onlyfans didn't kill the tube, it hollowed it out

the money moved first, then the users followed

start with the numbers because they’re the only part of this story that isn’t up for debate.

onlyfans went from around $1.4 billion in gross payments processed in 2019 to somewhere north of $5.6 billion by 2022, according to its own companies house filings in the uk. its platform cut alone (the 20% it keeps before paying creators) was reported at roughly $1.1 billion in revenue that year, with something like $525 million in pre-tax profit. that’s one platform, in one filing year, doing numbers that dwarf what any single tube discloses.

and that’s the catch: tubes don’t disclose. aylo, the company that owns pornhub, redtube, youporn, brazzers and reality kings, is privately held and doesn’t publish revenue. the closest thing to hard data is old mindgeek-era estimates putting pornhub’s ad revenue somewhere around $450-500 million a year, and that estimate is from before several state age-verification laws cut off entire population blocks and before similarweb started showing multi-year traffic declines on the flagship tube.

so the honest version of this comparison is: one side of the industry got dramatically richer and told everyone the exact figure, and the other side got quieter and stopped telling anyone anything. that asymmetry is itself the story. when a business model is winning, it doesn’t need to hide its numbers.

the subscriber economy isn’t a rounding error next to ad-supported tube revenue anymore. by 2023 onlyfans alone had over 4 million registered creators and past 305 million registered users. fansly and manyvids add more volume on top of that. none of those creators are relying on a tube’s category page to get found, and increasingly, none of them need to.

the carousel is not a fix, it’s a patch over a hole

tubes noticed the money leaving and did the thing large companies do when they notice money leaving: they bolted a widget onto the page instead of rethinking the page.

that’s what the “creator” tabs and promotional carousels on pornhub, xvideos, and a handful of others actually are. scroll a category page on any major tube now and somewhere in the middle you’ll hit a row of creator profile thumbnails, usually labeled something like “models” or “creators,” running an affiliate arrangement where the tube gets a cut for referring signups.

it’s not nothing. it’s also not discovery. a carousel is inventory placement, the same mechanism as the “you might also like” strip at the bottom of a news article. it shows you whoever paid for placement or whoever the algorithm decided converts best that week, not whoever matches what you were actually browsing for. you were three clicks deep into a specific category and it interrupts you with an unrelated upsell.

compare that to what actual category browsing on a tube used to promise: you pick a tag, you get scenes that match the tag, ranked by something resembling relevance or freshness. the creator carousel breaks that contract. it’s the tube admitting it can’t compete on its own turf anymore, so it’s renting out shelf space to the thing that’s beating it.

a creator on manyvids doesn’t exist if you’re standing on xvideos

here’s the part that actually matters for anyone trying to find something specific: discovery on the adult web is almost entirely platform-locked, and creator platforms made that worse, not better.

if a creator posts exclusively on fansly, and you’re browsing xvideos, that creator is invisible to you. not buried, not deprioritized, actually absent. there’s no crawl, no index, no cross-link. the tube has zero incentive to surface a competitor’s paywalled content, and the creator platform has zero incentive to be indexed by anyone outside its own app, because indexing means someone might find the content for free somewhere downstream.

the result is an industry where four or five completely separate discovery graphs exist side by side, each with no edges connecting to the others. someone who wants a specific act, a specific tag combination, a specific type of scene has to already know which platform hosts that thing, or they have to manually check five apps one at a time. that’s not a search problem, that’s an information architecture problem, and it’s gotten worse every year subscriber platforms have grown, because more of the industry’s actual output sits behind walls that don’t publish sitemaps.

here’s my actual contrarian take, and i think a decent number of people will push back on this: onlyfans didn’t just take revenue from tubes, it made discovery on the adult web strictly worse. the old tube model, ad-supported and gross as it was, at least indexed everything into one browsable category structure. the subscriber model optimized for creator payout and platform lock-in, and treated discoverability as something to actively suppress rather than solve. more money moved to creators. less of the industry’s content became findable by someone who didn’t already know what they were looking for.

why this makes cross-platform tools more necessary, not less

if discovery keeps fragmenting along platform lines while the money keeps moving toward platforms that don’t want to be indexed, then the tools that sit above all of it, the ones that dedupe, tag, and cross-link regardless of who’s hosting, become more load-bearing every year, not less.

the tube-era answer to “what should i watch” was: pick a category on a tube, scroll. that answer still half-works for tube content. it doesn’t work at all for the growing share of the industry living on subscriber platforms, and it definitely doesn’t work across the two.

that’s the gap prowl sits in. we’re not trying to be a tube and we’re not trying to compete with onlyfans on hosting. we index what’s public across the major tubes, dedupe the scenes that show up on four or five of them under different titles, and build a category structure that doesn’t depend on any single platform’s business model staying intact. the money already moved. the map didn’t catch up yet.