Chronotope’s avatarChronotope’s Twitter Archive—№ 72,240

                                                      1. Thing that has been frustrating me about the fake news coverage: Reporters tend to treat ad revenue as profit when it is gross revenue.
                                                    1. …in reply to @Chronotope
                                                      That NPR article & others report $10k to $30k/mo as 'how much fake news sites make'. Ok. So, how much does it cost to run a fake news site?
                                                  1. …in reply to @Chronotope
                                                    Let's play at some back-of-napkin math. He makes $20k in a month and we'll say he makes $0.04 per impression, similar to a mid-quality site.
                                                1. …in reply to @Chronotope
                                                  So that means 500,000 impressions on his fake news site a month (honestly not that much). So what are the costs?
                                              1. …in reply to @Chronotope
                                                Those are impressions that earn, Considering site quality let's say his traffic is prob 50% bots that don't earn. So monthly traffic is 1mil
                                            1. …in reply to @Chronotope
                                              So talking with some folks about these numbers and looking at the example site given in the NPR article which has 2 RevContents, 1 display
                                          1. …in reply to @Chronotope
                                            The Display is Q1Media. Looking at various reports it is placed as $15-$30 CPM. Now looking at RevContent.
                                        1. …in reply to @Chronotope
                                          Of the content networks RevContent supposedly has the highest RPM and lowest quality, both in content and bot-checking.
                                      1. …in reply to @Chronotope
                                        (Sorry, I meant RPM for Q1Media above)
                                    1. …in reply to @Chronotope
                                      Revcontent's own site says RPMs as high as $40, others report as low as $3/unit. Let's split the difference on both:
                                  1. …in reply to @Chronotope
                                    If we went down the middle on both and added up their RPMs it would actually be higher than $40/RPM. And that's not even counting clicks.
                                1. …in reply to @Chronotope
                                  So actually, I'm going to stick with my initial estimate. $0.04 per impression. But it was good to do the math. It might be way lower but...
                              1. …in reply to @Chronotope
                                ... like I said, this is back of napkin stuff.
                            1. …in reply to @Chronotope
                              As an aside, can I note how depressing it is that this fake news site has better load performance than most real news sites?
                          1. …in reply to @Chronotope
                            So that level of traffic is above your average shared hosting plan. So the host is likely a dedicated plan or AWS.
                        1. …in reply to @Chronotope
                          Let's make a safe assumption that these people are relatively unsophisticated technically, so not AWS.
                      1. …in reply to @Chronotope
                        Since it is a WordPress site, though it is not likely hosted on WPEngine, I'm going to just go there to get a guesstimate of hosting costs.
                    1. …in reply to @Chronotope
                      And I'm not an Ops expert, so if you are, feel free to correct me.
                  1. …in reply to @Chronotope
                    According to reports, a >1mil WP site on WP Engine would have a base hosting cost of $2,000/month
                1. …in reply to @Chronotope
                  Assuming that $20k/mo earnings, that means 10% of revenue minimum for hosting costs. Then we get into a bunch of other costs:
              1. …in reply to @Chronotope
                Is there a CDN? That has a cost. Also they apparently pay writers. That too has a cost. Though both are a lot harder to estimate.
            1. …in reply to @Chronotope
              Also unclear: how much time does it take to make fake news? What's the equivalent hourly rate these people are making?
          1. …in reply to @Chronotope
            I don't know here, it is real into the weeds depending on a lot of factors.
        1. …in reply to @Chronotope
          If we're going to be conservative, would it be fair to say paying for writers and home office costs is an additional $4,000/mo?
      1. …in reply to @Chronotope
        Who knows here, but let's just take that as a number for some basic costs, we'll say it includes CDN costs too. So yearly income: $168,000
    1. …in reply to @Chronotope
      Of course, our American fake news site-runner also has to pay taxes.
  1. …in reply to @Chronotope
    According to Wikipedia, Federal taxes for 100,000 to 335,000 are "$22,250 + 39% of the amount over 100,000"
    1. …in reply to @Chronotope
      So total taxes would be $17,846+22,250 = $40,096. So our American fake news maker would bring in $127,904 profit per year. Being generous.
      1. …in reply to @Chronotope
        (I'm assuming he properly deducts his costs, of course.)
        1. …in reply to @Chronotope
          In reality, costs are likely higher, and t/f actual yearly profit is probably less.
          1. …in reply to @Chronotope
            It would be cool if the next fake news report actually filled in some of these cost blanks. It would likely sound a lot less panicky.
