Ask HN: Is there hope for micropayments?
I've always heard that narrative that micropayments simply won't work and that people don't like it.
Recently, I've found myself paying for OpenAI's GPT-3 playground app in micropayments format. I use it for small tasks here and there and pay around 2-4$ a month. It feels totally fine and now I don't see why I wouldn't do the same for my Google searches or usage of maps.
Does anyone else have examples of services/products that they pay for in micropayments?
154 comments
[ 4.5 ms ] story [ 190 ms ] threadI am prone to microoptimizing things, and I would absolutely spend an objectively excessive amount of time and mental energy trying to avoid doing another search to save three cents.
One alternative that could work, is to make a subscription service, and then reimburse the unused amount.
Like make people pay 20$/month, then do your micropayment system in the background, by the end of the month, refund them to difference if they didn't exceed that monthly value.
It creates a positive feeling instead of a constant background stress with every interaction.
This is indeed the theoretical expectation I had regarding micro-payments, but I was positively surprised to find out that I didn't care at all and in the end, I'm always positively surprised by the GPT-3 bill (I got much more value out of it than they charged me).
It's easy to get users if you offer a paid service for free, but much more difficult to get people to pay for it.
(They have done surveys where people have said they would pay, but this is different to actually putting their credit card in and putting their money where their mouths are!)
I just also think there will be a non-zero proportion of people who absolutely promise and intend to pay, then won't put their credit card details in.
I also think there is a non-zero proportion of people who will search differently/less when every search costs them money!
Edit: similarity, when I'm building a side project now O avoid everything with a fixed cost as much as possible. $5/month for a relational DB? nope, let's see how I can store data and only pay for storage and access.
I don’t have that many subscriptions, so I’m not getting any fatigue ;)
If they have a monthly fee I'll hesitate since if I reduce my usage for a unspecified while I'll never know if I should cancel it or not. It requires being able to predict the future, which is a cognitive cost I can't overcome.
I used to optimise things excessively (eg. when grocery shopping, phone spending pay as you go) - then I stopped when I realised it's not worth the effort compared to what I make in a day.
I'm also annoyed I can't get a cell plan in my country that just charges me a flat rate per GB of data I use. Instead I pay a flat monthly fee that includes "free" call minutes and messages that I never use.
In both cases I'm sure I'm overpaying and cross-subsidising people who consume a lot.
Personally I prefer flat rate for unlimited everything if we're talking cell phone. It's not like I'll cancel it from month to month.
But for $random_site that is linked to from HN... no way I'll pay for a subscription. But I wouldn't pay 0.25 either for one article. Give up on the whale pricing and charge 1 cent.
This use case also needs a frictionless way of paying. No way I'm making an account for each site.
I also use my unlimited minutes to answer all spam calls and immediately turn my microphone off, just in case they pay by usage, but that is somewhat less critical :)
I can totally see the other side of it, though. My brain is just structured such that the money-to-mental-effort tradeoffs don't work out for me.
Micropayments are not "pay $2 or $3 per month" (either flat-fee or average), there are plenty of successful examples of that. Micropayments are "pay $0.004 or $0.02 for reading this article or using this service now". As Shirky noted, the problem with micropayments are not the technology or transactions costs. The problem is that they require a huge amount of user time and attention, relative to the value of the item being bought. This has not changed in the past 20 years.
[1] http://mx.thirdvisit.co.uk/2002/10/04/theacaseaagainstamicro...
Unless I'm misunderstanding OpenAI's pricing format for "bulk tokens" [1] prepaid for by credit-cards, this isn't web micropayments in the way many others think of that term. Likewise, buying audiobooks via Audible Credits by prepaying $229.50 to get 24 credits isn't really micropayments.
Some were hoping for internet micropayments at the granularity of transactions rather than pre-paying a large upfront amount to be parceled out in smaller amounts later. What prevents micropayments at the fine granularity is the high VISA/Mastercard credit-card transaction fees.
E.g. many want to micropay 2 cents for a news article without prepaying $10/month or $120/year to a "media aggregator" taking a percentage cut of that.
[1] https://openai.com/api/pricing/
I could imagine The Economist charging me at the end of the month for the 20 articles * 10 cent I consumed.
Let's say my site lets people purchase news articles for 1 cent. I get 4 readers in Chicago who each purchase 1 article per week. That's 208 transactions for a total of $2.08.
