Product Hunt Launch Mistakes: Your Thumbnail and Your Gallery Aren't Selling the Same Thing

Product Hunt asks for two separate images that most first-time launchers treat as one job. The thumbnail — a 240×240 square — decides whether someone in the daily feed clicks through at all. The gallery, starting at 1270×760, decides whether they stick around once they've landed. Those are two different moments, with two different jobs, and most solo launches only design for one of them.

Usually it's the thumbnail. It's the thing that gets screenshotted for a tweet, so it gets the attention. The gallery gets whatever's left of the evening — five screenshots exported straight from the app, maybe a border slapped on, uploaded an hour before midnight Pacific because that's when the day resets.

And the thumbnail problem itself is narrower than it looks. A full-screen product shot shrunk down to 240×240 turns into an unreadable mess of pixels — logos and icons survive that shrink, dashboards don't. So the fix isn't "make a better thumbnail," it's "stop trying to fit the whole product into a square built for a mark, not a screenshot." That's a five-minute decision once it's named, but almost nobody names it before launch day.

Product Hunt's own submission form only requires two gallery images before it displays at all. Most current launch guides land somewhere between five and eight, because that's roughly the number of beats it takes to walk a stranger from "what is this" to "I get it, upvote." That sequence matters even past launch day, since the traffic decays fast but the backlink it leaves behind doesn't — a gallery a stranger can actually follow is what gets that traffic to convert while it's still there. One over-designed hero image can't do what five plain ones can, because the job isn't to look impressive — it's to answer a sequence of questions one at a time.

What the gallery is actually deciding

By the time someone's looking at your gallery, they've already clicked. The thumbnail did its job. So the gallery isn't fighting for attention anymore — it's proving the click was worth it, fast, before the next tab wins.

That changes what the first gallery image should do. It's not the hardest sell in the set — it's the clearest one. Outcome in under two seconds, no annotation needed if the screenshot already makes the point. The second image can go one layer deeper into the actual workflow, the thing someone would actually click through once they'd signed up. If there's a genuine before/after or comparison, that's where it goes, not stretched across the header of image one where it competes with everything else for the same two seconds of attention.

A screenshot without a browser frame reads as a design mockup, not a working product — which matters more for a solo launch than a funded one, because "does this actually exist yet" is a real question a stranger has about a one-person project. A frame, even a simple one, answers it instantly, though a frame alone still just proves the product is real — it doesn't explain what it does, which is a separate problem the gallery has to solve too.

The other repeat mistake is padding. Product Hunt's gallery canvas is a 1270×760, 1.67:1 frame — wider than a typical 4:3 screenshot, narrower than 16:9 video. A raw screenshot dropped straight in without at least 10% breathing room on each side gets compressed against the edges, and if the source image is a different aspect ratio entirely, Product Hunt letterboxes it on upload. Both look the same way to a visitor: unfinished.

The checklist that actually matters for one person, one evening

None of this requires design skill, but it does require treating the five-to-eight slots as five different questions instead of one screenshot exported five ways — the same discipline behind feature cards that actually get someone to stop mid-scroll, none of which try to say everything at once:

• Each image shows a genuinely different product moment — not the same screen at a different zoom level, which wastes a slot that could've shown something new.
• No more than one annotation per image. One arrow, one label. A screenshot with three callouts fighting for attention is the same mistake as the over-designed hero, just repeated on a smaller scale.
• Consistent padding and framing across all five, so the set reads as one launch instead of five separate screenshots grabbed at different moments of a rushed evening.
• The thumbnail gets designed separately, not cropped down from whatever the biggest gallery image happens to be. Less than a third of Product of the Day launches even used an animated GIF thumbnail — a plain, high-contrast square consistently does the job without needing motion.
• File format matches the content: PNG for a static screenshot, GIF only where there's an actual short interaction worth showing. A GIF of something that isn't moving just adds file weight for nothing.

None of these are hard fixes on their own. What makes them add up on launch night is that they're each individually easy to skip when there's one person, one evening, and a growing list of other things — the maker comment, the tweet, the email to the waitlist — competing for the same hour. Building all five to that spec by hand, the night before launch, is the actual bottleneck — not a lack of design sense, and it's the same crunch six free device mockup tools get tested against on launch night. That's the specific gap List Pro's Product Hunt preset is built around: feed a screenshot in, and the headline weight, screenshot placement, and annotation density are already tuned for the 1270×760 canvas instead of guessed at from scratch at 11pm.

You built the product over weeks. The gallery gets maybe ninety minutes if you're being honest about it. What would the first screenshot need to show a total stranger, with zero context, for the other four to even get looked at?