Hacker News new | comments | show | ask | jobs | submit login
We’re sorry, and we’re not rolling out the fees change (patreon.com)
1195 points by JoshTriplett 11 months ago | hide | past | web | favorite | 539 comments

This is a remarkably direct and appropriate apology, along with an appropriate response, a rare thing indeed.

If not for their initial response to the first round of feedback, namely to put a "we're still doing it" at the bottom of their original blog post (with completely unreasonable and unjustified explanations for why), I'd give them full credit for responding appropriately. As it is, though, at least they fixed this. They're still going to need to deal with the giant pile of patrons they drove off; many people lost a significant fraction of their support because of this change, and that won't necessarily come back.

Between 1 and 2 years ago, I talked to some folks at Patreon about an unrelated topic (namely, trying to use Patreon for registered charities). They mentioned at that time that they had a change in the works to shift processing fees from creators to patrons, which they said at the time would be "better for creators" for various reasons, but that didn't want to talk about the details yet. (Since the issue is now public, I don't see any issue with saying that now.) That discussion naturally didn't include any information about fee structure or un-aggregation of charges with associated additional fees, e.g. the "$1 costs $1.37" problem, since that wasn't the topic; the impression they gave at the time was that the only change was to pass on the (presumably still aggregated) fees to the patrons, which wouldn't have been as serious a problem, and wouldn't have killed Patreon's suitability as a wildly successful microtransaction platform the way this change did. So, they've been thinking about this for a long time, and yet they sprung it on people without warning and without running the full details by anyone.

Update: there's now a survey from Patreon asking for feedback: https://twitter.com/Patreon/status/941010263385374720

Twitter was filled with screenshots of people removing their $1 pledges. It wouldn’t surprise me if the rollback of the new model was motivated more by economics than PR, where the forecasted revenue under the new model (accounting for loss rate of subscribers) was lower than the forecasted revenue under the old model.

Well, those things are inseperable in this day and age, especially for a company that depends on internet communities.

It's astonishing how a company with one product could misunderstand its own market so badly, but given the troubles of Twitter and to a lesser extent Etsy perhaps it's less surprising.

Especially Patreon. Patreon more or less takes people's money in exchange for warm fuzzy feelings. If they piss off their users, the users get less warm fuzzy feelings, and they'll stop buying.

in exchange for warm fuzzy feelings? wtf are you talking about?

"I'm voluntarily contributing a small piece to a body of work that I love and want to see more of."

I suppose that can be construed as a warm fuzzy feeling.

It's more than that though, you're paying for that artist to be able to create. In most cases, if people stopped paying, the artist would have to stop creating, or at least reduce the time they spend creating, so they could find alternative income sources.

Just because there's no defined benefit or exclusivity doesn't mean all you're paying for is "fuzzy feelings".

that's a weird way to summarize the purchase of art and entertainment, but i guess it's technically accurate.

You're not purchasing anything, Patreon is about patronage, as in donating/supporting with no expectation. Whether you get something is completely up to the creator to do.

Not necessarily. Patreon has two payment models - either pay per month, in which case you hope the creator sticks to what they say they'll be doing (and you withdraw your support if they don't and you're not happy about that); or the other option is where you pay when they publish a new piece of content (with a monthly cap). I support a couple on that model. So in that case it is more directly a purchase of some content, even if nothing stops creators from posting it outside of Patreon for anybody to enjoy.

There's also a fair bit of patron-only content from many people - bonus comics, behind the scenes stuff, exclusive videos, requested artwork, whatever.

So yes, you are purchasing arts and entertainment. Just in an unusual model.

If there is no obligation to deliver then there is no purchase. It's patronage, an already well defined concept and the basis for their name. Why must it be made more complicated as "purchasing...in an unusual model"?

You're most definitely purchasing something. It's just not a tangible good or service.

That’s the warm and fuzzy part ?

I read you want to put weight on how the very fact of giving or retracting donations messages an expectation to the creator and creates a kind of defacto commercial relationship.

But honestly I see it the other way round, where creators on patreons have already put something out in the world, and donations come after the fact.

There is an expectation that they go on creating, but if they said “fuck it I can’t do it anymore” I won’t be running after them to ask for my money back; I already got my warm and fuzzy feelings, the deal is closed on my side.

> That’s the warm and fuzzy part?

Yes. In the same way that you purchase something from a charity when you make a donation.

If you think you are purchasing something you are badly mistaken. The creator is not obligated to do anything. And your only recourse is to stop donating to him. There are no refunds.

Warm fuzzy feeling, one might contend :)

What would that be? Either way, purchase implies a guarantee to receive something in return but the creator has no obligation to do this, so it's a donation.

It isn't a purchase of art and entertainment. At least all the Patreons I supported were free content.

There's many which aren't free content or which have content locked off for higher tier patrons. There's many which are "exclusive for patrons for [X time]" too.

“Misunderstand”? Yeah, right. They knew most donations are $1, and just wanted a bigger part of the pie.

If they were out to solve a problem, they’d just do this for transactions say $10 and above.

Nay, Patreon knew very well what they’re trying to do, and seeing it fail, decided to use the “play dumb” card.

They certainly can't say they didn't have time to think this over.. so yeah..

On the other hand, it's possible that most of thevø revenue does come from 1$ donations, and they didn't understand how much creators desired having many pateons vs. few high paying patreons.

Not so astonishing, when you understand the business of venture capital. When you raise too much venture capital and try to grow faster than the market dictates, this sort poor decision making tends to happen.

This is what I told them in the survey. I'm a newbie to the platform, and I understand the value proposition better than the owner?! That's really screwed up, and it will take quite a while before I trust them. I still kind of think creators should take their accounts elsewhere. How can a group of people just not understand their own product so thoroughly? Baffling. Is there any precedent for people so badly confused able to learn which end is up? It really makes me suspect the company and its users just have irreconcilably different ideas about what is going on.

Is there an elsewhere for creators to take their accounts? I give via patreon because it is the only platform I'm aware of that exists that anyone really uses in this way. I give to some creators via their own systems - but as far as a generally accepted mass market solution - I'm interested in hearing about others but don't know them yet.

There is LiberaPay. Non-profit org, based in France. https://en.liberapay.com/

There was Subbable, but then it merged with Patreon.

Well, those things are inseperable in this day and age, especially for a company that depends on internet communities.

"Voting with your wallet." Or to paraphrase the way Polonius put it in Hamlet, one of the key differences between sane and insane is what people do with their money.

Vine actually died because of this.

Interesting, how is that?

