4 ms·
We have to assume they have considered this in the past and come to the conclusion it just doesn't work. I suspect tying a purchasing decision to every page vi
by dageshi 3y ago
We have to assume they have considered this in the past and come to the conclusion it just doesn't work.
I suspect tying a purchasing decision to every page visit just leads to people automatically backing out of the page with a very low conversion rate because it's just too annoying to decide whether to pay for something you're not sure the value of.
Despite all the gnashing of teeth, nothing competes with ads on the web.
- infecto 3y agoSince most news publications are failing businesses I would not immediately assume they have considered all avenues of monetization. I agree that ads have historically made a lot of money...but I am thinking of the paywalled industry. I am probably in the minority but I would be interested if any of the major publications had data on this kind of strategy. I realize there have been products including Flattr that did this but again I never saw it being using on a NYT level publication. Instead of me reading an archive link or just not reading the article at all I would be happy to pay some cents to consume it.
- amadeuspagel 3y agoThe NYT is very much not a failing business.
- infecto 3y agoThanks for the downvote but I said most not all. If you look at the news/magazine industry though it has not made a good transition to the internet age. Lots of closures and consolidations with some of the major outlets acquired as almost toys for the rich. It’s very much a shell of what it once was.
- JoshTriplett 3y ago> We have to assume they have considered this in the past and come to the conclusion it just doesn't work. Or that they prefer the world in which it doesn't take off because they expect to make more money, at least in the short term. If they believe that they can still convince N users to buy $195/year subscriptions, versus getting 10N users to pay a net total of $10/year in per-article fees, then they're better off trying to convince the N users to subscribe. Which might* work for the biggest dozen newspapers with brand recognition, but won't work as well for the long tail of news sites.