The one thing I find slightly grating about links to Kagi blogs is the top comments are almost always "I use Kagi and it's great!" rather than about the content of the blog. And I'm a happy Kagi subscriber!
I get that this is an option, but what it really shows to me is how broken the model for journalism is. In my view you almost always only get good-quality journalism if you pay for it. (Somehow - e.g. the BBC is paid for by the licence fee, although I'm a bit despairing at the quality of the BBC these days, but perhaps that tracks the significant cuts to their budget...) I'm pretty sure that there's a technically viable solution to micropayments, but there are too many competing interests for anyone to settle on anything. We can't even agree on a single setting that says you don't want advertising cookies!
The biggest problem with micropayments is how credit card payment processors work. You typically pay a percentage fee. But they thought of everything, so you also pay a flat fee per transaction like 10 cents. So they can get their money from tiny transaction.
So for something like a 20 cent micropayment, you pay more to your processor than you pocket.
One solution is a universal token system or something like it. You deposit say, 40 dollars, then get 10,000 tokens or whatever. Like Java Beans from back in the day.
That's not how a micropayments system would work, and you don't need a token system.
It would work similarly to how my transit card works. I use a credit card to put some amount of money on the card itself, say $20, and then every time I use transit, the payment system just deducts $2.85 from my balance. Once my balance goes below a threshold I set, it charges my card again to top up the balance.
So instead of seven $2.85 charges, there's one $20 charge.
A micropayments system would charge you that $20 once, and then deduct 20¢ from your balance every time you read an article. For 100 articles, you've charged the card only once.
It would work similarly on the website side: the micropayments platform wouldn't send a deposit over to the website owner every single time a visitor paid 20¢; it would aggregate payments and either pay out on a timed schedule, or when some minimum dollar amount is reached.
You don't need a token system for this; denominating the "wallet" balance in dollars (or whatever the user's currency is) is fine. A token system might be useful in order to make pricing easier from the website operator's side, but IMO they suck because they hide the actual amount of money the user is paying. It sort of like when you visit a foreign country where you're not familiar with the currency, and so you see that something costs 1,000,000 foobits, and it has no meaning to you, because the exchange rate with your own currency is something like 1:4285.
Whatever the solution is they'd better get to it because I'm not going to subscribe to every site that wants 20/month just because it shows once on HN or in search results.
Your comment illustrates the tension very well, I think. Businesses don't generally want customers who aren't their core market; those people though, often want one slice of a business's offerings. This can lead to big distractions for businesses, as putting in the effort to serve the wrong market is a waste of time. The publishing industry really fell for this when the internet became widespread; they dreamed of huge audiences and big ad numbers. Turns out, that was not a model that worked, at all. The model that works is focusing on serving a niche, and charging a subscription fee, so you can get predictable revenue.
That for some reason makes me think of those sites where you can insta subscribe online but you need to call and wait on hold for 6 hours to cancel :)
You know what's missing from this model? Back a million years ago when we all had pet dinosaurs, you could pay for a magazine issue on the street if it seemed interesting. And maybe if you found enough issues interesting you subscribed.
The only way that will work is if each publisher has their own token system, otherwise you have created a new middleman who will want to take their own cut. That middleman will also have to deal with similar problems as payment processors, such as fraud. Perhaps it will be to a lesser degree, because it would be a less valuable target, but it would still add to costs.
Then you run into all sorts of other problems. How do you price articles? I remember running into academic publishers charging the same for a 1 page editorial as a legitimate scientific paper. And how do you charge for the article? Is it 1 token per view or 1 token for unlimited views? If it is the latter, how do you ensure the buyer doesn't lose access? On top of all of that, how do you avoid incentivizing the wrong things? A 1 token article may be 10 times more popular than a 5 token article and cost 1/10th to write, but adjust the price of the 5 token article to 10 tokens and you may cut the readership in half. On top of all of that, the consumer probably doesn't want the cognitive burden of keeping track of all of this stuff.
But the biggest reason why it won't work is because publishers don't want it to work. They have demonstrated time and time again that they would rather sell monthly subscriptions than access to articles.
Of course they'll take their own cut, but that's fine. We accept (sometimes grudgingly) that credit card issuers and networks get their own cut, so there's easy precedent here.
> How do you price articles?
The same way anything else in life is priced: you price at whatever the market can bear.
> And how do you charge for the article? Is it 1 token per view or 1 token for unlimited views?
Also precedent there: online video rental/"purchase". You can decide what model you like better, and see if people want to pay what you ask. You can even offer both, just as is the case for many videos on Amazon, for example.
> but adjust the price of the 5 token article to 10 tokens and you may cut the readership in half.
Yes, that's one of the pitfalls of setting pricing, that every business has to deal with. Note that this problem is present with the monthly subscription model too: half of New York Times subscribers might cancel if the subscription price doubled.
That's a general problem with pricing: determining how much you can raise your price without losing an equivalent (or greater) amount of revenue due to losing customers.
(As an aside, sometimes doubling your prices and exactly halving your customer base is desirable. Fewer customers to deal with at the same overall revenue level might be a net positive. But of course losing a single customer then becomes twice as bad as losing a single customer at the previous pricing.)