That's over the threshold in Chicago's state, Illinois, for creating an "economic nexus". Their threshold is $100k sales per year or 200 transactions per year. So now I've got to register with Illinois and do regular tax filing. I may also have to collect more data than I want on my readers there, such as address information, to figure out the right rate (rates vary by location within most states).
23 other states also have that annoying "or 200 transactions" condition on their economic nexus laws.
Compare this to if I instead only offer my news articles through an aggregator. Now when those 4 readers in Chicago buy my articles they are doing business with the aggregator. I don't have to deal with any Illinois taxes. Or any of those other 23 states' taxes.
That's the aggregator's problem. That's fine because the aggregator presumably has not just my 4 readers in Illinois but also plenty of readers of the other content they aggregate, making their business in the state large enough that the costs of dealing with taxes are comparatively small.
As a seller, there should be a government website with an API that I can hook into where it calculates all applicable taxes and does all that nonsense and all I need to worry about is providing the product or service.
Such simplification might include a single rate per state and a way to do a single filing that covers all states one sold items into.
I.e., something similar to how the EU handles VAT. Here's how we as a US company handle VAT in EU countries. First, we registered with the tax authorities in a single EU country. We are using Ireland. This took something like 20 minutes to figure out and was all online.
Then, once a quarter we upload a file to the that country's tax office that simply lists for each EU country how much we sold, what VAT rate we used, and how much VAT we collected, and we give Ireland all the collected VAT. This only takes a few minutes I believe.
Ireland then deals with sending each of the other EU countries their share of the VAT we collected.
There is an ongoing effort to do something like that in the US with the Streamlined Sales Tax (SST) project. Almost half the states have joined that. The deal there is that if a seller agrees to collect tax for all the SST states then those states will pay the costs for the merchant to use some third party tax service, such as Avalara, for this.
That's nice, but does not include any rate simplification. With simple rates, like in the EU, I don't need to hook my cart up to any external tax APIs. I just need an internal table of the per country rate. I use an external API to keep up to date when a country changes their rate (which isn't very often), but that can be a daily cron job rather than something in the live cart path. With the bazillion rates per state model of the US you have to hook your cart up to a third party tax API.
SST also doesn't help with the states that aren't part of SST. Of course the third party tax services such as Avalara are happy to handle those for you too, but not for free like they do with the SST states.
An EU style system with both simple rates and a being able to deal with all states through one free service is probably too much to hope for. But maybe Congress requiring states to join SST if they want to force out of state sellers to collect would be achievable, and that would be way better than the current mess.
The only time I'm touching actual micropayments is when someone sends me $0.50 through Brave.
I wouldn't pay for my searches. I don't think I would pay for maps either. Plenty of alternatives.
Of course, we don’t do that anymore. Is Rand McNally still around?
I think Rand McNally is still around. I think I bought a topo map they made just a few years ago.
In the online map world, there are free alternatives if any one of them started charging.
[1] https://standards.openbanking.org.uk/
Saying that those costs should be coming down today compared to a decade or two ago because there is also much better surveillance, more bugs and backdoors/zero days have been found and with the global inflation stalking the planet, ironically it could be the inflation that helps make micro payments become more popular.
This is why advertising fees pay for search engine results and its an excuse to intelligence grab from businesses through their advertising criteria so its not just human end users who get surveilled.
I think the next step would be something similar but for blogging / text content.
PS: I myself wrote a RSS reader & podcast client that I released on the Umbrel app store. As you listen to podcasts, you can stream donations straight from the app running on your Umbrel node using funds you have complete custody of. It is pretty amazing.
PPS: In the Lightning world, micropayments can be as low as a millisat, which is 1/1000 of a satoshi. Routing fees are also very low and transactions are instantly confirmed.
Bitcoin with Lightning finally makes micro/nano-payments practical and it's going to flip the Internet economy on its head.
To me this is the first and best example as of yet where I feel that a cryptocurrency/blockchain is delivering real value (for value). And indeed, I like it.
This website is not applicable in this case.
There’s no one doing OFAC screening or compliance checks on these transactions, so of course it’s going to be cheap.
I would expect headwinds preventing this from going mainstream.
Most of my servers run for a long period of time, but sometimes I do experiment with running a certain setup that I kill after a few hours or days. Knowing that I only pay per hour at the same low cost allows me to experiment more.