I wonder how many creators they lost, too. At least one guy I follow, who makes rpg scenarios, said “screw it” - all future scenarios would go to drivethrurpg, and patreon backers would get discount codes appropriate to their backing level. He kept patreon as a notification platform, and drove —all- his revenue to an alternate channel.

Same with a webcomic creator I know. He basically stated "I do this as a side-job and have a steady income, please move your pledges for me to the creators you support that depend on Patreon for their income."

Wonder if he's still leaving.

Absolutely. I told anyone that would listen to cancel everything because I feared that would be the only effective method.

Nothing like voting with your wallet.

Like the new Coke recipes? Then the introduction to the original recipes.

I understand why they do it: they trying to bring down transaction costs with payment processors. These fees on small amounts is death by thousand cuts.

But, I believe, there is a way: they can introduce wallets for $1-2 patrons: put $10 in a single transaction, then charge the wallet once it is a time to pay to creator. This should aggregate small transactions into larger ones and significantly cut the costs.

You do not understand what happened.

They did aggregate up until this point, at least for most patrons and creators. If you did a regular pledge with a monthly charge, you only made one transaction a month. Similarly, each creator got paid once a month with one transaction. The exception was per-post patronages, which do indeed have more fees and are less suited to the micro-payment model. That is NOT what this was about, though.

The real motivation was to support gated content.

They want to bring creators into their subscription system, where you pay a fee to access content. They had a problem, though, in that people could pledge, access content, and then simply cancel their pledge before getting charged. To avoid this, they needed to do Charge Up Front (CUF) to ensure that any access to gated content was paid for. But that creates some confusion about when people get charged (do you pro-rate the first charge, with or without the next month, etc.) because people would pledge a certain amount and see a different amount charged. The alternative is to have people on their own billing cycles, which is much less confusing for customers, but eliminates the possibility of aggregation.

Basically, they want to abandon the goodwill/patronage model and become a subscription service, likely because the latter model is far more lucrative.

You're ignoring the fact that increasing the number of distinct fees charged to Patrons translates directly to more money for Patreon (because Patreon was charging a much higher fee to Patrons than they themselves are actually paying on the transaction). Their public explanation was that a traditional subscription model is simpler, but it's obviously worse for users and the only reason to do that instead of the alternatives (such as the wallet approach) is so Patreon can get their cut of those extra fees.

As Wikipedia would say, "[citation needed]". The transaction fees to patrons that Patreon proposed of 35¢ plus 2.9% are real familiar to anyone who's worked on a payment processor. Patreon was essentially passing the transaction fees they would have paid onto patrons, and they would not have been making a material amount of extra money for themselves if they'd stuck with the new system.

That’s a rate that a brand new company just starting out might get. For that to be the actual rate Patreon is paying they’d have to be extremely incompetent.

For comparison, both Square (which is not known to provide the best rates) and Stripe advertise better than 2.9% + 35¢, and that’s the rate you’d get without any negotiation at all.

35¢ plus 2.9% is for average e-commerce and small business clients. A company like Patreon that does large volume of transactions will almost always be able to negotiate a much lower transaction fees with payment processors.

I don't see how a 'wallet' doesn't handle the CUF... Patreon should already have the money in what is basically a pre-paid account. You just have a minimum amount that can be paid into the wallet. There should also be pay in/out schedules so things are aggregated/automated on both sides. They could even make interest off of the floating money and work a bit like a bank. This should be their bread and butter, optimizing transaction cost. Plenty of other sites do this sort of thing for real businesses, I don't see why it is so complex...

It really does seem incredibly straightforward. A patron adds $20 to a wallet and pays a low transaction fee. Can even configure their account to automatically add $20 whenever funds are low. Money stays in Patreon's bank account. Patron can't get their money back once it's in the wallet. Every month Patreon charges a percentage to disburse all money to the creator's bank account.

What am I missing?

It's simple to explain in the context of this discussion. But probably doesn't seem as simple to the average user, who wants to donate $1 to a creator, and will feel like some kind of scam is going on if they are asked to add $10 to their Patreon balance first.

Thinking about it some more, Patreon may end up eating the additional fees for the first payment (for CUF posts), increasing their cut a bit on other posts to make up for that, then aggregating the recurring charges.

I am guessing they can make more off of an increased number of credit card transactions with some sort of deal with Stripe or whomever. It also allows them to obscure when and how much they are getting paid when it really doesn't need to be that complicated.

If I want to donate $1 per month to somebody, cashing out $20 in advance is quite a bit -- it'd probably discourage a lot of people, since the initial investment is too much.

It was just an example. It could be whatever the user wants to pay; since they're paying the fees for each transaction, it makes sense to keep some money in a wallet.

> A patron adds $20 to a wallet and pays a low transaction fee.

I wonder if they have problems with anti money laundering regulations? Financial transactions are supposed to clearly indicate the beneficiary. If you're paying into a wallet that gets divvied up later, that's not possible so showing the paper trail of who benefits gets difficult.

How do those look now? I imagine the beneficiary of the transaction is clearly Patreon. The fact that they happen to have another transaction that pays money to someone else is fine. I mean, we don't expect to have to write several checks to various patients and families when we give money to St Jude. Or, more interestingly, look at how escrow works in a real estate deal, where the buyer and seller transactions go to the escrow company, which is then responsible for distributing it to the various parties.

Now, since it's basically acting as an escrow payment, Patreon might have some laws restricting what it can do with that money while it's holding it. For instance, they likely cannot legally earn interest on that money. Similar to how the security deposit held for a rental cannot yield interest unless that interest is paid to the renter.

Automated road tolling systems have the exact same economic problems as Patreon. And the pre-charged wallet with $xx minimum fillups is exactly the model they chose. Judging by the longevity of their systems, it seems to work fine.

as was mentioned below, you don't really have a choice with road tolling systems. i don't like having to pre-load $xx, but it's better then the only alternative. people can stop interacting with Patreon, but you can't really avoid toll roads very easily.

At least in the Illinois Tollway you do: pay more by stopping at each toll.

That's sort of the same thing here, right? Prefill a wallet and you don't have to pay extra for processing fees (equating to convenience going back to tollways).

you have no choice with road toll. unless you plan on teleporting or walking.

You could just take a different route

> To avoid this, they needed to do Charge Up Front (CUF) to ensure that any access to gated content was paid for. But that creates some confusion

But if you actually ask people whether they want the first couple charges to have weird timing, or whether they want to pay significantly more in credit card fees that don't go to the creator...

> They did aggregate up until this point

Still an issue if most people support 1-2 creators with $1-2 pledges.

