Re:A/B test of banner ads vs. traffic
by dangrossman on Sunday November 12, 2017 @09:38PM
(#15683858)
If you divide US digital ad spend in 2016 by number of US households, that's about $50/month. Same order of magnitude as my earlier guesstimate.
Re:A/B test of banner ads vs. traffic
by Houshalter on Sunday November 12, 2017 @09:52PM
(#15683910)
Sure but that's a very indirect calculation. Only a fraction of that is going to end up funneled to banner ads and the websites that run them. And that's the relevant metric; how much revenue you are losing by moving away from ads.
On a first order that's right: We don't need all the ad people now that there's no ads.
But then we have to consider that 95% of the costs/fees for running ad networks would still be needed for cost-sharing models too. You still need third-party JS.
Re:A/B test of banner ads vs. traffic
by Houshalter on Monday November 13, 2017 @04:22PM
(#15690374)
Patreon doesn't need js. A subscription model doesn't need js. Certainly not all the complicated tracking code and huge images and videos required for ads.
I was talking more about a model that works similarly to advertising (pennies/fraction of penny per page view)
I'm more of a fan for direct payments, though I think a lot of sites would struggle with just that.
How many people will subscribe to Android Authority directly? Lots of sites basically get views from searches , but little long term engagement. Those sites are still useful, but only have ads as a revenue model that's "viable"
One of my favorite channels is this super niche thing that only gets about 1000-2000 views per video and makes 2 vids a month. That adds up to maybe $4-8 in ad revenue using some rough estimates of youtube's payout rate. On patreon he's making over $500 a month. Which is really respectable for such a small audience.
Online ads have never generated that much money per user and were only profitable in large quantities. Given how scummy they are and this evidence that they repel users, they should be eliminated.