HypeDesk - a launch site where the demo video is the listing and projects that stop shipping disappear after 6 months
I've launched a handful of products over the past few years — a language-learning app, a voice tool, a prompt manager, a site of free browser utilities. Some launches went fine, some didn't. But they all had the same shape: a busy 24 hours, a small spike in traffic, and then a listing that was effectively dead by Thursday. Six months later the page is still there, still says "launched," and tells a visitor nothing about whether the product is alive, what changed, or whether anyone is still working on it.
Five days ago I started a full-time founder experiment, and the first thing I wanted to fix was that launch-and-vanish pattern. So I built HypeDesk.io, which opens this Wednesday, September 9.
Two decisions define it.
- The demo video is the listing.
Every project on HypeDesk is a 16:9 demo video, five minutes maximum, plus a launch date, product details, a category, and votes. No gallery of screenshots, no wall of copy. Someone scrolling the catalog should be able to tell in thirty seconds whether your product is real and whether it's for them. A tagline and a hero image can be written in an afternoon; a coherent five-minute demo can't be faked, and it forces the maker to actually show the thing working.
The upload flow is YouTube-style: you pick the video, it uploads directly while the browser checks length and aspect ratio and pulls ten candidate cover frames, then the form appears. You choose a cover, a category, a URL, and a launch date up to 90 days out. A human moderator reviews it and it goes live on your date. I'm the moderator right now, which brings me to the limitations later.
- Projects that stop shipping disappear.
This is the part I expect pushback on. You can relaunch a project once a calendar month, with a new video and a changelog describing what changed. The relaunch keeps your page, slug, domain, and votes — it's the same project, appearing again in that day's launches. And if a project has no published update for more than six months, it is hidden from the public catalog until the maker relaunches it. It's still in your dashboard with its history; it just isn't shown to visitors as if it were current.
The thinking: a launch site's real product is trust. A visitor should be able to assume that everything they see is alive. The six-month rule makes that assumption true by construction, and the monthly relaunch turns "launch day" from a one-time event into a cadence that rewards people who keep shipping instead of people who had the biggest DM list on one particular Tuesday.
What's not there, on purpose
No paid featuring, no promoted slots, no editorial picks. Today's launches are ordered by votes, the catalog can be sorted by newest or most voted and filtered across 20 categories, and that's it. I'd rather figure out monetization later than start with the incentive to sell the front page.
What's rough right now
The catalog is nearly empty, because the first launch day hasn't happened yet. It will look thin on Wednesday no matter what I do this week, and I'd rather say that than pretend otherwise.
Moderation is one person, so approval isn't instant. There's no staging environment; I'm running acceptance checks on production with my own test products. Video plays only after you press Play, which keeps my streaming costs tied to real viewers but means Google can't index the videos yet. And the five-minute, 16:9 requirement will filter out some people — if you only have a vertical phone recording, you can't list today. I think the filter is a feature, but it's also the thing that will keep the numbers small at the start.
What I'm asking
If you have a product and a demo video, I'd like you in the first cohort on Wednesday. Submissions are open now and you can pick Wednesday as your launch date; you sign in with Google, GitHub, or X, and listing takes about five minutes if the video is ready. It's free.
And the question I actually want answered: would you list on a site that hides your project after six months of silence? I've talked myself into believing that's exactly what makes a launch site worth browsing, but I'm the one who built it, so I'm the least reliable judge. Tell me where the model breaks.
Hiding dead abandoned projects protects visitor trust, but tying survival strictly to continuous feature shipping creates an awkward incentive trap.
Three specific dynamics to consider:
First, the stability penalty. Some of the most reliable utilities and micro SaaS tools reach feature completion. A database backup tool, a webhook forwarder, or a file converter might not need a new release or video every few months. Forcing those makers to record a fresh five minute video just to prove they exist turns maintenance into a video production chore.
Second, what buyers actually need to know is operational health rather than feature churn. A visitor wants to know three simple things before buying: is the app up right now, does the founder respond to bug reports, and does billing still work? You could solve this with an automated uptime ping check or a quarterly maintenance ping: click one button to confirm this project is still supported and actively monitored. That keeps the catalog clean of true ghost projects without punishing stable tools.
Third, video attention drop off. Five minutes is a long commitment during catalog browsing. Most viewers decide within thirty seconds whether a product is relevant to them. If you let makers highlight a thirty second quick clip or add chapter timestamps, completion rates across your catalog will be significantly higher.