I really don't get the motivation of the change, aside from basic greed.

I mean, it seems like a dream business model, you basically channel donations to content creators taking a cut, collecting goodwill along the way.

Their service cannot be more than a (refined and well designed as you want) CRUD app with payment processing, it's hard to imagine having such high costs that is not profitable.

> I understand why they do it: they trying to bring down transaction costs with payment processors. These fees on small amounts is death by thousand cuts.

Then they should have never strayed from the original model, where they charge you once per month for all your pledges, causing only one fee paid to the payment processor.

(If they needed to accommodate people pledging small amounts to very few creators, they could have introduced some sort of quarterly, biannual, or even annual payment option instead of just monthly. I'm curious if they evaluated this option and what they concluded.)

My coffee shop has a sign: $0.50 fee for card transactions under $5. I’ve seen that sign over the years at a dozen places. The idea isn’t entirely novel.

And you’re right, in that perhaps they should send $3 every three months or. $6 every six if you’re only giving to one person.

The reason you don’t see that everywhere is it’s specifically against the merchant agreement to charge cc customers extra. The only places I’ve ever seen getting away with it are very small shops and government agencies (like the auto tag office) that legally cannot pay those costs out of ordinary revenue.

Not between 2013-2016; there was a federal settlement that included the provision that merchants can add a surcharge for credit payments, overriding any merchant agreement.

Which was recently thrown out on appeal, so that’s fun for any place that was relying on it.

> so that’s fun for any place that was relying on it.

Places can simply work out a monthly total the cost of CC fees and increase their prices overall (or on specific products) to recover that cost.

Gas stations seem to still regularly have a credit card price that is different than their cash price.


Q. Why are there different prices for paying with cash vs. with my Visa card?

A. A merchant is permitted to offer discounts for paying in cash, however, the discount must be given as a reduction from the standard price.

It's semantics: I believe the wording or the typical agreement states that you may not charge a customer more for using a card. It does not, however, prohibit providing a 'cash discount'.

Which is why this whole debate seems a little silly from a practical perspective. It doesn't matter whether a fee is paid by the seller or the buyer or if a difference in price is a discount or a surcharge, it all washes out in the end. How Patreon or gas stations frame the difference is simply marketing. Patreon's mistake was just poor marketing.

OT, but why don't programmers say or write "It's syntax". Different semantics, in the programming language sense, isn't what that phrase usually means.

I'd say they were quite the opposite!

"Semantics" is about the meaning you're trying to express.

"Syntax" is the way you've chosen to express that meaning in this particular case.


    def blah

    function blah() {
      return 'a';
Semantically identical, but unrelated syntax.

That's exactly what this whole discussion is about though.

The difference between "card surcharge" and "cash discount" is treated as 'semantic' in the merchant contract context, but is it a distinction without a difference from the customer's perspective?

Google gives two definitions for "semantics". The first one, "the branch of linguistics and logic concerned with meaning", clearly matches the common usage of "It's semantics": it's primarily about words and definitions, and how they relate to the concepts they refer to. The second is "the meaning of a word, phrase, sentence, or text". I guess that makes sense as a microcosm of the whole field of study, in the same sense that people talk about "the physics of a situation" (say, involving two hard balls bouncing off each other) as how the general study of physics applies to that situation.

It happens that, with programming, explaining precisely how each part of a program corresponds to "meaning" (and there are different levels and dimensions of precision, e.g. "xs.append(x) means it'll add object x onto the list xs" vs "xs.append(x) means it'll look up the value of the symbol 'xs' in the lexical environment and then the global if necessary, and do the same for x, and then look for xs's attribute hash table for a key named 'append'..." vs "xs.append(x) means it'll check if the array backing the list has room for one more element, and, if not, look for a free block of the right size in its freelist, and, if not, use 'mmap' to grab more memory from the OS, and barf if that fails, and then copy the array data into a new array of twice the size, and free the old array...") also gives a precise description of the behavior of the program (assuming hardware perfectly obeys the abstractions it's meant to obey), which is all the programmer is interested in.

This is a field where machines reliably turn symbols into actions; hence, the meanings of symbols are the central object of study. Semantics are extremely important and are almost all we need to think about. This is not the case in most fields. (I guess one other profession that comes to mind where semantics is important is law.)

To bring it back to your original question... yes, semantics is important to us. If we want to call something unimportant in that way... "It's a matter of definition" or "We're arguing over definitions" strikes me as the best fit.

(One other distinction: Programming language terms have semantics for machines. They also have semantics for humans. Whether the computer recognizes something as a closure, or complains about a syntax error or executes something completely different, is very important. Whether human programmers call it a "closure", a "function", a "procedure", or some other term, is generally unimportant. Thus, the maximally correct dismissive comment might be something like, "It's human semantics".)

I think that's not corporate policy and is something individual franchisees do, hoping no credit card company compliance officer will refuel at their store and notice (or possibly not even knowing it's prohibited).

That's illegal on my country, they cannot change the price for different payment methods.

Which country it is?

Payment cards are not a legal tender, so you can do whatever you want, provided you didn't sign a contract not to do.

In France for instance if you accept credit cards, it is illegal to add a fee for accepting it.

You can impose a minimum purchase amount for accepting it though.

Certain professions (i.e., taxis) are even forced to accept credit cards.

> In France for instance if you accept credit cards, it is illegal to add a fee for accepting it.

Are you sure, that it is by law, though? In most countries, it is a contractual obligation: when you make a contract with your bank or payment processor, one of those terms & conditions is not to discriminate against the card payers. The purpose is to force the processing fees on the merchant, otherwise the cards would be at a disadvantage and would not get popular. For a case study, see Southeast Asia (1. cash rules there; 2. you can pay with a card, but you will have to cover the fees too; 3. as a consequence, everyone uses cash, see also point 1).

In the case of taxis, I would understand if they voluntarily had terminals for their own protection (so they would not get mugged for few banknotes), but forcing them... that sounds wrong.

>> In France for instance if you accept credit cards, it is illegal to add a fee for accepting it.

> Are you sure, that it is by law?

It is: Article L112-12 from the Code monétaire et financier. "Le bénéficiaire ne peut appliquer de frais pour l'utilisation d'un instrument de paiement donné." (The seller shouldn't charge any fee for the use of a given payment instrument).


More generally, the law requires that the total to pay is to be presented to the client before he choses how to pay.

Interesting, thanks.

Spain, and its the same thing as the user from france. They cannot charge you any extra fee for using cards or any payment method.

