Georgii EmelianovGuides

How to Make a Product Hunt Launch Video

Seven guides rank for this and all seven treat the video as the hero asset. On Product Hunt it is not — the gallery is, and the difference changes what you build first.

A white torn-paper title card headlined “How to Make a Product Hunt Launch Video”, with one large deep-violet layer sweeping across the upper-left and three small magenta fragments along the lower-right edge.

Every guide ranking for product hunt launch video is published by a company that sells a video tool, and all of them tell you the same thing: the video is the centrepiece of your launch, here is a beat structure, here is our product. Not one of them cites Product Hunt's own documentation, and the documentation says something that reorders the whole plan.

On Product Hunt, the video is an embed a visitor has to choose to play. What renders on the page is the gallery. What renders in the feed is a thumbnail. And GIFs in the gallery animate on hover rather than autoplaying, which means the first frame of your GIF is the still that most people see. The video matters, but it is the third asset in line, and building it first is why launch-day asset packs end up unbalanced.

This guide is built around Product Hunt's stated requirements, checked on their help centre on 21 August 2026. No upvote statistics appear below, because none of the pages promising them cites a test.

What Product Hunt actually accepts

Get this on screen before you make anything. All of it is from Product Hunt's own "How to post a product" help article.

video       YouTube links only. No MP4 upload exists.
            Full URLs — shortened links will not load.
            Public or unlisted, and uploaded before launch day.
gallery     Minimum 2 images. Recommended size 1270 x 760.
thumbnail   Square. Recommended 240 x 240.
GIFs        Under 3MB. Animate on hover, they do not autoplay.
            Strobing and quick cuts are discouraged and may be edited by PH.
description 260 characters.

Two of those lines decide the production order. Only YouTube links are supported means your video needs a channel and an upload, not an export. And GIFs animate on hover means frame one of any animated gallery item is a static image people will judge you on, so it has to be a designed frame rather than wherever the encode happened to start.

The short version

  1. Build the gallery first. Two or more images at 1270 x 760, the first one carrying the whole pitch.
  2. Write a 45-second spine. One problem line, one product line, one flow, one close.
  3. Capture the product working. Real interface, real motion, one flow completed.
  4. Cut for silence. Burned-in captions, no intro card, product on screen from frame one.
  5. Upload to YouTube days early. Public or unlisted, ads off, full URL saved.
  6. Assemble the post the day before. Thumbnail, gallery, video URL, 260-character description.

Step 1: Build the gallery before the video

The gallery is the asset with the highest guaranteed impressions, because it is what the page shows without any action from the visitor. Product Hunt requires at least two images and recommends 1270 x 760.

Treat the first gallery image as the whole pitch. It should carry the product name, one line of what it does, and a real screen. Someone who reads only that image should be able to describe your product to a colleague. Everything after it can be detail.

An animated GIF in the gallery is worth the effort, with one condition: because it animates on hover rather than autoplaying, frame one is a static thumbnail. Export the GIF so frame one is the frame you would have chosen as a still, and keep it under 3MB. Product Hunt also discourages strobing and quick cuts, and notes those may be edited by their team, so a calm loop beats a montage.

The thumbnail is separate and square, recommended at 240 x 240. It shows in the feed at small size next to competing launches, so it is your icon or a single bold mark — not your logo with a tagline, which is illegible at that size.

Step 2: Write the 45-second spine

Four lines, in this order. Write them before you record anything.

  1. The problem, in one sentence. Named as the reader would name it, not as your positioning doc names it.
  2. What the product is, in one sentence. Product name, category, the thing it does.
  3. One flow, completed. The single task that proves the claim.
  4. The close. What to do next, in three words.

Forty-five seconds is convention rather than a rule — Product Hunt publishes no duration guidance, and the 45-to-75-second range that circulates comes from community practice, not from a test. What is true is that your audience arrived deliberately, so you have more attention here than on any other surface in a launch. The tradeoff is that a Product Hunt visitor is comparing you against four other launches that day, so length spent on atmosphere is length spent while they scroll.

The lines above are also the caption cards. Writing them first means the edit is assembly rather than authorship.

Step 3: Capture the product actually working

The one thing a Product Hunt audience is unusually good at detecting is a video that avoids showing the product. This is an audience of builders. A launch video that opens on a logo animation and cuts to a testimonial reads as a product that is not ready.

So: real interface, real motion, one flow from start to finish. The specifics of getting a clean take — setting the app's state, removing debug UI, recording in scenes rather than one run — are the same as for any product video, and the app demo video guide covers them in detail.

Three things specific to this surface:

  • Seed the data before you record. Empty lists and "Test User 1" undercut a launch more here than anywhere else.
  • Record each beat separately. One take of a whole flow means one fumble costs the run.
  • Get one shot you can freeze. You need a gallery still and a YouTube thumbnail out of this session, and pulling them from footage you already have beats a second session.

Step 4: Cut for a silent hover

Assemble in spine order, then apply three constraints that come from the surface rather than from taste.

Product on screen from frame one. No logo intro, no loading screen, no black. The first frame is also your YouTube thumbnail candidate.