Filtering out screenshot mockups with real working software is a great direction. Just be careful not to conflate lack of new features with an abandoned product.
Building on the point about stable products: would you consider a lightweight “still maintained” check-in instead of requiring a new demo and changelog? You could show “last feature update” and “last verified active” separately. That would preserve the freshness signal without asking makers to ship something just to stay listed.
Good idea, will consider something like that.
most ppl list on directories for a link that stays up, and u hide the page at six months. the videos dont index either, so whats left after launch day?
Yeah, agree, valid point.
You asked where it breaks, so here's the sharpest one: your trust mechanic and your cold-start fight each other, and the 6-month rule loses at exactly the wrong time. Your value prop is "everything you see is alive," enforced by hiding stale projects. But who does that rule remove? The maker who launched, got no traction, and went quiet — which is the default indie outcome, most projects stall. So your rule systematically deletes the majority of listings. At a mature catalog of thousands, hiding the dead ones still leaves plenty. At launch, with dozens, hiding the dead ones leaves a ghost town. The rule that makes a big catalog trustworthy makes a young one empty, and you're young for the next year. Trust-by-subtraction is a scale feature you've shipped at zero scale.
The deeper break: "still shipping" isn't "alive," it's a proxy, and it punishes the wrong outcome. A product so good it's finished, stable, profitable, not changing, gets hidden for the crime of being done. Meanwhile a project that never worked but relaunches monthly stays visible forever. You've optimized for cadence, but the best possible result — a product that reached done — trips your hide rule. You're punishing doneness and rewarding churn.
And the relaunch treadmill reintroduces the fakeability you designed out. Your video can't be faked, that's the whole thesis. But "relaunch monthly or disappear" forces makers to produce a changelog to survive, and a monthly changelog is trivially fakeable, "minor improvements," ship a comma, stay visible. You made the honest signal (video) mandatory and then bolted on a dishonest one (survival-cadence) right next to it.
The fix keeps your spirit and drops the flaw: hide dead, not quiet. The un-fakeable liveness signal is the one your whole site already trusts — is the product still actually running? Ping the URL, check it still loads and operates. A finished stable product passes; an abandoned one fails; and you never punish someone for the sin of building something that didn't need a changelog this month. Liveness is a property of the product, not the maker's posting frequency.
So the real question back: is your six-month rule detecting dead products, or just quiet makers? Because right now it's the second, and the second is a supply gun pointed at your own foot on day one.
The six-month rule is clever because it creates a measurement problem worth solving: the relaunch rate becomes a proxy for "does this platform matter to makers?" But the measurement itself needs precision. "Amount of relaunches" counts tactical relaunches (just keeping visibility) the same as relaunches paired with real changelog updates. And you can't know if a maker relaunched because HypeDesk matters or because they're trying to stay visible on every platform they list on. The signal that actually proves your hypothesis is narrower: makers relaunching here AND removing old listings from PH or Indie Hackers - actively choosing HypeDesk. That's the measurement that proves the six-month rule created preference, not just compliance.
Your own upload flow argues against the five minutes. You say someone scrolling should be able to tell in thirty seconds, then the format is a video of up to five, and the thing actually doing the thirty second job is the cover frame they pick from your ten candidates. A cover frame is a hero image. So the artefact you say can be faked in an afternoon is still the one carrying the catalogue.
The other half: a demo is not un-fakeable, it is expensive. It filters for makers who can produce video, which tracks polish and budget rather than shipping. The one who ships weekly and hates being on camera is exactly who the six month rule is meant to protect, and the video wall is what keeps them out. Two rules pulling opposite ways.
Said as someone whose own site has no product screenshots at all, so your premise lands.
I agree, I made rules a bit less strict but in future I will reduce max time to 2 minutes. Also I do not agree that demo is expensive. With all AI tools we have right now it takes couple hours to make a good video.
The six-month rule is the real bet. If projects disappear, what evidence would tell you visitors actually trust the catalog more—and that makers still see enough value to keep relaunching?
Yep, I have 20 products launched on ProductHunt and 90% of them is dead for long time but PH still keeps listings for no reason.
That’s a strong reason for the six-month rule. What would tell you makers value the catalog enough to relaunch or maintain projects rather than just treating it as another launch directory?
I think the amount of launches and re-lauches would be a strong indicator.
That’s a clean signal to watch. I’d be interested in seeing whether those relaunches actually become a recurring behavior. If you’re open to it, what’s the best email to reach you on?