Those are not cryptocurrency micropayments, and the actual transfer aggregates to one accumulated, monthly payment.
Similarly Amazon charges by the byte of storage on S3, but bills monthly.
(We may have also been the first online music store to have Creative Commons licensing built in, as our launch coincided with CC's. I'll never forget at their launch party, when the nice, awkward teenage kid I thought was just some attendee's son got up and was introduced as one of CC's developers, Aaron Swartz.)
The thing that killed the momentum then is the same thing that still kills it - card transaction fees. Bitpass got around this by allowing you to buy Bitpass credits for like $3, which you could spend anywhere. It worked great for music, and Mperia was originally seen as a good gateway (and, frankly, loss leader) for getting people to adopt our system.
Alas, it never took off, and Bitpass's brilliant CEO and founder got sidelined by investors in favor of some ronin CEO from the ad world who bogged it down in awkward partnership deal negotiations until the money ran out. I'm still convinced, all these years later, that if they'd focused on the indie media angle, it could have taken off.
(I also wish that this band who played their first gigs ever at my coffeeshop open mic in Vegas called The Killers had put their record up pre-record label deal, as I asked them to. I think Brandon was down but their shitty manager told them not to, and later they sued him for being shady af, which I did warn them about.)
Never trust the suits, kids. Never trust their bullshit. And never give your heart to the business. It'll get broken every time.
(But even then still never trust the suits)
If only there were some permissionless internet payments system that has near-zero fees and realtime settlement that works across borders and has frictionless setup.
The real problem is not the market or the technology, it's that the state wishes to surveil and ultimately exercise veto over all payments in society. Lightning payments solve this perfectly and could across the whole web but it's illegal in the US to "transmit money" without doing all sorts of market-killing, expensive customer intrusion and surveillance (and buying expensive licenses in every state).
There's nothing technically stopping a browser extension from letting you anonymously pay a penny per request or something in one click to pass a paywall. It's just the state's incessant hunger to want 100% transaction surveillance and veto.
Yeah, I'm pretty sure that's not a thing. Bitcoin transactions can easily cost tens of dollars each, and Lightning balances are ultimately settled via Bitcoin.
Payment channels are a little complex so it's common for people to not understand them, but they do indeed solve this problem completely.
In EU there are also free and instant bank transfers.
CC payments can be reversed though (which is a cons in my mind, I don't understand why merchants should lose the money by default in case of a dispute).
Sellers can price in and insure themselves against fraud. The protections offered by CCs are that insurance for consumers.
Which, incidentally, is why it's the blockchain with the second highest number of users. Basically free and instant transactions enables seamless integration with the web.
[1] https://www.theblockcrypto.com/linked/144639/solana-restarte...
[2] https://u.today/solana-network-goes-offline-again-now-ddos-a...
While the 2nd part of your statement is on the surface true, it misses the point. You can gets hundreds of thousands of transactions out of a channel without ever closing it. Batching on-chain transactions is now easier than ever so you can close and rebalance hundreds of channels cheaply. You can very easily do automatic swaps between onchain and lightning channel balances. You can be the counter party to either side of these transactions depending on if you need inbound or outbound channel liquidity. These transactions can also be easily batched allowing you to unlock liquidity where you need it for a fraction of the price of a single spend bitcoin transaction.
There is still a long way to go to scale bitcoin to billions of users while keeping the blockchain accessible to all.
For further interesting scaling solutions in the works checkout.
- cross input signature aggregation: A way to super charge batching. - channel factories: a way for a lightning channel to be trustlessly owned by multiple parties. - MuSig: multisig that is indistinguishable from a single signer transaction. - federated chaumian mints: a way to do privacy preserving custodial wallets. - drive chains/space chains: a bitcoin merged mined side chain. - eltoo: new lightning channel upgrade that cuts the space complexity of channels from O(n) to O(1). - covenants congestion control: allows a sender to commit to spend while the receiver gets to choose when to receiver. Exchange withdrawals during high demand times while a customer can wait for a lower demand time to fully receive the coins.
Happy to answer any questions on these.
With anything cryptocoin, everyone on the planet can now analyze where my money flows to and from. That's decidedly worse than governments which need court orders to do so.
I don't believe this is correct for Monero, ZCash, and anything based on mimblewimble (BEAM, Grin).
[1] https://de.cointelegraph.com/news/privacy-coins-no-more-ciph...