            1. …in reply to @Chronotope
              So bluechoochoo notes that I'm not even accounting for purchased traffic costs, which def exist - bluechoochoo/807294369090834432
              1. …in reply to @Chronotope
                Seems increasingly unlikely that even the high performing fake news maker in the NPR article breaks 100k/year.
                1. …in reply to @Chronotope
                  So Wonkette's commiegirl1 notes their effective revenue per impression is $0.0037. commiegirl1/807295794378342400
                  1. …in reply to @Chronotope
                    To get the revenue reported in the NPR article, fake news sites at that rate of revenue would have to create 5.4mil page views a month.
                    1. …in reply to @Chronotope
                      For the record that would pop minimum estimate of hosting cost based off WPEngine numbers to $7,200. So 36% of profits minimum for hosting.
                      1. …in reply to @Chronotope
                        So even if they spent zero dollars on paid traffic (highly unlikely) we're talking ~50% of revenue going to business costs.
                        1. …in reply to @Chronotope
                          Taxes still paid on the over $100,000 slot. But now about $2-3k/mo. Which means they're making mby $8,000/mo profit? Still being generous.
                          1. …in reply to @Chronotope
                            So, and this is assuming the fake news sites actually get >5.4 million monthly impressions--which seems unlikely, annual profit is $84-128k
                            1. …in reply to @Chronotope
                              But that's before accounting for paid traffic costs. If we counted those in, that number gets *way* lower. I'd bet less then $80k by a lot.
                              1. …in reply to @Chronotope
                                BTW, $8-7,000/mo is very close to the NYT article's number. So maybe *that* fake news person was reporting profit not gross revenue?
                                1. …in reply to @Chronotope
                                  So yeah, not as crazy profitable as it initially sounded to run a fake news site.
                                  1. …in reply to @Chronotope
                                    That said, sad truth is if he worked 40/hrs per week on fake news the profit is still more than the average salary for a single reporter.
                                    1. …in reply to @Chronotope
                                      $72k > highest average reporter salary by location reported by Pontyer in 2013 poynter.org/2013/why-an-average-journalism-grads-salary-might-not-be-an-average-salary-where-you-work/202172/
                                      OpenGraph image for poynter.org/2013/why-an-average-journalism-grads-salary-might-not-be-an-average-salary-where-you-work/202172/
                                      1. …in reply to @Chronotope
                                        But on the flip side, he has to pay his own healthcare costs in that case? So perhaps it comes out about even.
                                        1. …in reply to @Chronotope
                                          ($72k based on the $6k/mo number in the NYT article)
                                          1. …in reply to @Chronotope
                                            Though worth remembering, all these numbers get much lower if he's paying for a decent chunk of traffic, which is very possible.
                                            1. …in reply to @Chronotope
                                              So, to sum up: fake news person in NPR's coverage prob makes enough to underpay a few writers and pay himself less then avg salary for DC.
                                              1. …in reply to @Chronotope
                                                Sure, it would be better if he made $0. But in the grand scheme of things the alleged largest player is still a fairly small fish.
                                                1. …in reply to @Chronotope
                                                  It's likely WAY less. From Aug-Election day BuzzFeed reported 8.7mil Fake News conversions from FB. Makes the 5.4 number seem unlikely.
                                                  1. …in reply to @Chronotope
                                                    Also, seriously makes me doubt the numbers self-reported by fake-news-writing teens in the NBC story nbcnews.com/news/world/fake-news-how-partying-macedonian-teen-earns-thousands-publishing-lies-n692451
                                                    1. …in reply to @Chronotope
                                                      I mean sure... party night on AdSense bucks. But buying BMWs? Seems unlikely. Mby NBC should have been a little less trusting.
                                                      1. …in reply to @Chronotope
                                                        I mean... I dunno. Maybe let's not trust people who are making a living out of writing fake news to tell the truth about their earnings?
                                                        1. …in reply to @Chronotope
                                                          And even if we double their FB traffic on their best months, we're talking about a total of 6 million impressions on Fake News a month?
                                                          1. …in reply to @Chronotope
                                                            I know it looks like a big number... but in grand scheme of things: relatively minor. The avg John Oliver YouTube clip reaches more people.
                                                            oh my god twitter doesn’t include alt text from images in their API
                                                            1. …in reply to @Chronotope
                                                              I mean... seriously. One LastWeekTonight YouTube video might have more plays than the entirety of impressions for fake news over a month.
                                                              1. …in reply to @Chronotope
                                                                (Note: that BuzzFeed number is *engagements* which means not all of them are even monetized click-throughs.)


Search tweets' text