One in five launches has no domain

Of 14,354 Product Hunt launches with a working web address, 2,932 do not have a domain. They point at an App Store page, a Play Store listing, a Chrome Web Store entry, a GitHub repository, a Gumroad checkout, or a subdomain on somebody's hosting platform.

That is 20.4%, and it has been roughly 20% every quarter for two years.

Two different kinds of nothing

The 2,932 split into two groups that behave differently and are worth keeping apart.

An app-store listing means the product has no web address of its own at all. apps.apple.com/us/app/kitchen-pantry-tracker is where the product lives. There is no site to visit, and the store page is doing all the work a homepage would do.

A platform subdomain means the product does have an address, and never registered anything. nextrelations.vercel.app is a real site at a real name. The name is simply rented from Vercel rather than owned.

Apple accounts for 588 of the store listings and Google 539, the latter split between Play and the Chrome Web Store. GitHub contributes 314, which are mostly open-source tools where the repository is the product. Gumroad adds 175, which are mostly paid templates and downloads.

On the platform side, Vercel is dominant with 477, then Netlify at 151, GitHub Pages at 128 and Lovable at 84.

The line that did not move

The interesting number here is the one that stayed still.

Across two years, the share registering a domain never leaves the band between 78.9% and 82.5%. It begins at 79.9% and ends at 80.3%. In the same period the number of launches per quarter went from 284 to nearly 4,000.

That stability is the finding, because plenty happened during it that should have disturbed it. Hosting platforms made shipping without a domain trivially easy and free. AI app builders arrived and put a working deployed URL in the hands of people who had written no code. Lovable alone accounts for 84 launches here, all of them on lovable.app, and did not exist in this dataset's first quarter.

Platform subdomains did grow, from 3.2% to about 8%. They roughly doubled, then stopped. From the third quarter of 2025 onward the figure moves between 6.8% and 8.3% with no direction in it. Whatever population is content to launch on a rented name appears to have been reached.

Reading it honestly

Product Hunt selects for a particular kind of launch, and that matters more here than in most of our research. Posting a product on Product Hunt is an act of promotion, and someone willing to do it has already decided the product is worth showing to strangers. The genuinely casual end of building, the weekend project shared in a Discord and nowhere else, is underrepresented by construction.

So read 79.6% as the domain-registering rate among products whose makers thought them worth launching, not among everything built. The first number is the one that describes intent.

One more caveat on the store-listing group. A mobile app with a real website is counted here by whichever address Product Hunt holds, and that is often the store. Some of those 1,849 launches do own a domain and simply did not lead with it, so 12.9% is an upper bound on the truly domainless mobile products rather than a measurement of them.

What it suggests

The free subdomain turned out to be a complement rather than a substitute. It absorbed a class of product that would otherwise have launched with nothing at all, and it stopped there.

For the four in five who do register, the decision looks less like a cost question than a signal one. A .vercel.app address says the thing was built quickly and may not be permanent, which is accurate for a large share of what appears on Product Hunt and is exactly what a founder building something they intend to keep does not want said.