402 Payment Required, and why micropayments are doomed
piszek.com
piszek.com
I've pointed out before that the enthusiasm for micropayments is from people who want to collect them, not those who want to pay them. Unlike credit cards, which, when introduced, were easy to sell to buyers but hard to sell to merchants, since the merchant pays fees.
Cryptocurrencies are not a magic bullet here. The technology for preventing double-spending is incredibly inefficient and slow. Transaction costs are high. Right now, a Bitcoin transaction costs at least US$5. It's been as high as $20.
That's right, it will all cost the same as a Bitcoin transaction.
And how about receiving a lightning payment? Well... The merchant need to have a balance so he can accept your payment(!) Yes, that's another Bitcoin transaction right there. Or if you're lucky you can find someone who'll give you liquidity (for a fee of course).
It's debatable if micropayments will ever take off, but lightning never will.
I don't think any of them make micro payments easy right now, but throughput is not a real problem until transaction volume goes WAY up.
What is bitcoin's throughput? 2MB every -10 minutes-? That's less than 3KB/s. Let's not do this whole dance where the real numbers never come up in all the fear mongering.
A single 10TB hard drive would take over -92 YEARS- to fill up at that rate. And again, most people don't ever need to sync with the chain, they let servers do it for them.
What is your point?
Yes, there is a risk of double-spending, but since this is a small amount, the cost of the attack may be higher than the benefit. The payment processor can also deploy more full nodes in different geographic locations to monitor and detect potential double-spending transactions in the transaction memory pool.
And then a million fraudulent microtransactions will add up. Of course, that would never happen in practice, because the Blockchain cannot handle a million transactions in any reasonable time frame.
If the attacker sends the legit transaction and double-spending one at the same time, it will be detected due to the few seconds wait by the payment processor.
and in fact it is quite common to accept "double-spending" since that is really the only way to increase fees to push through a "stuck" transaction. Otherwise you have to wait an indeterminate (potentially forever) period of time until it exits the mempool of all network participants
[1] https://bitcoin.stackexchange.com/questions/81624/can-two-or...
This page (https://github.com/ethereum/wiki/wiki/Design-Rationale) contains a good overview of the UTXO vs Account model with each's strengths and weaknesses.
There is a nice article which hopefully somebody can link that explains that visa was a success because merchants reduced their costs, too. They had to do informal credit accounting for all their clients who only paid for the purchased goods once a month. With visa, that accounting position, which could be a full-time employee, wasn't needed any more. Furthermore, they reduced their risk of clients defaulting on those micro loans.
There is also a time savings. You don’t need employees counting and recounting the cash with witnesses when the cashier changes.
Cash needs to be physically secured. A cash register full of credit card receipts is pretty worthless.
You're right about crypto not being a magic bullet though.
I'd suspect each BTC payment is actually $60-70. When block rewards end, someone's gonna have to pay that.
Ignoring the fact that's going to occur in 2140, what you've stated is simply not true, users decide what fees to pay, no one else determines the price.
I send zero fee tx's regularly. If you are in no rush the cost ranges from tiny to non-existent sending money across the planet.
It's never going to be used for micro transactions without a second layer but the whole premise of this comment is out of whack with the reality.
The actual cost of a transaction can be represented as the following equilibrium:
(E * pkWh) = F + (R * pBTC)
- E is the average amount of power required to process one transaction, a function of hash rate (~700kWh)
- pkWh is the average price paid by miners for power (~0.08 $/kWh)
- F is the average fee paid for a transaction directly (~5)
- R is the average block reward per transaction (6.25/3500)
- pBTC is the market price of a Bitcoin (~9500)
If (F + (R * pBTC)) is ever less than (E * pkWh) as happened after the halving, something's gotta give. Either the fee has to go up, the price of a Bitcoin has to go up, or the hash rate will drop. The halving demonstrated (3).
Make no mistake, though, the cost of a bitcoin payment is ~$50, you just have to distribute it over the parameters.
