4 ms·
Just like no one wants to think about spending money each time they turn on a light in their house or take a shower? Yes, it'll be a barrier for people for a we
by billytetrud 5y ago
Just like no one wants to think about spending money each time they turn on a light in their house or take a shower? Yes, it'll be a barrier for people for a week, and then they'll forget about it like everything else we pay for individually.
> your ratings system would perpetuate the sort of think-piece journalism in which authors proclaim whatever their readers currently believe
The idea is intended to solve the problem of low-quality news, not thought-bubbles. People will always gravitate towards their own biases, this won't solve that, nor will it make it worse. It could make it a bit better for people who are actually looking to escape their biases, because those people could rate articles more highly if they actually convince them to change their views, or at least teach them about an alternate viewpoint in a non-infuriating way.
- ghaff 5y agoMicropayment startups have come and gone. I agree with you that pay-as-you-go has been normed with electricity, water, and probably even gas for cars. But maybe it's because people feel they need to do those things for some definitions of need. Whereas media, even newspapers, is more in the class of entertainment. Even music, which was trending towards digital purchases for a while has largely moved to all you can eat subscriptions.
- billytetrud 5y ago> Whereas media, even newspapers, is more in the class of entertainment But why would pay-as-you-go be bad for the class of entertainment? The vast majority of entertainment options out in the world are pay as you go (live shows, movie theaters, vacations, restraunts, etc etc). > Even music, which was trending towards digital purchases for a while has largely moved to all you can eat subscriptions. Music has been an interesting outlier here because people are increasingly listening to music playlists and not hand-picking their own music. Hand-picking music is time consuming and a pain sometimes, so why not let the algorithms decide what to listen to? By contrast, people still hand-pick news articles, TV shows, movies, and things like that - and they'll likely continue to do that because they're activities that generally require engagment - whereas music is generally listened to in the background while doing something else. The movement towards the subscription model has primarily been about improving things for businesses, not for customers. The rhetoric is "customers don't want to think about paying", but we can clearly see countless counterexamples in the world. Businesses, however, like the subscription model because it makes it easier for them to predict their revenue and it puts up a (small) barrier in place for customers to stop paying them. However, while the subscription model is great for single items (eg high-end software) where the subscription is basically just a way to pay-as-you-go automatically, the subscription model is actually a pretty bad payment model for networks of content when you think about what it incentivizes. The user pays a fixed price for unlimited content. This incentivizes the user to consume as much as they can. A user who use the service 100 times more than another user both pay the same price. This means it necessarily excludes users who don't consume a lot, and raises the price for people who consume a below average amount. There is a deadweight economic loss there. In addition it provides no net incentive for the subscription provider to increase the options available in the system or the quality of those options. In fact, there is an incentive to release low-quality content as filler so that their high-consuming users don't run out of content (which they do anyway). Take a video streaming service for example. If they increase the number of options available to customers, they don't earn any more money but each item available in the system gets watched less, which means owners of the content would would to be part of that subscription network less and less as the content grows. Either that or they charge customers more as the library grows - which customers wouldn't pay for since most people wouldn't actually consume more just because the library's bigger. What results is that video subscription services have necessarily fragmented and will stay fragmented for as long as people keep drooling over the subscription model. People are locked into one or two networks of content which are primarily all they watch. They don't have access to other content without paying significantly more or making a burdensome leap of ending service with one network and beginning with another (which loses them access to content they had there). News is even worse, where individual newspapers have their own siloed subscription services. That's great if you read most of the content on there, but its terrible if you just want to read an article someone sends you ever once in a while. There's some more aggrigated news services that have come out recently, but these will also have the fragmentation issues that video streaming has because of the subscription model.