Could be. There are a lot of places with minimum orders for credit card too. That probably works easier for the business anyway.

Maybe it's just me, but, I think it's so fucking stupid that it costs money (which goes to private businesses) to spend money in the modern world.

(and no, bitcoin doesn't fix that at all)

I think if we were in an alternate universe where the government managed a free utility for digital payments no one would find it strange for a second.

It has always costed money to to spend money and there are pros and cons to each method. All that stuff isn't free even if it isn't directly measurable most of the time.

Cash costs money too--you have to count it, handle it, deposit it, protect it from theft, verify it's authenticity, make sure you have enough change, spend time counting change, etc. As a consumer, if a merchant fucks you over you have little recourse besides suing. Cash takes no time to clear so you can spend it right after getting it, it is anonymous, it is hard to trace so you can skip on taxes, etc.

Checks can bounce, you have to deposit them, they take time to clear, they take forever to write, they can be fake, etc. However, it is hard for your cashiers to skim off the top, it uses exact amounts so no change to keep, it is a single slip of paper to carry around instead of a pocket of paper currency, as a consumer you can stop payment on a check if the merchant fucks you over, etc...

Credit cards are super quick to use at the register. They don't require any change (unless handing cash back). As a consumer, if a merchant fucks you over you can issue a chargeback. It is easy to track your spending as a consumer because all transactions are recorded electronically. As a merchant you don't have to handle change or cash, your cashiers can't easily skim off the top, etc...

It's all trade offs and I'll bet if you did an NPV on all the different methods taking into consideration all their pros and cons, they'd all wind up "costing" similar amounts.

And in the meantime, banks are making money on investing, lending out funds deposited. And not doing too badly: http://money.cnn.com/2017/03/03/investing/bank-profits-recor...

Major US banks made $171B _profit_ last year. Lets not pretend that they're already bending over backwards to lower costs on transaction processing.

> Lets not pretend that they're already bending over backwards to lower costs on transaction processing.

It's a free market and you as a merchant are free to pick your vendor. In addition, you don't even have to accept credit cards if you want--you can just take cash or check, or even just cash. Of course people entering your store might not buy anything from you because who the hell carries around cash or checks anymore, but no jackbooted thugs will force you at gunpoint to accept credit cards.

The credit card industry is a huge, highly competitive industry with little barriers to entry. Massive market forces are at work squeezing the margins for transaction processing fees to as little as possible.

My bet is that transaction fees are priced at what they are because:

- its cheaper than handling cash or checks

- it results in more revenue, even with the "cash back" discount given to CC users.

- it is more traceable and easier for your bookkeepers to manage

And to be honest, I believe many small business who take cash only do so because they aren't reporting all their income. When it is off the books cash-only, you can pocket the sales tax, underreport income tax, pay your vendors cash under the table, etc. I also believe that if their margins are so thin they can't afford a 2% transaction fee they probably have no business staying in business anyway. Whatever they are doing is more likely than not a value-destroying NPV-negative project to begin with. (see also: just about every business shown on those "save my bar / save my restaurant / save my hotel" TV shows)

>Credit cards are super quick to use at the register.

At least they used to be. Chips ruined that. Now I have to stand there staring at the screen for half a minute waiting to respond to prompts and to pull the card out at the end. For most transactions, cash is probably faster than using a chip.

This used to be the case for me until very recently. I've noticed that the two grocery stores I frequent most are suddenly significantly faster. Like almost instant (less than 2 secs). This is in Seattle. So, it'll probably be that way everywhere eventually.

In Sweden, in most stores, you insert the chip and enter the PIN just after the transaction starts, and then once everything is scanned you just confirm the total price (which is very quick).

In Denmark you just use NFC/Contactless. Hardly anyone ever types in PIN numbers and it is also very quick.

When I was visiting Sweden the terminals did not support NFC or I was just doing something wrong.

NFC support is getting more common in Sweden now. Retail stores has been slow with it until now, though.

This is why I’m a big fan of NFC payments when available. I find it to be pretty much instant, instead of having to wait what feels like 30 seconds for the chip transaction to process.

Though I’m in the US. I hear chip transactions are way faster in other countries where they’ve been using that system for a while.

It's even worse when your chip starts to fail, as one of my has started to do very recently. You have to put the card in three times(and wait for it to fail 3 times) before you're allowed to use the swipe function. Took longer than cash.

Recently I tried the chip and it failed, then I tried the swipe and it said to use the chip, then I used the chip and it said I had to use the swipe, and finally the swipe worked. First time I ever had a chip fail, and it's a pretty much brand new card, so it was most likely the machine's fault.

I don't see the problem. If waiting a little is such a big deal then you are in too much of a hurry already and/or too stressed out. Relax. Take it easy.

Where are you, out of curiousity? I find most of the chips are no slower than using a swipe card was before.

I lived in Spain for a year, and everyone had chips there. My US card always drew eyerolls as they had to find the /old/ machine and dust it off and find someone who remembered how to use it. I was excited that we were getting chips here in the US, but of course they fucked it up and for whatever reason it takes ages for your chip to be read, and there are frequent chip read errors.

In the US, it depends entirely on the point-of-sale system. Some are just as fast; some are a lot slower. My guess is that the slow ones are using the same old CPUs that the previous "stripe-only" ones did, which just aren't powerful enough to handle the added cryptography requirements in a reasonable amount of time.

Chip cards work super-slow in the US. I often pay in cash because it's so much faster. I take the same chip card to the Netherlands and it works in a few seconds.

I have the same experience in Orange County.

Everyone hates the chip.

The US

I have been used cards with chips for years and now contactless payments and they are a lot quicker than swipping the card. In less than 5 seconds is done

At least in Spain

I assume spain is chip and pin?

In the US it's chip and signature, and you can't sign until after the chip verification (either signing a paper receipt, or with a stylus on the terminal screen).

In addition chips are new and some terminals are painfully slow (30 seconds) to verify. I'm not sure I've ever seen one finish in under 5 seconds, though a few merchants have terminals that come close to that.

The sad fact is the terminals are painfully slow because US banks don't think US customers want or are capable of using PINs.

When you use a PIN you get a nice two factor signature from the card chip: it signs the current timestamp and the PIN you knew, and can do both as quickly as chip's processing capability and the bandwidth between the chip and terminal allows.

US banks came up with a dumb compromise just like most of their websites use Wish-It-Were-Two-Factor auth and secondary "Security Question" passwords, the chip cards in the US are doing their own Wish-It-Were-Two-Factor: sign a timestamp, wait some amount of wall clock time, sign a different timestamp.

