheduling, not hustle. Before midnight, split your supporter list into three or four waves instead of one blast, and set reminders to send each wave at a different hour of the day rather than all at once.
The Week Before: What Has to Exist by 11:59 PM
Everything that happens on launch day runs on assets and relationships built in the week before it. Seven days out, work through this in order:
- Finalize the gallery images at 1270x760. Older guides still circulate the outdated 635x380 spec; the current Product Hunt image requirement is 1270x760, and a mismatched asset gets cropped oddly on the homepage card.
- Write the maker's first comment before launch day, not during it. It should explain why the product exists and ask one real question, since the first comment sets the tone for every reply that follows.
- Tag the Product Hunt link with a UTM code. Without one, the traffic a launch sends shows up as vague referral noise instead of a measurable number; building a UTM code takes about five minutes and is the only way to separate launch-day visitors from everything else in your analytics the week after.
- Line up five to ten people who'll leave a real comment, not just an upvote. People who already used the product during testing make the best early commenters, because they can answer other users' questions in the thread without you; a list of beta testers built before launch does double duty here.
- Pick the day deliberately. Tuesday through Thursday brings more competing launches and a higher bar to clear. Saturday and Sunday draw a smaller crowd and a lower bar, but also a smaller total audience watching the homepage that day.
- Build a one-page command center. A doc with the hour blocks below and a link to your analytics dashboard. Add a place to log bugs reported in the thread so nothing gets lost in a fast-moving comment section.
- Confirm the submission sits inside your broader plan, not as a standalone event; a startup launch checklist keeps Product Hunt from becoming the only thing that happened that week.
The Hour-by-Hour Script
Pacific time governs the whole day regardless of where the team sits, and Product Hunt doesn't adjust the clock for your timezone. Set every reminder to Pacific before you forget and send a wave three hours early by accident.
| Window (Pacific) | What's happening on the platform | What to do |
|---|---|---|
| 12:01am to roughly 4am | Homepage sorts close to randomly; points haven't kicked in yet | Submit at 12:01 and post your own first comment. Notify only your closest 10 to 20 supporters |
| 4am to 10am | Real point-based ranking begins; East Coast and Europe wake up | Send wave one of the email list, reply to every comment within 30 minutes |
| 10am to 2pm | US workday peak traffic on the homepage | Send wave two, staggered by a few hours from wave one. Check the thread every 15 minutes |
| 2pm to 6pm | US afternoon lull; Europe in the evening | Send personal messages to specific supporters rather than a mass blast |
| 6pm to 10pm | West Coast evening; most of Europe asleep | Final public push, thank commenters by name in the thread |
| 10pm to 11:59pm | Last hour before the leaderboard resets | Post a thank-you comment, screenshot where you landed, draft tomorrow's follow-up post |
The numbers behind that schedule are worth sitting with for a minute. A Q1 2026 reading of leaderboard thresholds put Product of the Day at roughly 550 net upvotes on a Saturday and closer to 1,050 on a Tuesday, with headline counts running 10 to 30 percent above the net figure that actually decides rank. A supporter list of 40 people spread across a 20-hour window is two upvotes an hour. Without an email list or a social following doing real work on top of that, the schedule above is organizing a push that was never going to be big enough to begin with.
Check that number before picking a day, not after watching the rank stall at 3pm.
The Comment Reply Rule That Actually Moves Rank
Every guide repeats the same instruction: reply to every comment within 30 minutes. Nothing else on this day is fully yours to control. This is.
On launch day, we had the whole team on board to respond to messages almost instantly. This was a round-the-clock effort by the team to show that we were actively engaged and interested in what feedback we were receiving from the Product Hunt community. We left no questions or comments unanswered.
Tanya Rai, Product Manager, LastMile AI, in a Product Hunt launch retrospective
A six-hour gap in replies between 11am and 5pm reads, to a stranger scrolling the thread for the first time at 4pm, like nobody's home. Nobody comments in an empty room.
The fix is a rotation, not a hero schedule. Assign two people, not one, to own the comment thread in overlapping shifts, so a single lunch break or client call doesn't open a gap a new visitor notices. Check the thread on a timer, not when it occurs to you, every 15 to 20 minutes through the busiest stretch of the day. Fast beats polished.
A Realistic Upvote Target
Before the day starts, write down an honest number: how many people on your list will plausibly upvote within the 20-hour window, not how many people are on the list.
Most Product of the Day winners in 2026 finish with 500 to 1,200 quality-weighted upvotes, and the range depends heavily on day of week and comment engagement, not just the raw size of a founder's following. An email list of 2,000 with a 15 percent realistic open rate, and a fraction of openers who'll click through and vote, gets a founder nowhere near a Tuesday threshold on its own.
Do the math first. That math changes the decision that matters most: which day to launch on. A smaller list aimed at a Saturday's lower bar beats the same list aimed at a Tuesday's higher one, every time the goal is a badge rather than exposure on the busiest day of the week.
When the Launch Doesn't Land Top Five
Most launches don't. That's the unglamorous base rate behind every highlight-reel recap, and it's not a reason to buy a block of fake upvotes from a service pitching an easy badge.
Product Hunt actively polices exactly that behavior. A sudden cluster of new accounts voting for one product in the same few minutes is the pattern its own system is built to catch, which can get a launch's engagement devalued or the listing pulled entirely. The homepage staying "an authentic place for makers" isn't a slogan. It's the stated reason the ranking formula is kept private in the first place.
What survives a quiet launch day regardless of rank: the page itself, and a comment thread full of real questions and answers. A backlink from a domain with real authority outlasts the rank entirely, the kind a comparison of launch platforms is useful for weighing against, since a single bad day on one homepage isn't the only distribution channel worth having.
There's also a channel that doesn't depend on anyone else's community voting at all. A listing on The Board is ranked strictly by what's been paid, nothing else, disclosed as exactly that every time it's mentioned here, and it can go live the same afternoon a Product Hunt launch undershoots.
A dollar buys the bottom rung of that board. It's a different kind of reach than a badge earned by strangers' votes, and it's worth having in reserve precisely because it doesn't depend on timing, timezone, or who else submitted the same Tuesday.
The 48 Hours After Launch
The day doesn't end when the leaderboard resets. Two things need to happen in the 48 hours after, before the thread goes cold and the traffic spike looks like a rounding error in the monthly report.
Pull the UTM-tagged numbers first: how many sessions actually came from the Product Hunt link, and how many of those signed up. Compare what they did in that first session against every other channel that week. That comparison, not the final rank, is the number that tells you whether to do this again next quarter.
Post a recap comment in the original thread thanking specific commenters by name, and answer anything that came in after the day ended; the thread stays indexed and visible long after the badge stops mattering. Email the people who signed up during the launch window with a short welcome note while the product is still fresh in their mind, not a week later once the context is gone.
If the numbers came in soft, the fix isn't a louder push next time. Check it anyway. Pick a quieter day and a tighter supporter list. The reply rotation has to hold for twenty hours straight, which is the part almost every retrospective admits it got wrong the first time around.
A Product Hunt launch is one day on one clock. Launching a startup is the longer project that day either feeds or doesn't, and the full run of launch posts covers everything else that happens before and after this one day. The 48 hours after decide which project it turns out to be.