You didn't think miners were given you "almost free" transactions for their health or something did you? Your very-low transaction fee is the tip of a giant cost-burg.
The only way it doesn't waste inordinate amounts of power and support a reasonable number of transactions is if you pretend.
Waste? It's evidently not waste, because miners are compensated. Missed opportunities for that energy to have done work on something else, or at least done more efficiently so the surplus could do something else? Perhaps, but that can be said of anything. I'd instead criticize the far greater missed opportunities of not coating deserts with solar panels, not peppering the world with nuclear reactors, not having satellites beaming microwaves down, etc. etc. We'll never become a Type 1 civilization with attitudes suggesting we lower global energy consumption or try to allocate it only in approved ways.
Well, I mean, if you're trying to launder money or manage a ransomware product, PayPal and Visa aren't great candidates.
> Waste? It's evidently not waste, because miners are compensated.
People are compensated for all sorts of wasteful garbage. I could pay some dude to dump gasoline directly into the rain gutters. Or move a mountain 6 feet to the left using only a shovel. That doesn't instantly turn it into a productive activity.
> Missed opportunities for that energy to have done work on something else, or at least done more efficiently so the surplus could do something else? Perhaps, but that can be said of anything.
There's no massive energy surplus. If we turned off the 55TWh/yr that's being burned on multiplayer Excel, it would stop a non-trivial percentage of our greenhouse gas emissions.
> I'd instead criticize the far greater missed opportunities of not coating deserts with solar panels, not peppering the world with nuclear reactors, not having satellites beaming microwaves down, etc. etc.
Sure, I'm down. Sounds great!
> We'll never become a Type 1 civilization with attitudes suggesting we lower global energy consumption or try to allocate it only in approved ways.
Not with that attitude :)
To me, it's like there's a giant smoldering tire fire in the town square, and everyone's walking around praising the tire fire for the nice warm glow it creates.
Any delta is due to the 45% drop in hash rate after the halving. Extensive methodology is provided.
None of those get easier with micropayments. If anything they get more difficult.
It might be just out of my ignorance for what the downsides are, but quite often I do wish I had a way to make micropayments, and I feel like lots of others do feel similarly.
and if you just leave one big channel with lots of money in it, then you end up with this fun situation! Hope you don't have a power outage... ever.
https://coinspice.io/news/bitcoiners-warn-after-user-documen...
BTW a follow-up to the story you link is this one: https://cointelegraph.com/news/guy-who-lost-4-bitcoin-has-re...
Sample size of one, but I do!
I frequently come across paywall'd articles I'd like to read. I'm not willing to add more subscriptions to my life, but it would be great if I could pay 25¢ just for access to the one article. If the payment process was easy, I'd use it a lot!
If such a hub does this too much then they'll have trouble getting others to commit funds to a channel to them.
More likely users won't have to deal with all that because there will be micropayment wallet providers that can aggregate the transaction costs of buying bitcoin and maintain large value LN channels.
And this is why most publishers will use subscriptions for their normal content, except some dubious ones who will use clickbait headlines to attract one-time buyers.
Of course, any reasonable solution would take, say, $10 from you, and allow you to just press "pay" anywhere a paywall is encountered. Of course it would just auto-pay for articles on sites you put in a whitelist, within a limit you set. All the clearing is done in non-micro amounts, say, monthly.
Such solutions do exist. The problem, of course, is the buy-in from site owners. I suspect that running better ads just has a higher ROI.
The only thing lacking with UPI is proper integration with merchants; large retailers rely on weird locked in services like PhonePe etc. for integration with their billing systems, but the rest seem to use a QR code and manual verification of payment.
My local grocery store (a small mom and pop operation) uses a QR code system with a speaker the recites out loud any new transaction so that both the merchant and the customer can verify it easily.
And besides, integration with merchants is not the job of UPI. It's the job of the application implementing UPI.
Consider me one of the mythical latter people. I really really really want to support quality journalism, but I don't spend enough time on any one news site to justify a monthly subscription (most of which are fairly expensive). I would be totally fine spending sub-$1 per article read, though, if that were an option.