Captions burned in. The feed is silent, and even on the product page most people will not have sound on for a launch they are skimming. Text cards between beats carry the argument. Keep them out of the bottom fifth, where YouTube's player controls sit.

No dead air. Cut the travel between a tap landing and the next tap starting. That single pass usually takes a 70-second assembly to 45 without removing a single beat.

The strongest signal you can send on launch day is a video where the product is doing something in every second of it. That is a cutting decision, not a production budget.

Music is optional and low-stakes here. A voiceover is worth it only if you are personally the pitch — a founder talking over their own product performs well on this platform in a way it does not on an App Store listing.

Step 5: Upload to YouTube and wire it in

Product Hunt states plainly that only YouTube links are supported. There is no MP4 upload, and shortened links will not load.

  1. Upload to YouTube at least two days before launch. Processing at higher resolutions takes time, and you do not want to discover that at 12:01am Pacific.
  2. Set it to public or unlisted, never private. Private videos do not render, and Product Hunt's help centre calls this out specifically.
  3. Copy the full watch URL. Not the share shortlink.
  4. Pick the YouTube thumbnail deliberately. It is a second poster frame, and the auto-selected one is usually a mid-motion blur.
  5. Play it back from the pasted URL in the launch preview before you schedule.

Step 6: Ship the assets the day before

The final assembly is short if the previous steps happened. Thumbnail at 240 x 240, at least two gallery images at 1270 x 760 with the pitch image first, the YouTube URL, and the 260-character description.

Two hundred and sixty characters is tighter than it sounds — roughly three sentences. Spend them on what the product does and who it is for, not on the story of building it. That story belongs in the maker's first comment, which has no such limit and is the asset most first-time launchers underuse.

Where this goes wrong

The video was made first. The gallery ends up as three screenshots pulled from the footage at the last minute, and the asset with the highest impressions is the one that got the least thought.

Frame one of the GIF is a blur. Because it animates on hover, that frame is what most visitors see. An encode that starts mid-transition looks like a broken image.

The YouTube video is private. The single most common launch-morning failure, and Product Hunt's own documentation warns about it.

The thumbnail has words in it. At 240 x 240 in a feed, a tagline is a grey smear. Use a mark.

The video shows five features. A Product Hunt visitor is deciding whether to look further, not learning your product. One flow completed beats five sampled, every time.

The launch is the only marketing. A Product Hunt day is a spike, not a strategy, and the assets you build for it should be reusable on the App Store, the landing page and social. Where the launch sits in the broader plan is covered in the solo developer's app marketing playbook.

Doing it without the recording session

The reason launch videos get made the night before is that the capture session is a real block of work that cannot start until the product is in the right state.

If you build for Apple platforms, that block can be automated. Reely drives an iOS feature in the Simulator, records the run, and composites the cuts, camera moves and captions from a spec file rather than from a timeline. Because it authored the run, the recording carries a sidecar of labeled interaction marks, so the caption lands on the tap and a re-run after a UI change retimes everything downstream. It also renders the same source at several canvas sizes, which is convenient when the launch needs a 16:9 for YouTube and a portrait cut for the posts around it.

The bounds, stated honestly: Apple platforms only, so an Android or web launch is out of scope. The fully generated path films a mock of your feature rather than your shipping binary, though if you already have a recording it can wrap the motion design around that instead. Creating is free; exporting is the paid part.

The general version of the same job — script, take, cut, export, for any surface — is in the app promo video how-to.

Frequently asked questions

Do you need a video for a Product Hunt launch?

No. Product Hunt requires at least two gallery images and does not require a video at all. A launch with a strong first gallery image and an animated GIF can do well without one. A video helps when the product's value is in motion — a workflow, an interaction, something a still cannot show — and it is a poor use of launch-week time otherwise.

Can you upload an MP4 to Product Hunt?

No. Product Hunt's help centre states that only YouTube links are supported, and that shortened links will not load. Upload the file to YouTube, set it public or unlisted rather than private, and paste the full watch URL into the launch form. Do the upload at least a day early so processing has finished.

How long should a Product Hunt launch video be?

Product Hunt publishes no duration requirement. Community practice lands around 45 to 75 seconds, which suits an audience that arrived deliberately and is comparing several launches the same day. The useful constraint is not the total but the density: every second should have the product doing something, which usually pulls the cut shorter than planned.

What size should Product Hunt gallery images be?

Product Hunt recommends 1270 x 760 for gallery images and requires at least two before the gallery is viewable. The thumbnail is separate and should be square, recommended at 240 x 240. Animated GIFs in the gallery must be under 3MB, animate on hover rather than autoplaying, and should avoid strobing effects and quick cuts.

What to do this week

Make the first gallery image before you make anything else. One frame, at 1270 x 760, carrying the product name, one line of what it does, and a real screen. If that image cannot stand alone, no video will rescue the launch, and if it can, the video's job gets much easier.

Then get the footage. If the product is a native iOS feature and the capture session is what keeps sliding, Reely generates it from a description of the flow — free to create with, paid only when you export.

← All posts