The Pre-Launch Checklist

Product Hunt features digital products and services that are currently available. A waitlist that does not let a visitor in is not available, and does not get featured.

The named exclusions: directories, templates, boilerplates, podcasts, courses, reports, events, books, services, commerce sites, deal platforms, non-tech submissions, non-functional Kickstarter projects, and hardware without a prominent digital component. Read that list twice if you built a directory.

Product Hunt avoids minimal, undifferentiated products that seem primarily focused on immediate monetization rather than providing long-term value. Write one sentence naming your long-term value. If you cannot write it, you have your answer.

Featuring is judged on Useful, Novel, High Craft and Creative — and a launch does not need all four: often they spike in one or two. Pick yours, write it down, and make the gallery argue that one case instead of four weak ones.

Company accounts are prohibited. A launch is posted by a person, and the comments are answered by that person.

Two launches from the same company scheduled for the same day may be rescheduled to different dates — by them, not by you, and not at a time of your choosing.

Featuring is decided from the data you submitted, after more than one review, and reaching out is rarely necessary. There is no appeal worth planning around. Fill every field before you schedule, not after you are turned down.

Not followers. Not list size. The count of people you could send a personal message to on launch day and reasonably expect to open it. Everything below depends on this number, and it is almost always smaller than the one in your head.

Launch boards differ by more than an order of magnitude in how many products they take a day and in what a visible placement costs. On the biggest, hundreds of products compete and a visible spot costs three figures of upvotes. On smaller boards, the entire day’s field is smaller than the big one’s front page.

It is an observed floor — the lowest score among the launches that were featured on a day somebody counted. Product Hunt publishes no threshold and featuring is an editorial decision, so clearing it guarantees nothing. Falling well short of it does mean the odds are against you, which is the half you can act on.

A founder who can move thirty people will not reach the big board’s featuring line, and on most days is at the top of a smaller one with exactly the same thirty. Winning a small room is a better asset than vanishing in a large one.

Decide this now, while you are calm and nothing is at stake. Nobody picks a good second option at three in the morning on a day that is going badly.

A developer tool does better in front of developers than in front of a general audience ten times the size. Match the audience first, then the size.

One email to people who opted in outperforms a day of posting into a feed. If you have no list, this is the four weeks in which you get one — and if four weeks is not enough, that is a reason to move the date, not to skip the step.

Ask them what the product does. If three cannot tell you, rewrite it. Do this four weeks out, while rewriting still costs nothing.

Widely repeated advice with no published evidence behind it, and cheap enough to be worth doing anyway: the people who watched the thing get made are the people who turn up for it.

Comment on other launches, as yourself, for weeks before your own. An account whose entire history is one launch reads as exactly that — to the community, and to whoever reviews it.

Thirty to eighty names in a file, with the channel next to each. This list produces your first hours. Writing it on the day is how it ends up at nine names.

Tell them the date, tell them exactly what you will ask for — a visit and a comment — and ask whether they are in. A yes given two weeks early is worth ten messages sent at midnight.

Not what it feels like, not what category it sits in. If your tagline would fit a competitor unchanged, it is not a tagline, it is a genre.

Export it, open it on a phone, hold it at arm’s length. Most thumbnails are a wordmark that is illegible small, made by a founder who never looked at it small.

It is the one image most visitors will ever see. Show the product doing the thing, not a headline on a gradient.

Each one earns its slot by showing something the previous one did not. Five screenshots of the same screen is one screenshot, repeated.

The received wisdom is that the product should be doing its most impressive thing within five seconds and that the whole thing must work with the sound off. Sensible, universally repeated, and unsupported by any published figures.

The third one is the one nobody writes and the one that makes the other two credible. Naming who should not buy it is the cheapest trust you will ever buy.

Pick the topics your buyers actually follow, not the ones that look most impressive next to your name.

Website, socials, app stores. One dead link on the day is a conversion you never find out about.

Redeem it yourself, in a private window, with an account that has never existed before.

“How is this different from the incumbent?” “Why is this not a thin wrapper?” “What happens to my data?” You will be asked these in public in the first hours, and an answer written calmly two weeks early reads very differently from one typed at four in the morning.