Most of the wait in a chip purchase in the US is artificial just to make sure that two timestamps are "sufficiently" different. US banks should just give people PINs and stop this silliness.

> The sad fact is the terminals are painfully slow because US banks don't think US customers want or are capable of using PINs.

Is this conjecture, or do you have actual citations to back this up? Those same banks have been issuing debit cards with PINs for a couple decades.

It's a bit intentional hyperbole, but not by much.

Every chip card I've received to date from several different major US banks has included some variation of "Great news! You don't need to learn or use a PIN to use this card."

My personal reaction every time has been, "But what if I want to use a PIN?" and this far I've never seen a satisfactory answer in those same letters or on those banks' own websites.

Admittedly that is purely anecdotal, as far as citations go, but in my mind it seems pretty clear what these banks think about PINs for credit cards.

Why can't you though? I feel that is an implementation detail that is leaking out. The machine can do two things at once and only use the signature if the transaction is successful.

Yes, chip and pin, and for amounts smaller than 20 euro you don't even have to put the pin if you configure the card

Chip & signature sounds awful (and pointless!) Why would they not use PIN?

We now have contactless payments for under £30 in the UK, similar to Apple pay but you just place your card on the reader.

that's because of shitty software. chip should get faster: https://paymentweek.com/2017-5-2-wait-no-more-quick-chip-tec...

Yeah but a >1% fee is ridiculous. We all know visa, Mastercard, PayPal are drowning in cash.

I had high hopes for bitcoin like currency having a flat fee in cents and an optional percentage few for insurance if you do want a charge back and other nice credit card like features.

But bitcoin is riddled with its own charge mania nowadays.

> Cash costs money too--you have to count it, handle it, deposit it, protect it from theft, verify it's authenticity, make sure you have enough change, spend time counting change, etc.

Holy God. I might as well say it's all equally free, since when the heat death of the universe is done with, there will be no difference between something having existed or taken place or not.

With the new chip system I swapped to cash for many transactions because it's faster.

I use contactless chipped card everywhere and it literally takes less than a second for most payments.

I am not going to use such an insecure payment method.

What's insecure about it? Heard of a lot of cases where money were stolen from people banks accounts/cards using scamming, viruses etc (and witnessed one right when it was happening), never heard of problems because of card being contactless. It also lets you not to flash your pin in front of the whole store for many of the payments. I think on the convenience/security scale it's very high.

They limit it to low value transactions specifically because it's not as secure. Feel free to dig into it if you don't trust the industry's assessment.

They limit it to make it secure and make it pointless for the thieves to try exploit it, at the same time covering many quick payment needs - convinience/security balance I mentioned and you ignored. And feel free to think about industry's assessment - they make more and more cards with contactless payment capabilities and it's used very extensively in some countries of the world already.

to make it secure Security is not binary. It's trusted less than physical contact readers for large value transactions for very good reasons. They care less about low value transactions, but that's not a security question.

You may not care about security, but all I said was it's worse, and not good enough for me which you have not denied.

Apple Pay takes all of about two seconds. No way you are paying with cash faster.

Paying for lunch is pull out cash and go vs. handing card to waiter waiting on them then signing and go.

In that case cash can save ~5 minutes.

That was true before the introduction of chip cards, though. The waiter's having to carry the card to the staff area of the restaurant, complete the transaction (by swiping), then return the card to you, probably interweaving some of their other duties along the way, took up most of the time. Their having to wait an extra few seconds for the chip reader seems a drop in the bucket.

I noticed this process has gotten noticeably longer after chip cards where required.

Waiters used to take cars and swipe them ASAP presumably to get better tips by being prompt. Now, put this off until they finish a full round and/or put cards in and start doing something else because it's going to take a while.

Granted, this is far from a complete switch, but it was still noticeable.

You don't have to hand anything over with near field stuff. You put it near it, ding, done. Saving you 5 minutes + time it takes for you to exchange cash. If you're dealing with a waiter for lunch, you're already wasting a ton of time.

The trade off on what food I eat is not a waste. Wasting time on slow payment methods when better options exist is.

Near field is also limited by transaction size so it's far from sipe for anything.

I don't think it's stupid that it costs money, but I am surprised fees are still as high as they are. I suspect it's got something to do with how opaque the processing fee is for the customer.

If I understand it right, merchants often aren't allowed to pass on the fees or (much the same thing) give discounts if you chose a cheaper payment option. No wonder there's little competition on the fees, and Visa[1], Mastercard[2] and PayPal[3] are all having record profits.

[1] https://www.cnbc.com/2017/10/25/visa-quarterly-profit-rises-...

[2] https://www.reuters.com/article/us-mastercard-results/master...

[3] https://www.reuters.com/article/us-paypal-hldg-results/paypa...

Merchants absolutely can pass on the fees or give "discounts" for cash, at least in the USA. Gas stations are one of the major businesses that practice this.

Anecdotally, I worked at a company store years ago where we would charge customers paying with cards a 2.75% fee, which is what we were paying. And that store was operated by a Fortune 500 company.

Also, virtually every small private business I go to in Los Angeles charges some fee, or has stipulations, for credit card transactions.

This wasn't originally the case. Passing on the fee was for a long time against most merchant agreements. Which makes sense from the credit card company's point of view: they don't want shoppers to have any reason not to whip out the card.

The big change came in January 2013 after a big court judgment. Adding fees, though, is still illegal in some states: https://www.creditcards.com/credit-card-news/business-surcha...

Thanks for that link.

Made me wonder about the stores/restaurants I frequent here that do charge, apparently they're still ok to do so after 2015(link from the comments in your linked post): https://oag.ca.gov/consumers/general/credit-card-surcharges

Where? I haven't seen any in years, at least in the southeast.

There are more stores not accepting credit cards than charging extra.

Maybe I should've mentioned "Los Angeles" higher in my comment, but looking at the sibling comment to yours, apparently there are plenty other states where it is legal to pass on the fees as a surcharge as well. Obviously your situation may vary.

In London, there's quite a large number of merchants who have now stopped accepting cash - requiring cards/NFC only.

PayPal fees are high because 1) people pay it and 2) fraud.

Credit card "processing fees" however are high because of the stupid rewards programs that Americans are so addicted to because it makes them feel like they are "sticking it to the man" and getting money back, to the point of having dozens of plastic cards in their wallet. Of course with processing fees at >1% of gross value no cashback or airline miles program is ever going to make you come out ahead.

In the EU there are no silly rewards games and fixed processing fees of at most 0.3%.