I understand why it is not an option, but that's not really the point. If it were, I'd do it. Hell, I'd pay 50c per article just to avoid the annoyance of finding a paywall workaround.
Many CCs are great for micropayments, its just bitcoin that sucks - it shouldn't be that popular.
0-conf is also reasonably safe for small amounts. We must always remember that credit cards are perfectly acceptable, despite them only being "confirmed" after several months (so they cannot be reversed any longer).
This is true of bitcoin. There are better cryptocurrencies which have already solved these problems completely. Have a taste on https://nanospeed.live/ to do a fee-less live transaction in milliseconds.
It is pretty transparent to the user, there is no per-transaction fixed cost (though the margin taken by Ad provider is pretty huge), and some people are living pretty nicely with it.
I agree with the various points this article makes, but it doesn't mean micro-payments are a no go. Just make it as seamless as ads!
The article mentions subscription, but doesn't mention the possibility of subscription-based micro-payments?
That's what scroll.com, and I believe there solution has none of the issues mentioned in this article, even though it would still make sense as a 402 Payment Required! Of course you can't generalize scroll.com everywhere, for many reasons. But I believe /that/ idea is a start of making micro-payments a thing.
https://www.paypal.com/us/smarthelp/article/what-are-micropa...
https://www.paypal.com/us/webapps/mpp/merchant-fees#fixed-fe...
Without this, my bootstrapped company that employed over 15 people would not have been possible.
0.05x + 0.05 = 0.029x + 0.30 : x=$11.90
Fees would be:
$0.10 vs $0.33 on $1
$0.15 vs $0.36 on $2
$0.30 vs $0.45 on $5
$0.55 vs $0.59 on $10
$1.55 vs $1.17 on $30
$1.55 is not a micropayment.
Been doing that for around a decade. Make a substantial fraction of my income from US$1 payments.
The stories about PayPal freezing everything on a hair trigger are indeed worrying, but after using PayPal to collect the majority of my income for 12-15 years, my attitude is that these stories are just another example of how the internet magnifies the bad/abnormal and buries the normal. There's no doubt that they happen, but in the context of the total amount of activity that PayPal is involved in, my sense is that they are a tiny, tiny fraction.
In the macropayment land, a competitor could come in and offer 45c to process $10 instead of PayPal's 60c. No big deal.
In the micropayment land, a competitor could come in and offer 10c to process $1 instead of PayPal's 33c. MUCH bigger deal.
That said, I'm not in agreement overall. I think there are likely many things that can be billed in "sub-minimum-transaction-increments," it's just that they only become worth charging when they exceed the floor. So I guess in my head most of it is tracking "microdebt," and only sometimes is combining enough of the outstanding "debt" enough for it to be worth a payment.
It's basically a permutation seen in every game with (much hated) in-game currency. Maybe there's dirt cheap things to buy, but you still have to exchange to IGC at a $1:n floor.
There's some overhead at the sending/receiving institution as well, but it has a similar order of magnitude (0.0005-0.005€/txn mostly depending on your scale because of the fixed costs); the total cost for a proper interbank transfer is less than a cent if you do things right. The price is mostly determined by other factors, not the cost - it can be (and often is) zero for competitiveness purposes, recouping the costs from other services, or it can be something like the same 30 cents as the original article lists if someone is willing to pay that.
Chargebacks and distpute resolution and all that stuff is just something people in the us rely on, because using credit cards makes fraud easy. If I use my chip and pin card, then I can be sure that I did the payments. If I rely on a number printed on my card that I have to give someone, then fraud is just too easy.
I wish all online credit card transactions would use 3D Secure, but for some reason they don't.
The same for when people complain that european bank cards don't have any cashback or other "free" stuff associated with them. People don't realise, that they themselves are paying for that, no? I'm really glad the EU legislated caps on card transaction fees, because moving some bits around can only cost so much.
Debit cards come with chargebacks. Chargebacks are a feature provided by the payment networks themselves, so it doesn't matter whether you're using a credit or debit card.