- ghaff 5y agoThe entertainment people pay for a-la-carte is relatively high-end though. It's $5 to $10 to $100+ . People are fine with making that a conscious decision. Not so much with a $.05 for an article. I'm not justifying this pattern. Just observing that it seems to be the case (as have others). As for video services, lots of people argued against the $100/month bundle as augmented by the $5/each Blockbuster payment. I actually prefer the current fragmented services plus a la carte even if it's annoying sometimes. Certainly saves me money. ADDED: >In addition it provides no net incentive for the subscription provider to increase the options available in the system or the quality of those options. Not sure I understand that. While there's certainly inertia, if your subscription isn't providing me with value, I cancel it. (Some companies make it difficult to do so but that's a separate issue.)
- billytetrud 5y ago> Certainly saves me money. I think a lot of people believe its saving them money, but in reality its actually costing more and at the same time lower quality than it would be with a correctly-priced pay-per-view model. You can't compare to blockbuster because that system was old and there have been many things that can bring the cost down these days. > if your subscription isn't providing me with value, I cancel it Yes, that's why streaming video services have some quality content, and its why they price them so cheaply - so its worth it despite the fact that most of the content is low quality and lacks a variety of quality content. If they increased the options available, they'd have to charge more money than its worth to cover the costs of that extra content. If they increased the general quality, same thing. There is a limit to how large these subscription content networks can get before it must fragment.
- seanp2k2 5y ago>” Just like no one wants to think about spending money each time they turn on a light in their house or take a shower?” No, in the same way that Netflix is different from renting movies on Amazon Prime. It’s SVOD vs TVOD vs something like youtube without a subscription that is AVOD. There’s a reason that Spotify is hugely popular while it’s a lot less common to buy an album on iTunes. It’s the reason AYCE KBBQ and buffets are popular. Even if pay-per-use ends up cheaper, flat-rate subscriptions are easier for people to see value in, and they can consume without concern of overages.
- billytetrud 5y ago> There’s a reason that Spotify is hugely popular while it’s a lot less common to buy an album on iTunes. It’s the reason AYCE KBBQ and buffets are popular. Yes, but the reason is not the subscription model. People like variety. The subscription model actually prevents good variety in a network. Read my comment here: https://news.ycombinator.com/item?id=27804511 https://news.ycombinator.com/item?id=27804511 . You bring up Spotify, but music is a fundamentally different product than news or videos. > flat-rate subscriptions are easier for people to see value in What you're saying is that people are getting duped into paying higher prices for lower quality content. While you might be right, this isn't a point in favor of the subscription model. > they can consume without concern of overages There is no reason a service can't help a user track their consumption and let the user configure a warning if they're hitting the configured limit. The subscription service is a very inflexible way of ensuring people stay within their budget. The real problem is that pay-per-view content is massively overpriced. Of course people opt for subscription services when for the price of 3 pay-per-view movies, they can watch 30 movies on a subscription service. If pay-per-view content were priced commensurately with subscription services, it would change the game.
- Wowfunhappy 5y agoThere's a lot less decision-making involved in turning on a light, or running the faucet—it always does the exact same thing. It's hard to know in advance if an article will be worth reading. People have also had an entire century to become inured to utility pricing. A better comparison might be data-limited cell phone plans. Even though spectrum and network capacity are legitimately scarce resources (especially in the early days), basically all of the carriers have been forced to offer some kind of quasi-unlimited data plan, where you might be throttled but you won't be charged overages, or you won't be charged on specific streaming services.
- billytetrud 5y agoThus the ratings system
- Wowfunhappy 5y agoYes, that works super well on Youtube, doesn't it? ;)
- billytetrud 5y agoYou're comparing apples to orangutans. Like, sure I get you're joking. But your joke implies that you think that youtube upvotes are at all comparable to what I'm suggesting. Its not.
- Wowfunhappy 5y agoWhy is it different? One reason I used Youtube is that they started out with a five star rating system, and moved to the current one because it worked better. Users only ever rate things either five stars or one star, when given the choice. The only exception I know if is Goodreads, where people seem to put a great deal of effort into their ratings (which I find super interesting). You might argue this is a better analog because the audience would have more overlap with newspaper readers, but people aren't going to put as much thought into a short article versus a long novel.