In places like the EU and Australia, credit/debit card interchange fees are capped very low (like 0.3% low).

I would, I in no way want to give that kind of power to government.

I would 100% opposes to a government run digital payment system, the government is too closing integrated to todays electronic systems as it is, I in no way want them to own and directly control it

If you receive your paycheck electronically, the transfer likely occurs over the ACH system, the primary operator of which is the Federal Reserve. The Fed also runs FedWire. The Federal Reserve has actually been at the forefront of electronic payment and settlement systems since their advent.

Though, the Federal Reserve is quasi-public/quasi-private. But for any systemically important transactional system you can be sure that the government is plugged in one way or another. The Clearing House company is the main alternative to the Federal Reserve electronic systems for interbank transfers (they're the "private" ACH operator), and they're heavily regulated, with the Federal Reserve given substantial oversight authorities.

Ok, And.

Just because the current system sucks and needs to be replace, does not mean that replacement should be another Government Run System.

As I stated in my original comment the government is too closly integrated to today's electronic systems as it is

I want the government out, not more ingrained

I'm still optimistic that we'll have a digital currency solving this problem well in the future. Distributed, auditable, scalable, international transactions at ridiculously low fees.

It costs money because there are significant fixed and incremental costs to handle it: Fixed because of the labor and value of the technology, and incremental because of fraud. Much of the incremental cost is fraud insurance.

There are many ways banks could reduce the fraud, but US banks don't need to since nobody is protesting the fees. In Europe however, where interchange fees are regulated to low levels, they have been quick to adopt/mandate features like chip-and-PIN and 3DSecure that lower fraud.

And yet somehow US banks manage to struggle by not even making it to $200B/year profit, having to survive on $170B or so...

Can you name an American consumer financing company that even breaks $100B in revenue?

Some places have systems you'd like. For instance Denmark has the "dankort" debit payment network where small merchants only play a flat annual fee and no per-transaction costs whatsoever.

They couldn't even do it with physical mail. You're paying the government every time you send a letter for the cost of that postage stamp. And the postal service in the US is pretty much constantly under-funded.

What makes you think we could do it digitally?

Like Europe, right?

Payment costs in the Netherlands (ABNAMRO bank, english): https://www.abnamro.nl/nl/images/Content/022_Zakelijk_nieuwe...

Note that nearly all fees are flat fees.

You pay to have some handle the transaction details for you. It's a service. You are free to mail dollar bills, though you'll use a service you need to pay for there, too.

They do... It's called ACH.

The only reliable way to do large transactions internationally without percentage based fees are wire transfers.

Even they have a $15 fee and sometimes a $15 receiving fee as well.

Ideally you'd have zero fees for transactions under $5. $5-$X0000 would have a percentage based fee and thereafter a flat fee for large transactions.

USA wire transfers have so high fees because banks don't want to offer such payment methods, so they charge extreme markup (their cost is less than 1$, so 15$ is 1400% markup) on that.

In general, real time payments for large amounts can be made for <1$ (both fedwire system for its participants and various EUR systems do that), and consumer payments that can be (a) batched and sent in bulk and (b) get delivered e.g. at the end of day, not in real time, those have a cost of ~$0.01 (perhaps 0.02 in case of small volumes) to the bank.

So, in EU, everyone has access to something that's pretty much equivalent to USA "wire transfer", and it might cost e.g. some 40 cents or can be offered to consumers either "for free" as part of a common services bundle/maintenance fee.

EU blows my mind at so many levels. Many countries with different cultures and languages working together for common good of citizens.

I have really hopes for the Euro

Then people would game the system. If someone wanted to send $100, they'd send 20 payments of $4.99 and 1 payment of $0.20... at least I would :)

It seems very likely that the change was motivated by payment processor fees. Patreon uses Stripe to process payments, which charges a flat rate of 2.9% + $0.30 per successful charge. Compare that to the proposed fee patron fee structure of 2.9% + $0.35 — where presumably the extra $0.05 per transaction is the cut Patreon takes off the top for themselves.

> the proposed fee patron fee structure of 2.9% + $0.35

The problem was that was per pledge. So if, like me, you support a couple dozen creators at $1-$3 each, you would expect to pay 24 x $.35. But all those payments are charged in one transaction. So they are only paying $.30. It isn't 5c per transaction they were planning to skim, but $8.

Part of the change was no longer aggregating the charges but running them through individually. They wouldn't have been pocketing most of the fees.

Which honestly makes it even more baffling an idea.

I believe the idea was to give the money to creators immediately rather than end of month when they charged everyone.

...which is still baffling! The core value of Patreon is microtransaction aggregation. This isn't just "product management screwed up a feature", it's product management doesn't understand their own product at all. Crazy.

It's like designing a fancy new electric car but leaving off the wheels.

It's like designing an electric car, but requiring you to use a diesel generator to recharge it.

You're assuming that Patreon will issue all of those charges at once.

However, they talk about that problem in the article. What if a Patreon supporter pledges/makes a subscription on Day 1, then another on Day 2, and a third on Day 3, and so on?

In their original blog post [1], they spoke about how, ideally, they'd issue each of those charges to the supporter immediately; and then begin recurring billing 30 days later after that date. However, if those charges are made on different days, and the anniversaries occur on different days going forward, then they don't have the opportunity to condense them into a single transaction.

Patreon are looking for a solution where someone can create a subscription, be billed for it immediately (i.e. not wait for beginning of next month), and then continue from there with recurring monthly payments. With a naive system each subscription would have its own cadence, preventing transaction consolidation. They also talk about how, if they have a standard monthly billing period, then there are issues with waiting until the next period to make the first charge.

Perhaps they didn't explain this as well as they could have, but it made sense to me. It seems like a "Patreon Wallet" could indeed be a solution to a lot of these problems. Refill your wallet with a single large transaction, then draw funds from it when pledging to support creators.

[1] https://blog.patreon.com/updating-patreons-fee-structure/

If I back a new creator, it charges me immediately (I think, at least it seems that way from my billing history), but going forward I'm billed with the rest of my support in one transaction (definitely). I really don't see the problem.

Patreon's docs until recently noted that they get charged 1.9% by Stripe, so there's more than $0.05 being skimmed there.

Don't forget it's now 2.9% of $1.37 instead of $1.00. Reducing their piece by another 1 or 2 cents depending on how Stripe rounds

Patreon does enough volume to qualify for discounts - their Stripe rate is 1.9%, not 2.9%. (This used to be on https://patreon.zendesk.com/hc/en-us/articles/204606125-How-..., but it's gone now.)

