Pornhub's removal of all non-verified content led to a 41% drop in traffic within 1 month. Much of the displaced traffic was absorbed by competing platforms, including less regulated fringe websites (+55% visits) and mainstream competitors (+10%). papers.ssrn.com/sol3/papers.... - ThreadSky | a Reddit-style client for Bluesky

florianederer.bsky.social • 58 days ago

Pornhub's removal of all non-verified content led to a 41% drop in traffic within 1 month.

Much of the displaced traffic was absorbed by competing platforms, including less regulated fringe websites (+55% visits) and mainstream competitors (+10%).
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5106235

Comments

5h09un.bsky.social•57 days ago

Cool paper! Any info on the types of content that people migrating from PH to other sites are looking for? I get the DMCA thing, but another major problem that site has had is stuff involving minors and sexual violence.

leonardomadio.bsky.social•57 days ago

Unfortunately we have no information about content consumption and share of “bad” content available to rivals.

leonardomadio.bsky.social•57 days ago

Some more information can be found in the transparency reports that the VLOP sites ("very large online platforms" according to the EC) are obligated to publicly provide everyt six months.

Three of the tube sites considered in our analysis are subject to these obligations.

marcellograz.bsky.social•58 days ago

Currently cooking dinner, so a bit lazy. Did diverted traffic land on other Adult Empire sites?

leonardomadio.bsky.social•57 days ago

Yes, we use non-adult sites as control group. Traffic was diverted to other adult sites, mostly fringe sites, just after the shock.

Our analysis shows that aggregators and search engines had a pivotal role in driving traffic to such sites

marcellograz.bsky.social•57 days ago

Thanks for answering! Were the fring part of Aylo? (PH owner)?

leonardomadio.bsky.social•57 days ago

No, we distinguish between

(i) Aylo (former MindGeek) sites. Here, we have PH and small sites.

(ii) non-MindGeek sites, where we distinguish between mainstream (the top 3 rivals of PH) and the fringe. We have data on a limited number of fringe sites

leonardomadio.bsky.social•57 days ago

The shock affected negatively all Aylo sites since all Aylo sites were "moderated".

In the analysis, we show that while PH managed to mitigate its initial losses, the small sites did not.

marcellograz.bsky.social•57 days ago

zetaof1.bsky.social•57 days ago

Super curious question, did you try using a Gravity model on this? Coming from International Trade I obviously always think in terms of Gravity. But that could be fun here

leonardomadio.bsky.social•56 days ago

Thank you for the interest in our work.

No, we have in mind a search model (and we present it in Appendix A) since we want to rationalize traffic diversion

zetaof1.bsky.social•56 days ago

Thanks, just skimmed it, looks good. I still think it could be fun to throw a gravity model at it. I'll have a bit of a think about it. In my view you have a lot of the marks for it, resistances for site switching, logged-dependent-variable, etc. ... I'll think on it. Cool application though.

leonardomadio.bsky.social•56 days ago

Thanks a lot! We’ll think about it as well!

god-of-abf.bsky.social•58 days ago

"The Net interprets censorship as damage and routes around it" -- John Gilmore, 1993.

Attempts to police this type of content on a single site are, of course, utterly futile; it will just migrate to competing platforms.

The only "solution" is pervasive China-style censorship.

Which is coming.