It might be slow but if you have no strong reason to be on Tumblr besides everyone having been on Tumblr you will see a slowly building migration away from Tumblr.
And losing 1/4 the user base overnight is a great way to kick off that migration. It only takes one blog in a web of follows jumping ship to motivate all the other producers and the fans of that blog and its related blogs to move to alternatives.
Perhaps this is a neglected downside. By banning porn you will turn a part of your platform into advertisments for your compeditors. Every post from a popular person saying "I'm moving to X" is an advert for X on your platform.
I suspect the bottom line comes down to the bottom line, though; Tumblr has spent over a decade being a pit that absorbs money, and Verizon is not the kind of company that's going to run a service like that as an experiment. They're making a bet that kicking off the pronz will increase ad revenue, and I'm sure they have projections to back that bet up. That doesn't mean the projections are right, of course, but I don't see how a money-losing Tumblr would have continued under Oath. If this bet doesn't work and they shut it down, everyone will say "ha, we told you that you shouldn't have kicked off the pronz," but the pronz ain't never brought in the bucks for them, and we can safely assume "explicitly set out to monetize the pronz" was never in the cards.
However, now it's very likely that I'll find a mediating app -- or I'll be scraping my own feed -- to avoid the inevitable increased influx of advertisements so Verizon will still lose that possible income from me. I suspect many others will take similar measures (less known apps like Tyblr in iOS for example rid you of ads and make your feed sorted by timestamp and not by an "algorithm").
There are former Tumblr staff members on Twitter saying it will lose 50% of the userbase. Losing 50% of the userbase will unquestionably kill the site stone dead.
In 2016, Motherboard actually did the digging, and discovered a few interesting numbers, categorizing 130M Tumblr users into "producers" of porn, "consumers" (those who follow the producers), and "unintentionally exposed" users who don't follow porn producers but see reblogs from them.
* Only about 1% of users are producers. * About 22% of Tumblr users are consumers. * About 28% of Tumblr users are unintentionally exposed.
So, what happens?
* Obviously, they lose the producers. * They lose some of the consumers, but not all. (Not all of the consumers are there exclusively for porn.) * They lose some of the unintentionally exposed, but probably not most.
They may lose a quarter or more of their userbase this way, sure, but Tumblr's estimates a few years ago for their total user count were over 200M. Maybe that drops down to "only" 150M, or even 100M. But now their hosting costs have dropped dramatically, too. So if advertising revenue increases or even stays the same as it is now, they win. And 100M users is absolutely a big enough audience to bring in advertisers.