A Quarter of Tumblr's Users Are There To Consume Porn, Data Scientists Estimate (vice.com)
On Monday, Tumblr announced that it will permanently ban adult content from its platform on December 17th, alienating a large portion of the site's users who enjoy sharing and consuming NSFW content. Motherboard has surfaced a study conducted in 2017 by two Italian universities and Bell Labs, which found that roughly a quarter of Tumblr users were on the platform largely to consume pornography. From the report: This study was based on the behavior of 130 million users, about half of Tumblr's entire user base. Of that number, "adult content consumers are 22 percent of our sample," the study said. "At the time of the study, roughly 30 million active accounts were consuming adult content, either re-sharing it or following the accounts of those producers," Luca Aiello, one of the study's authors and now a senior research scientist at Nokia Bell Labs told Motherboard in an email. "I expect this audience to experience a noticeable drop in engagement: some of them will just churn out, many of them will likely reduce considerably the time spent on the platform."
Another 28 percent, or roughly 40 million users, encountered pornography unintentionally on Tumblr. That means they didn't seek out the porn, but they followed someone who pushed it into their feed. "The extent of this exposure is hard to estimate but it's probably not major," Aiello said. "Therefore, I believe some people in this segments would be happy to have a cleaner Tumblr feed but I don't expect a significant lift in their engagement, overall." Crucially, the study found that Tumblr's userbase was more female than many social networks ("we estimate that the average user age is 26 and 72 percent of the users are female," they wrote.) They found that these demographics held up between porn consumers and non porn consumers on the site, and that, in fact, young women between the ages of 20-25 were consuming porn on the site at a higher rate than young men. This means that Tumblr's crackdown will likely disproportionately affect women porn consumers.
Another 28 percent, or roughly 40 million users, encountered pornography unintentionally on Tumblr. That means they didn't seek out the porn, but they followed someone who pushed it into their feed. "The extent of this exposure is hard to estimate but it's probably not major," Aiello said. "Therefore, I believe some people in this segments would be happy to have a cleaner Tumblr feed but I don't expect a significant lift in their engagement, overall." Crucially, the study found that Tumblr's userbase was more female than many social networks ("we estimate that the average user age is 26 and 72 percent of the users are female," they wrote.) They found that these demographics held up between porn consumers and non porn consumers on the site, and that, in fact, young women between the ages of 20-25 were consuming porn on the site at a higher rate than young men. This means that Tumblr's crackdown will likely disproportionately affect women porn consumers.
They're being polite. They mean definitely no less than one quarter. The problem with a community built like Tumblr was the intent to blur the line between dialogue, art, and pornography. Mind you, even pornographers are opposed to this, because it also destroys their profit model. This new profit model still requires pornography to survive, but does so in a non-obvious way that appeals better to the subliminal.
... are just too shy to admit it.
And perhaps more seriously though the study does specially states that the aforementioned 28% are there largely to browse porn. Which at least suggest while that remaining 72% are not there just for the porn we don't know how much of them still do use the site for it's adult content.
Ultimately I believe this is a sign of Tumblrs impending collapse, given that this entire situation stems from the fact that their administration was too technological inept to properly administer their site so they just decided to discard a large part of their user-base instead of just fixing the problem.