Dan Greenberg et. al— I appreciate the responses. Just a note I *agree* with what’s being said here about how banning these sites cuts off revenue streams & opportunities to spread. In this regard, the title of this piece is not aligned with the area I was looking at, which was the lack of display ad/paid keywords/paid *upstream traffic* for these fake news/propaganda websites.
As I gathered more data on traffic/category breakdowns, I ended up focusing on the share of fake news sites’ social/direct/search engine traffic because it suggested things like emailing/IM, email newsletters and coordinated PageRank-boosting (i.e., linking) might be driving more traffic than otherwise expected. So, my bad.
But I do agree that “banning” fake news from network $ cuts off an important part of their revenue streams, and that it does help curb the “fake news” problem. I did a comprehensive data analysis related to this adtech, and as you correctly point out, I found a lot of evidence that social tech (e.g., Facebook “like” buttons) plays a role in supporting fake news sites’ ability to track users/coordinate content/spread content/(and of course, monetize). I might update the title so it doesn’t lead readers into feeling like I’ve overlooked the revenue-stream/adtech side of the equation. Thanks again for pointing this out.