Looks like they wanna be nice guys running a marketplace at 5% and claiming to be good guys.

And then VC dogs biting them to make fatter profits so they find ticket master like shoddy processing fee scams.

minus their likely bargain with Stripe. I doubt they pay as much as a small customer.

Weird that they’d even use stripe as they could likely skip the middleman and save more money. Stripe is great for small business. Patreon isn’t that.

Stripe isn't really a middleman. In fact, as they also handle the acquiring process with efficient high-volume rates, generally fewer parties are involved than normal.

Even for very large businesses, actually processing cards directly is pretty difficult and generally not worth doing. Instead you can just get a processor like Stripe or Adyen to give you good rates on an interchange-plus model.

Whereas Paypal will apparently do 5% and 0.05, which makes the dollar go to $1.10 instead of $1.38.

That's what they did before the announced fee changes. After the fee changes were supposed to go into effect they we're no longer aggregating pledges. Madness.

Would that put any extra requirements on them? They're not just a middleman in that situation, they actually hold the funds. That sounds close to the PayPal for a long time: we've got accounts and transactions and deal with real money but we really really are not a bank, promise, please don't regulate us as one!

Alternately they could have patrons pay in arrears: You give your donations and then you pay at the end of the month. One transaction on the card and then it gets divvied up.

That is how it currently works... I get one patreon transaction a month even though I have weekly pledges. That's part of what made this change so frustrating and greedy

Maybe they felt the cost of fighting fraud there would be too high? Scammer signs up as a patron to his own account with stolen credit card, gets paid, Patreon can't charge the (now-cancelled) card at the end of the month?

They could delay payments to creators by a month to combat this but then they're holding the funds for a month, which brings you back to the original problem of seeming like a bank.

This seems like something Stripe could address, maybe.

EDIT: Just saw the other reply that explained this is how it currently works. I'll leave my original comment up though

Isn't that how Apple App Store / Google Play end up working as well, though?

> put $10 in a single transaction, then charge the wallet once it is a time to pay to creator

They already group the transaction on the 1st of the month for me. I use PayPal, is it different for other payment types?

If you keep people's money and move it around for them, you start looking like a bank, and you get regulated like a bank. I think this is why online games let you buy tokens with real money, and the tokens sit in your account. (It's probably of critical importance that there exist no way for the tokens to be converted back into currency - which is why regulators were and are far more concerned about online gold farming than the game devs themselves. It's good to be the bank.)

Having worked many years ago at There, a now defunct-in-all-but-name Second Life competitor: you're absolutely right. We could let people buy "Therebucks," but it would have been illegal for us to let them convert Therebucks back to cash.

WRT Patreon, I'm 99% sure that this is the reason they didn't propose something like a "balance card" where patrons just give them $X and both pledges and fees get deducted from that -- since that money's being held and then paid to someone else, it could end up being too close to a "money transmission service." (I'd thought there was a non-zero chance that this was a motivating factor in this change to start with, actually; coming so soon after a huge investment round sure makes it seem like there was a condition to that investment that spurred this change. Christie Koehler wrote a good post about that on her blog, although I don't have the link handy.)

How do this work for services like Playstation or Steam, which have wallets like you're describing? Is it just the fact that you can't take money back out or what? (I have no idea about laws regarding banking/finance/etc)

I'm pretty sure it's the "can't take money back out" that's the deciding factor, yep. While I'm not positive, I think this is even true for (non-banking) government agencies -- I can load value on a transit card, but I can't take it back off as cash.

(Edited to add: of course, this means that Patreon could have done a balance card thing if they were willing to say "but you can't get your balance back if you cancel," but I doubt that would have won much more love than the per-transaction fee idea did...)

Or indeed Apple App Store, Google Play.

Those aren't the same, though. You can pay for a subscription with the App Store, and you can buy in-app "tokens" for games. But the subscriptions are being charged when they turn over each month (Apple will bundle charges together that occur on the same day, but that's it, AFAIK), and you can't exchange tokens back for cash. So Apple is never "holding" your money.

I was thinking more along the lines of buying iTunes cards to load money onto your "account", which you can then later use to purchase digital goods, at which point the funds are disbursed to the content creators, minus Apple's rake.

How is that different from putting money into a Patreon account, which can later be allocated to content creators, at which point the funds are disbursed, minus Patreon's rake?

Ah, I see. But I think that's still the same "one-way conversion" thing that avoids the issue, right? If you buy a $50 iTunes card, you can't spend $25 of it on iTunes and get the other $25 back as cash.

Patreon could certainly do that, but either they haven't considered it or--probably more likely--they don't like the idea. There could be practical reasons for not wanting to do that; a lot of people have talked about Patreon's problems with chargebacks, for instance, when people want to cancel pledges but they're bundled together with pledges they want to keep going all in one credit card transaction. In this balance card scenario, it'd be easy to have someone say, "Hey, I wanna cancel the monthly pledges I'm giving and get my balance back," and they could lose a lot of goodwill if they say, "Yeah, you can't do that." (While many of us would rather receive cash than gift cards, we generally understand that an iTunes gift card is something we can only use by spending at iTunes.)

You can go creative (though, I'm not lawyer, nor accountant): have a minimum payment limit: 5$, for instance; if pledge is $1, then ask to pre-pay 5 units in advance (and be transparent why this has to be this way).

The problem if I understood correctly their previous strife is when you finance patrons across taxation boundaries. If they have to collect vat, they need separate transactions, even if that drived up transaction costd

> If they have to collect vat, they need separate transactions

At least in Europe that's not true. If I buy milk (6% VAT) and beer (19% VAT) in the supermarket here (The Netherlands), I don't have to split that in two transactions.

sure it's true for goods and services if sold to a single state, but if you sell across state boundaries you're required to register and pay your vat to each state where the sale happened and track those transactions along with the source

Yeah, they could have avoided this altogether by checking with their user base but this is a good response, although radio silence for five days is a bit crazy.

Shocking you community's members with a sudden change - no matter the size - is almost always seen as hostile and drives members away.

I recommend the talk[1] Joe Peacock of fark.com gave about the time they deployed a redesign unannounced and made their infamous "You'll get over it." post in the thread full of confused and angry members.

[1] https://www.youtube.com/watch?v=YnVeysllPDI

>Shocking you community's members with a sudden change - no matter the size - is almost always seen as hostile and drives members away.

But the size of the change matters. No matter how much warning they gave there's no way I would tolerate a flat $0.35 fee plus a percentage on a dollar donation.

What VC board of directors is going to support that?