Launch traffic lands on your site, not on your listing. Fix the bugs, then walk the whole path a stranger walks — signed out, on a phone, on a slow connection.

A separate tag for the listing, the email and each social post. Otherwise the only thing launch day teaches you is the total.

The one day you get concentrated traffic should not be the day you discover the limit.

Product Hunt’s own answer is blunt: The best day to launch is the day on which you’re most prepared. Tuesday to Thursday is a convention people repeat, not a rule anyone published. Being ready beats being on a Tuesday.

Their wording: 12:01 am Pacific Time is the best time to launch for makers that are planning ahead. The board runs a 24-hour cycle. Every hour you launch later is an hour you never get back.

From Central European time, 12:01 Pacific is 09:01 in the morning. Both regions shift their clocks and they do not shift them on the same date, so check the offset for your actual launch week rather than the one you remember.

Why you built it, who it is for, what you want back. Written in daylight a week early — not composed live while the first upvotes arrive.

One send, early, with the link and one sentence of ask. Scheduled, so that a bad morning cannot cost you the largest single lever you have.

Never for an upvote. See the rule further down — this is the place founders break it without noticing.

Product Hunt gives one away free. Put it up before the day, not after it.

This is an eighteen-hour shift, not a morning. A meeting at two in the afternoon costs you the stretch of comment replies that most launches abandon anyway.

Someone should hand you food. This item is not a joke, and founders who skip it stop answering comments around hour seven.

Do not pay anyone to hunt or send traffic to your product — it goes against our guidelines and will result in us taking action to remove your product, and you could be permanently banned from the site.

Same sentence, same consequence. Every service that offers to guarantee you upvotes is selling you the exact thing that gets launches removed.

You cannot ask people directly to upvote your product. Instead, ask them to visit and comment. That covers your direct messages, your captions, your team channel and your mother.

Paying for engagement in favours rather than money is still paying for engagement, and it is the behaviour the guideline describes. The votes are discounted at best and the account is gone at worst.

Product Hunt encourages makers to hunt their own products, and there is no demonstrated advantage to a third-party hunter. The version of that strategy that would actually help you is the version that is banned.

Paste the file you wrote last week. It is the frame everyone who arrives later reads the launch through.

One send. Resending to the non-openers later in the day costs more goodwill than it returns.

In your own words, one person at a time, from the file you wrote four weeks ago. Spread across the hours rather than fired off in one block — partly because it reads as human, partly because a launch that peaks at hour one has nothing left at hour ten.

Every one, including the sceptical ones, and especially the sceptical ones. A good answer to a hard question is the most persuasive thing on your listing.

Product Hunt states the leaderboard moves on upvotes, comments, time since submission, and other factors. Comments are one of the few inputs they name out loud, and the only one that stays entirely within your control all day.

Product Hunt does not publish how the ranking works, and says why: doing so would make it easier to be gamed. So every precise figure you have read — votes per hour, a first-four-hour threshold, a percentage chance of finishing top five — is somebody’s guess dressed as a finding. Do not build your day around one.

Several hundred products launch on a normal day on the big board. The ten on the front page are not the competition, they are the outcome. Knowing the real field size is the difference between a disappointing result and an understood one.

People who already helped like being told it worked. People who have not helped do not respond to being asked twice.

Placement, upvotes, comments, signups, revenue — and which channel each one came from. Today it takes ten minutes. In a month it is unrecoverable.

Founders who report an honest disappointing number keep the list. The ones who go quiet after a bad day have to rebuild it before the next launch.

These are the warmest leads the launch produced, and the shortest-lived.

A day in front of a few hundred strangers who owe you nothing is a research budget most companies have to pay for.

Homepage, deck, changelog, hiring page. The day ends; what you do with it does not have to.

A launch that finishes in single digits usually says one of two things: the room was too big for the reach, or the list was too short. Both are fixable, and both are fixed before launch day rather than after it.

Product Hunt permits relaunching as often as you have new significant product iterations available. Significant is doing the work in that sentence: a new landing page is not an iteration, and the second launch is judged against the same featuring criteria as the first.