> I wish all online credit card transactions would use 3D Secure, but for some reason they don't.
You realize it's the same kind of “free stuff that comes with a hidden cost” you are complaining about in your rant don't you?
Newsflash: perhaps I'd like to support more than a couple sites. Those coffees add up.
Also, I don't do Starbucks coffee. It's overpriced.
I'm happy to do it. It's worth it to me to support people who are often working full time to produce quality content. But it's no surprise that only a microscopic fraction of people do. It would be nice if a higher fraction gave a tenth as much.
It's absolutely a fork in the eye to record labels and AdSense. :)
Would you pay $5/month indefinitely to every site that has an article linked to on HN that you read? No, didn't think so.
Would you pay 5 cents per article read though? If the technical possibility to do that existed?
I'll never subscribe to the new york times because I'm not even on the same continent and I don't care about their news generally.
They could get some pocket change out of me when HN links to them though.
If you removed those two assumptions, micropayments make sense.
Crypto currency actually can solve those problems elegantly. You can have low transaction costs. And, transactions can be anonymous and irreversible (depending on the currency).
Critics will rightly point out those promises are not always kept (ethereum gas prices, for example), but they are the goals. Lower transaction fees, I would assert, are not the goals of PayPal.
The problem is not technical. It's that there isn't mass adoption. If a well funded startup did a "Fresno Drop" they might be able to overcome that. That would be interesting and could be very effective.
https://99percentinvisible.org/episode/the-fresno-drop/
Full disclosure: I'm working on such a system and have a patent pending on the process.
This is a misunderstanding of how mining works. It takes virtually zero energy to create a block (ie. commit a transaction). However, it takes massive amounts of energy for your block to be considered valid. This is because each block also comes with a block subsidy (currently 6.25BTC), which many miners are competing for.
With cash you don't have that problem - the transaction is physcial and immediate. But I doubt any micropayment system will get approval to transact money without anti-money laundering laws and transaction reversibility.
Streaming music consists of a single transaction that some platform logic would then divide behind the scenes. The consumer isnt forced to make multiple payment decisions. If and when they are, it probably won't work. Even having to pick and choose what portion of some pool of funds one wants to give to which creator creates cognitive friction that I don't think we've seen solved yet.
Humble Bundle would probably have good stats on that: how many people bother to alter the default proportions of the devs/charity/HumbleBundle split?
This is why Netflix is so successful, but Apple's pay-by-the-episode model is far less engaging, even though I wager for most people it'd be much cheaper.
In Europe, this is not the case. The EU introduced a cap of 0.2% of the transaction value for consumer debit cards, and 0.3% for credit cards for the amount that banks can charge each other. So as a retailer, you can get 1.6% or better rates here, with no 30c minimum. People use credit cards every day to get bus tickets worth around $1-2 - that would be like 25% of revenue if there was a fixed 30c amount or similar.
I still can't figure out how Amazon's Prime/Chase card can give me back 5% of everything I spend at Whole Foods. It's absurd.
For one, it requires a $120/yr Prime membership to get that rate - the non-prime rate is 3% (which is seen on many cards). I have an Amex with an annual fee that gives 6% back at all grocery stores (including Whole Foods), though I also wonder how that can be profitable for them. The break even point where you earn more on the 6% card after the annual fee compared to the no-annual fee 3% card is only a little over $3000/yr (~$60/week), which isnt much even for someone single.
Probably the same they would feel when a national ID card scheme or a single-payer healthcare system like in literally all other Western countries would be introduced. For a society that takes pride in unrestricted capitalism to eliminate inefficiencies there seems to be a widespread acceptance of utter waste in these areas.
The Sainsbury's card earns "1 point for every £5 spent elsewhere", 1 point is 0.5p, so that's worth 0.1%, and can only be spent at Sainsbury's or certain other vendors.
This was to be my next company, and it was effectively banned.
We have all of the tech to anonymously demand a real-time irreversible penny payment for a pageload, a download, or an API call. Pervasive, suspicionless financial surveillance demands of the government have made the resulting businesses mostly impossible.