> On top of all of that, the consumer probably doesn't want the cognitive burden of keeping track of all of this stuff.
This is the real issue from the consumer side. People like all-you-can-eat plans. They would much rather make a payment decision once, and then get "free" articles to read on an ongoing basis, than have to make a payment decision for every single article they think they might want to read.
> They have demonstrated time and time again that they would rather sell monthly subscriptions than access to articles.
Yup. Recurring subscription revenue is stable and predictable; one-time per-article revenue is not.
> Of course they'll take their own cut, but that's fine. We accept (sometimes grudgingly) that credit card issuers and networks get their own cut, so there's easy precedent here.
In that case, I fail to understand the benefits of using a token system over a credit card. The publisher is loosing a cut to another third party. The consumer is dealing with an extra layer of complexity, and paying more to offset the cost to the publisher.
> This is the real issue from the consumer side. People like all-you-can-eat plans. They would much rather make a payment decision once, and then get "free" articles to read on an ongoing basis, than have to make a payment decision for every single article they think they might want to read.
Do you blame the consumer? Not only do they have to decide whether the article is worth reading, they have to do so sight unseen. While there are some who simply want all-you-can-eat plans, there are plenty who would be happy to pay by the article if they knew the article was digestable (at least in cases where it would be cheaper to buy one article they are interested in, rather than for a month at the buffet table).
Each card transaction comes with the ability for the buyer to do a chargeback. How would you price this feature, if not per transaction?
The solution here is a government operated utility that works like cash to reduce transaction costs, by not having the protections a card transaction has.
Stablecoins are a pretty clean solution to this. Negligible fee, sub-cent transactions supported, no need for kyc when paying. Fill up your browser periodically, and pay from an embedded wallet.
Lots of smart people have tried to implement micropayments like this, and it has never worked.
Publishers would rather have the chance at $10 than give up the chance at $10 for 10c. And readers don’t want to be thinking about their credit balance every time they read an article.
It's a real shame the grifters took over cryptocurrencies. I'd love to have a pre-minted ZKP USD style coin. Uncle Sam can take a small transaction fee as consumption tax, people can get wallets as they please. Fully private, semi-decentralized (you don't need full for a government currency), and gov even gets to "tax" transactions on illegal goods (Al Capone). But the two major flaws are that kinda kills a lot of the CC company businesses and it's not great for a country that wants to spy on its citizens. On the flip side, my nightmare is they implement that same system without the ZKP and now your grandma giving you lunch money gets tracked. Weirdly we want with grifting instead. Weird move...
Everyone knows about Monero. If someone doesn't we can count on someone ensuring we do. But I can say with certainty that anyone who knows what a ZKP is knows about Monero even though Monero doesn't use ZKPs
Used to be you would trade a yak or a horse for something. Then you could trade little pieces of metal. In the future, why not something smaller? I mean, people are trading virtual things (like an app). Why not something small for something as small as a website view?
I'd argue it's still a product issue, one that will eventually be solved.
The reason a payment is too much friction today is because it's surrounded by a raft of cruft: is there an extra "processing fee", am I signing up for a recurring charge and didn't realize it, I have to enter my card details and address, or I have to log in to some system, I have to OTP with my bank, etc.
It hasn't been solved yet, but someone will solve it to the point that it's a single swipe to buy an article for 25c, and it'll work. I would argue the NYT is in the best position to do so, but they are addicted to their scummy business practices where it's hard to cancel, you pay "$2/mo introductory rate" that suddenly escalates to $40/mo, and so on.
> The one thing I find slightly grating about links to Kagi blogs is the top comments are almost always "I use Kagi and it's great!"
Couldn't agree more. I don't know how others perceive it, but I find myself becoming more and more wary of broad endorsement. Maybe it's the amount of astroturfing you see on the internet, but I can't help imagine if there is a concerted effort behind it. When someone shares something meaningful and insightful about a product, there's no better endorsement of it. But low effort platitudes are probably doing more harm than intended.
I use kagi. The results quality is pretty good, but I probably use the uprank and downrank feature all the time. Being able to report stuff as AI and get little indicators that a site is AI slop is good.
As an example, serious eats is my go to recipe site, and I've got it tuned so it returns higher in my results.
I also really appreciate I can write my own custom css, and style every box and widget on the page.
And marine most of all, I love that their AI tool will not show up unless you end your query with a "?". Meaning I'm explicitly opting in to using it.
My last reason for liking it is ideological. No ads, no manipulation because of advertising tech. Just ... the stuff I searched for.
I do not like the image search page, however. It feels clunky in a way I can't put my finger on.
There's a few other things I really like, but it's hard to type them all out. So I often go, "kagi is great, love kagi."
You mainly get this because honestly I was already happy with the existing product. The main focus of this post is not something I was really looking for and is a nice addition but I would still be using the existing product regardless.
Basically I just want search that I can actually filter are returns relevant useful results.. This adds an extra facet that I might filter on from time to time but the main ones, which are time and exact strings, are the reason I use the product.