Card fees are consequences of their pricing decisions, not marginal cost of transactions. And, the reason they have "pricing power" is the size/market position of their networks.
> browser extension from letting you anonymously pay a penny per request
Of course, and there is no need for crypto currencies. Paypal could have implemented this for twenty years.
Monero is the only payment method I would consider to be most likely anonymous. It has very low transaction fees (<0.001%) and completes a translaction in about 20 minutes.
Check out nano: https://en.wikipedia.org/wiki/Nano_(cryptocurrency)
Granted it is a cryptocurrency, but it is feeless and ~1 second confirmation.
I too thought it sounded like magic when I first heard of it.
Also, the browser extension you wish for exists, it's called Coil: https://coil.com/ but it simply doesn't have enough buy-in to work. The real problem with microtransactions has nothing to do about some bigwig execs or some deep state preventing it crap, it's simply that EVERYONE has to buy into it for it to work properly.
Well, the US could follow Europe here if it wanted - in 2015 we've been capping CC fees at 0.3% and debit card fees at 0.2% and at the same time capped per-transaction fees at 5 cents [1], and whoops, suddenly the acceptance of cards in Germany exploded. I can pay the 50ct fee for loos on train stations or 1€ for a soda on a vending machine with cards.
The solution to micropayments are not shitcoins, the solution is to squeeze the rent seekers.
[1] https://www.consilium.europa.eu/en/press/press-releases/2015...
Compared to the US's 1.15%~2.7% +10ct interchange (not including other fees!) it's still quite cheap.
https://www.zettle.com/gb/pricing
> The solution to micropayments are not shitcoins, the solution is to squeeze the rent seekers.
First, "shitcoins" and blockchain-based tech for p2p payments are orthogonal points. You can have cheap transfers without requiring specific tokens.
Second, have you ever considered that the best way to squeeze the rent seekers is precisely be developing a technology that disrupts their business and forces them to lower their margins?
Crypto is a multi billion dollar market, why hasn't a single solution for micropayments managed to achieve commercial uptake? Everyone would benefit if such a solution existed and was cheaper than Visa/MC.
Crypto having low transaction costs at anything resembling commercial scale TPS seems to be a myth. There's always some giant gaping caveat that goes unmentioned, like "lightning network Bitcoin" or "eth 2.0 with PoS will solve this".
Somehow that has to get paid, and it’s via higher interchange rates.
It was amazing all the indie music I was able to purchase, load up the wallet and spend micro amounts per song.
I think the situation may have improved, with some card processors like Zettle who only take a flat 1.75% of your total daily taking, which makes it a bit easier to do many small transactions.
Btw I am from India. Earlier,every time I see people swipe the cards in the retail stores, I felt sorry for the shopkeepers as they had to bear the 1.5%-2% as transaction charges of the total amount paid.
It all changed after the launch of UPI payment system, now I can just pay 5-10 INR(7 cents to 13 cents) to a shopkeeper without feeling the guilt of payment charges. It has been more than two years when I used my last VISA card payment transaction for any retail shopping. I think other countries should also develop such not-for-profit payment network system that can be used to pay efficiently digitally without making any side the victim.
Pix is not even 2 years old and it's already changed the way people deal with money. It's incredibly reliable and astonishingly fast.
Who would say that the biggest Brazilian fintech revolution of all times would come from the Central Bank?
[1]: https://en.wikipedia.org/wiki/Pix_(electronic_payment_system...
Instant transactions, zero fees, available to anyone with a bank account.
Seems like some of the Pix => cash features are different.
Sounds similar to what I used in New Zealand quite a few years ago. However their system wasn't confined to mobile apps. Is pix mobile only?
Also in that system payments reconciled that evening, it wasn't instantaneous. To be honest I preferred that. I was able to fix a payment with a mistake before it went out.
I used to feel that way. Then I discovered how much time/effort is spent counting cash. Credit cards are all electronic and so the money goes directly into your bank accounts without needing to count and recount all that cash. Companies are money ahead vs cash once they account for all the labor cash costs. That is before we get into all the costs of theft that you avoid.
The above assumes a straight 2% transaction fee. If you are paying $.25/transaction then for small purchases the transaction costs are way too high.
I think the people aren't saying you're wrong because it's untrue, but that's like a niche of a niche.
Check out stacker.news as a live example.