Why not use “credits” where buyer makes a single payment and then uses those credits to make smaller transactions
For anything that hopes to become as ubiquitous as HTTP, it needs to be standard, decentralized and distributed, and extensible, a system where anyone can stand up an endpoint without centralized approval. I think the governmental stranglehold on payments in most of the developed west is going to be a hard barrier to such a thing ever developing, and instead we'll see the continued growth of centralized chokepoint platforms like Facebook, Medium, Uber, PayPal, Square, et c.
It's a real tragedy, because all of the tech is now in place for a total any-to-any model of content, APIs, and payments. It's just illegal.
(And not even "a little bit" settle-your-way-out illegal like web spidering under a super strict interpretation of copyright law, but SWAT-machine-guns-in-your face, "terrorism and child porn and money laundering!" pearl-clutching dozens-of-years-in-jail mega super illegal.)
Because that's, let's face it, is exactly what such a system would be used in a non-negligible way. Bitcoin got a lot of early adopters with Silk Road and the likes.
You'd also be crazy if you thought that e.g. PayPal is not used "in a non-negligible way" for the purchase of controlled drugs or child pornography or sex trafficking victims.
Anything widely used by millions of people is going to be used for the kinds of payments that millions of people make, whether legal or illegal. That's not an argument against the system existing.
Illegal drugs are a major source of funding for terrorism, e.g. Afghani heroin, or the Latin American cocaine gangs.
> Anything widely used by millions of people is going to be used for the kinds of payments that millions of people make, whether legal or illegal.
It also creates a paper trail which any criminal worth his salt wants to avoid. I have never heard about Paypal being used for child porn or paying a trafficked hooker, the latter is a purely cash based business anyway..
Had that option ever arisen, I'd have 'sent' x cents directly to the account of hundreds of creators. Because it was simple and person-to-person. Now and then, I'd have 'sent' x dollars to -someone- who'd created something of extraordinary value ... to me.
In my view, the rewards collected by brokers are usually not deserved, and the rewards for streamers highway robbery. A 'facilitator' at cost+ (limited) is a different ballgame.
The idea - simple, direct, voluntary, user-chosen 'rewards' P2P - never became a reality. I regret that. A cent, a nickel, a dime - anyone could play - from 100,000 people in a year (or even longer-tail) amounts to a very tangible reward, and valuable feedback. I don't need anyone telling me what something is worth to me; the creator does.
They have a browser add-on that does streaming micropayments directly to creators as you view the content.
"Direct to creators" means from your (financial institution of some type) account direct into the equivalent for the creator. No middle-men, close to zero fee, no cut, no membership.
I think the system will need to allow a user to allocate a flat monthly total they're willing to spend on micro-payments. Then, as the user accumulates more and more people they want to pay, the total gets split evenly (or at a designated ratio) across all.
Sort of like the saying "give what you can, take what you need".
I think there are several opportunities here:
1) I think government needs to be involved so they can create a new "designation" so we can provide rules this new designation follows. Most obvious of which would be taxes.
2) I think as the world becomes more and more connected, there will be a real need for tooling to allow users to manage the 10's or 100's of thousands of people they're paying with micropayments.
3) Probably would be good for standards to exist, so interconnected, de-centralized systems can keep "subscribers / payers" up-to-date with the latest from all applicable platforms, with all of the people they're currently paying.
Tons more, but not sure anyone will read this so won't go into too much detail...
As long as you can't do payments well under $1 things are doomed.
Just look at camming sites.
You buy 100tokens for 10€ and can "tip" people ¢10 if you like.
This works pretty well, and I think people would be happy to pay like ¢10-¢50 en masse for all kind of stuff if it was possible.
What can be considered a "high" rate in Bangladesh is going to be very different from Norway or Switzerland.
If you're doing business globally, the question of an appropriate "rate" or "price" is vexing, and isn't solved with glib western-industrial myopia in which you limit yourself the world's 0.001%. Unless of course that is precisely what you want.