I know what you mean, but it comes across - even to me, a paying customer who does really like Kagi - a bit like talking to someone involved in an MLM who can't wait to steer the conversation around to their product. I'd just prefer discussions to stay somewhat on topic. There's something interesting in this announcement in that, in my view, it deliberately omits an attribute that's a mild proxy for high-quality journalism, and I always find it a shame when the topic at hand is drowned out by generic boosterism.
> I'm pretty sure that there's a technically viable solution to micropayments, but there are too many competing interests for anyone to settle on anything.
If I could easily use ApplePay to pay 20¢ to read an article, without any tracking or other strings attached, I would probably do that to avoid the whole song and dance of archive workarounds. It’s those competing interests that really get in the way. Giving an email, signing up for an account, storing payment information… money aside, those costs are too high for what I’m getting from a site I don’t read regularly.
It has been tried e.g. https://en.wikipedia.org/wiki/Blendle but it just doesn't work, even a tiny fee will already deter most people from "buying" an article as it will always feel as a loss. Whereas with a subscription you don't feel bad in the moment for losing money but feel good for using your already paid subscription.
Publishers want them and many have tried. It is the other way around, people don't want to make micro payments. Some say they do, but the evidence proves they're wrong.
I think the majority publishers have tried it in the past 10-15 years.
If you are a publisher, something which gains widespread attention might happen in your region, or an article might become viral for other reasons. You know that a lot of people would like to read that article and not become subscribers. So if you can sell access to an article 50 000 times for 1 dollar, that's $50 000 in income you wouldn't otherwise have.
But unfortunately, very few people are willing to pay a dollar or 50c for an article which they are desperate to read.
The more I think about it, the less sure I am that the model for journalism is actually broken. For some unknown reason, people do not want to subscribe to news media, but they'll happily pay for Netflix, Disney, HBO and Spotify.
The comment is always: I want to be able to pay for this one article, but streaming doesn't work like that. I can watch a movie on Netflix and pay €25 per month, then watch what ever else I want. Or, I can pay €13 and watch just that one movie on AppleTV and nothing else. Newspapers are more expensive, typically, because their subscriber count is way lower than that of Netflix, but their cost remain roughly the same.
You can't say: I want to pay $1 for an article, unless that article is massively popular, that's not going to cover all the other less popular, but important news for that day. Journalism requires a stable income to produce things of value. Investigations may that months or years. Using the cost of the Netflix/AppleTV example you should either subscribe, or pay around half the monthly subscription per article.
So no, the revenue model for journalism isn't broken. Our ability so recognize the value of good journalism is. It's easier and more profitable to write an "article" about Kim Kardashians oatmeal with a few dusin ads slapped on, than it is to do a multiple page article about the governance problems of the UK, which you need to sell, because advertisers don't want to be associated with politics and the target audience big enough anyway.
So, we can't have good, independent journalism, because the amount of people who care aren't enough to cover the expense or they want to shop around and only pay for the topics and views they care about and agree with. Micro-payments is the complete wrong way to go, we need more people to start subscribing to newspapers again and the papers needs to take that money and funnel that money directly into actual journalism.
> The one thing I find slightly grating about links to Kagi blogs is the top comments are almost always "I use Kagi and it's great!" rather than about the content of the blog. And I'm a happy Kagi subscriber!
Okay, new topic then. Have you heard of InstaPot? ;)
Apple news manages to give access to a variety of mainstream sources for a fixed price. Unfortunately there is no way to block the ads and without an ad-free tier it's a no go for me.
The foregone conclusion is that journalism is somehow helpful or valuable to me. I have never read a piece of journalism or "news" and thought to myself, "Wow, how useful! I'm so glad I consumed that."
And I sat here struggling for 5 minutes to think of a single instance. I couldn't come up with one.
So maybe the underlying problem is that journalism isn't nearly as valuable as it thinks it should be.
> We can't even agree on a single setting that says you don't want advertising cookies!
We can't, because the website owners don't want one. They want there to be no standard, because then it's more likely that some people will accidentally allow them.
>you almost always only get good-quality journalism if you pay for it
Often true, but sadly doesn't work in reverse - if you pay for it, it often still isn't good-quality journalism (although your mind will tell you it is, just because you paid for it).
> only get good-quality journalism if you pay for it
True for independent media but for corporate media the same stories end up in so many places and most paywalls can be easily removed anyway you aren't missing anything. It also feels wrong to pay for mainstream/corporate media in general, why pay for your own psyop, some PR/AD/NGO/lobby/agency/government already paid for the journalism why should I have to pay to be manipulated? I rather give my money to Democracy Now or Drop Site News.
I get that this is an option, but what it really shows to me is how broken the model for journalism is. In my view you almost always only get good-quality journalism if you pay for it. (Somehow - e.g. the BBC is paid for by the licence fee, although I'm a bit despairing at the quality of the BBC these days, but perhaps that tracks the significant cuts to their budget...) I'm pretty sure that there's a technically viable solution to micropayments, but there are too many competing interests for anyone to settle on anything. We can't even agree on a single setting that says you don't want advertising cookies!