Spotify and paying cloud companies are micropayments in the same way metered water and electricity are. In that they aren't. Both bill/pay out on either a regular schedule or a threshold.
Edit: my mistake was assuming any taxes are simple, see here for example: https://news.ycombinator.com/item?id=31389210
https://en.wikisource.org/wiki/The_Machine_Stops
I've been waiting (both with anticipation and dread) for a day when micro-payments are ubiquitous, as I think it's the only way to compete with the subscription and ad business models. I just don't know if people will really accept it given previous examples of it have played out. But I still hope someone gets it right, and I can pay 5c to read a newspaper article without needing to subscribe, and things like that.
This is an open system so lots of competition and zero lock-in is possible. It's su successful that world's 40% digital transaction are hapenning in India.
[1] https://en.m.wikipedia.org/wiki/Unified_Payments_Interface
In order to make micropayments work, you need to
1. have some sort of personal money-holding folder (it's so hard to avoid the word 'wallet' here), that you can easily fill (e.g. via your credit card). This moneyholder needs to be sufficiently safe for use, and should be password-protected. Ideally, it's a function or tightly integrated with your browser.
2. The website itself needs a moneyholder, though this could conceivably be implemented with limited functionality.
3. An easy to implement way to have small transactions from your moneyholder to the moneyholder of a website. To avoid creating issues around money laundering, the maximal height or a daily limit for transaction to one destination could be considered.
4. A legal way to retrieve funds from that moneyholder.
That's the easy part. Now consider the following:
1. The system should be useable regardless of the participants countries of origin, even in countries that have a cash culture and little implementation of credit cards...
2. ... but must adhere to international and national legislation (looking at you, Russian/Iranian/Cuban sanction regime)
3. A certain level of anonymity is advisable. The newsman at the newsstand has no idea what my name or address is when I buy a newspaper, and neither should the micropayment merchant.
4. Transactions need to be failsafe and fast. I don't want to first click through four layers of login screens and read an email before I get to reading the article I was about to buy.
5. Somehow, whoever manages this system needs to make enough money to support infrastructure and engineers, plus a profit. (The most easy way would be to demand a 5% processing fee for sending money into the moneyholder... I can see that approach failing for psychological reasons, though ("I put 20 dollars in there, why is my balance only 19$?")
All current net payment schemes fail in one of these fields. Paypal is comparatively fast and ubiquitous, but fleeces people with their fees. Cryptocurrencies are often slow - and processing fees are no joke, not to being with the volatility issues. I think GNU Taler [1] is interesting, but that by necessity would become a regional solution linked to local banks at best.
[1] https://taler.net/en/index.html
I don't think you need to make the same tradeoffs because less decentralization is needed for micropayments. I think GNU Taler is a good solution but it is competing against a monopolistic surveilance network that the banks profit from.
I don't think the world/country can depend on a single company like PayPal. Once you get vendor lock-in, they will hike up the fees. GNU Taler would implement the same interface/API across different banks/vendors/wallets, so the individual switching cost (of going to another bank, for example) is low.
The only way consumers can use micropayments is to set up a subscription via a credit card (or similar). Either with the vendor themselves or via a thirdparty. They don't want to do that.
Publishers don't want to deal with the hassle of accepting 100000 small transactions, some percentage requires them to add sales tax, some percentage will fail, and a non-zero percentage will be reversed. In the face of these problems, getting a check every month for advertises seems like a much better bet.
The closest thing to micropayments for content makers is something like Patreon, where consumers pay a small amount up-front for access to future content. But even here, a small amount is usually a few dollars rather than cents.
That would only work if the ad networks made more money off you with micropayments than they do with ads. If that was the case then you could be sure the ads would become so irritating to virtually force you to pay.
Plus, instead of just having your browsing habits, the ad networks would now have your credit card details. That sounds less than optimal to me.
I think back to when people used to tip dogecoins via a bot on Reddit. It was low-friction, low-cost, and seemed to seek no rent. A similar model may work well for micropayments (instead of microtipping), but I don't have enough faith in the current web3 sphere to develop this without it turning into a scam.
You seed your account with $20. Sometimes charging is free, sometimes it's $1-2 dollars, and sometimes it's more. When your account gets low, they bill your credit card.
EZ Pass: The electronic toll network on the East Coast of the US. Works similar to Chargepoint. If you only drive a few miles on a toll road they deduct a few cents from your account.