A shorter way to say that is that something can fail thousands of times, but only has to succeed once. If the right technical, social and business factors come together, you get working micropayments.
Best movies of 2020 would be "pay 0.10 for #100-90 and then .10 for each #90-#80 etc". Stratechery would be "pay 0.10 for the above the fold part and then 0.10 for the rest".
Authors used to get paid in micropayments (by the word or books were serialized) so we don't have to guess about the effects. Victorian authors (whose works were all published serialized) stretched things out to the point where modern readers often find them unbearably long.
- Form a consortium (just like the competing banks did in the 50s to create Visa).
- Google, Apple, and Microsoft already have expertise in collecting payments and running services, which they can lend to the consortium.
- Each member creates a UI in their own browser for micropayments that is backed by the consortium.
- The UI has the ability to add funds (processed by the consortium) and set budgets and have little check boxes for "ask me every time" or "automatically pay if below 1 cent" or "allow up to $5 / mo on this web page" or lots of other budget options so that you don't have to think about it every time.
- When you go to a page, they can throw up a request for payment (which if you've already made a budget decision you don't even see as a user), you send a token, they pass the token to the consortium who moves the money from your account to theirs.
- The consortium pays out to web pages when they have earned more than $10 (just like Adwords!) and generates a 1099 or equivalent. They take a piece of pie at this point.
And now if you want an ad free experience, or want to use a site that requires micropayments, you must use one of those browsers. And the site owners can offer both. Send back a payment request, if it fails, show ads.
Heck, even if Google didn't want in, Apple, Mozilla, and Microsoft could pull it off without them (they still have about 38% of the market).
You can secure it using standard crypto. The browser generates a private/public key pair. The website generates a keypair. Both sends the public key to the consortium. When you make a payment, you generate a message that says "I pay foo.com 1 cent" and then sign it. Foo.com takes the message, wraps it up and says, "I want 1 cent" and signs it. The consortium verifies the signatures, makes sure the amounts add up, and then sends back an "OK" response if the user has the funds available.
They key here is that if those four agree on a UI and agree to a central payment processor, they have enough pull that website owners will want to use the service. Especially if there are no per transaction fees.
This also solves the onboarding problem for website owners. It pushes it to the consortium. Who can push it right into their parent companies. Google could automatically tie payment to your Google account, and so could Apple and Microsoft. It would be part of the UI.
In fact it would be a good workaround for Google in the new "do not track me" world.
People block ads because they are annoying, and trackers because they get no perceived benefit.
But payments provide benefit to the user, so they would be more willing to make the payment, thus providing tracking data. And for Google it would help cut Facebook out of the tracking because then they could encourage blocking tracking while accepting payments!
The legality makes it difficult but with a few hundred thousand you can get all the licenses. It's tough to find a partner bank but with any sort of funding it would be easier to get meetings.
You don't even need to start online. It's my world and what I'd want to do, but brick and mortar stores would kill for lower fees. Could start in a single state at physical retail and really cut down on the MSB licenses needed.
We mostly tried to sell to news/magazines publishers. The points that the article author brought up are all good points; but what we kept seeing was that it wasn't the consumers who were unwilling to pay, it was the publishers who were unwilling to try a new model. Whenever a site with quality content integrated with us, the percentage of visitors who paid was more decent than you'd imagine. What we couldn't do was convincing a sizable national-level publisher to use us. We had various niche content sites and some local region newspapers integrate with us and they actually saw a pretty good payers to visitors ratio. Some niche content sites even made pretty good revenue using our paywall product.
Large companies have absolutely no reason to partner with an unknown. Worse than NIH, is "why fund a parasite on top of our business model?"
There is a TON of context I am omitting in that post. In particular, I think that niche sites can run micropayments because early adopters WILL use them. But the economics of it means that you need to cross the chasm to the early majority in order to break even. But the early majority will be more conscious about these tiny amounts of cost.
Alternatively, I can be totally wrong, but this whole thread is an awesome discussion :)
[0]: https://lsat.tech/