Are we supposed to believe they didn't game the numbers out on this? That attrition wouldn't hold sway? I don't believe it, and any attempt by them to "I'm just a simple caveman, unfamiliar with your society" an excuse lies in high tension with "$100MM in funding."

I think the simplest explanation is that moving to focus on bigger-ticket patrons destroyed their brand.

If you ask the user base if they are OK with a price increase you are going to get a big NO every time.

Their user base includes both patrons and creators. Obviously patrons would say they don't want a price increase, all other things being equal, but they might be okay with it knowing that increase goes to creators. And asking a handful of high earning creators could have given them good feedback on why it's a bad idea.

But the increase was going to the company Patreon not the creators. That was mostly the point - they need/want to make more money.

It didn’t really seem like Patreon was making significantly more money from the new model, fees were higher because they chose to no longer batch charges. The fact that this belief is so widespread is a major testament to Patreon’s failed messaging.

Not batching charges would be so needlessly wasteful that I refuse to believe that was really their intent.

Fair. It's true that they didn't announce the unbundling until after the initial wave of complaints.

"fees were higher because they chose to no longer batch charges. "

Why did they decide that?? (sorry did not read all blogpost in case it is there)

Scroll way down on this page https://blog.patreon.com/updating-patreons-fee-structure ... apparently people were getting confused by the batched charges

I feel like Patreon is trying to please everyone, which in the end pleases no one.

If Patreon came out and said hey, we've built a platform for years, this is incredibly expensive, we need to show a profit or shut this entire thing down. We're going to start charging a bit more, AND give you this thing you've been asking for in return.

Then everyone would be happy. They need to follow the shit sandwich strategy of good, shit news, some more good stuff for you guys.

Instead they dropped the shit on the community, arrogantly hoping they'll simply eat it.

The original post about this change pitched it entirely in terms of "more money going to creators".

I am a creator and my immediate response was to think about the math for a few seconds and go OH HELL NO.

Creators need to treat it like business income and realize that whatever fees patreon takes is just the cost of business.

No better business than forcing your supplier to reduce prices (or preventing them from increasing them).

That's not true. I'm completely fine with the idea of paying processing fees!

Once per month.

I give them leeway on the 5 days. I guess they could have said something general like "We understand the new payment policies are not what our user base has expected of us. We are formulating an in-depth response and will have it out in the next few days. We appreciate your response and patience."

It'd have been dangerous to say even that if they didn't know they could fix the problem in a satisfactory way, or they'd get blasted even more. But yes, they should have handled this far, far better.

I'm just sharing in here b/c I've always appreciated when others "peel back the curtain" to show the thinking behind decisions.

At the end of the day, it was obvious that we fucked up from the immediate feedback, but internally there was already a strong camp who felt this change was wrong, which greatly helped our ability to move quickly and plan out how we'd call this off (all the code was already in production behind feature flags).

On the data side, our churn ticked up, but was actually lower than we predicted and quickly returned to normal levels https://imgur.com/a/inFOE -- so the reversal was largely driven by the complaints of creators and patrons alike, and our own realization of how tone-deaf the decision was to begin with.

Speaking of which -- in my mind there are some obvious solutions to alleviating fees that didn't require moving the entire burden onto the goodwill of supporters. We're going to tackle a more comprehensive roadmap in the coming days and weeks (and talk to creators much more during this process), but if anyone is curious or wants to offer feedback, here are my thoughts:

* A big problem (that we arguably created ourselves) is a patron could pledge to a creator on the 25th of a month (granting a patron 5-6 days of access), get charged, and then get charged again on the 1st of the next month (granting the patron 30-31 days of patron-access). Pro-rating didn't seem like a right solution for the first charge since becoming a patron unlocked all the content immediately. So imo we should just (like a Netflix or any other subscription) have your first pledge grant you 30 days of access, whenever it was created -- and have this forever be your "anniversary date" of charge -- if you pledge Jan 25th, you'd be charged again Feb 25th.

* The question is around how we aggregate pledges if this patron pledges to another creator. If we say the patron pledged $5 to the 1st creator on the 25th of Jan, then pledged $5 to the second creator on the 10th of February, I'd want to charge the patron $10 on Feb 25, and in the receipt call out that you're paying for the 1st creator for the period of Feb 25-March 25, and for the 2nd creator from March 10-April 10 (because you'd have already paid for Feb 10-March 10 when you pledged to the 2nd creator on Feb 10). This way you're ALWAYS getting 30 days of access for every pledge you pay.

* Now that we can aggregate payments across multiple creators, I'd want to allow for the purchasing of multiple months/year at a time for a creator or creators, which further reduces fees -- additionally it allows creators to setup rewards (we're building this system out now) for when patrons have pledged a total of X amount or pledged for some time period.

* Finally, once we can pledge across multiple time periods, I'd love for cryptocurrency to be able to pay for patronage. Currently it's trivial for us to accept bitcoin via our Stripe integration, but we wouldn't be able to do recurring payments, and the ultimate thing that kills it for me is that creators wouldn't actually get bitcoin, but rather we'd have to convert it to fiat immediately. I'd rather patrons be able to purchase multiple months of patronage and creators have the option to convert to fiat immediately or hold actual bitcoin and participate in the wild ride that is cryptocurrency speculation :)

Ok that's all -- ideas are rough above, but that's what's on my mind lately -- happy holidays.

As you say, pro-rating first charge has the content unlock problem and "anniversary date" of charge defeats aggregation.

TLDR: Aggregate on a single day, but pro-rate the second month, not the first, so there isn't a "runt" payment up-front.

I start a monthly $20/mo pledge on Feb 15th. I'm billed $20 on the Feb 15th, same day. This covers one month, Feb 15th - March 15th. Let's say the first of the month is aggregation day. Come March 1st I'm already paid up through the 15th, so I get billed for the pro-rated $10 for the second half of March. Come April 1st and all aggregation days thereafter I can now be billed on aggregation day for the full monthly charge.

> TLDR: Aggregate on a single day, but pro-rate the second month, not the first, so there isn't a "runt" payment up-front.

That was my immediate thought reading this text. Of course it'll confuse some users, but since it's the fairest way, it should be the way to go. Also, just don't charge for the prorated period if the feed is going to be as big as the contribution (e.g. for a couple days for someone contributing $2 a month)

This is essentially the same feedback I left in the form: pro-rate the 2nd payment, and waive it entirely if it's too small to justify the fee. The payment processing confirmation email that Patreon sends out could put an asterisk next to a pro-rated pledge and explain what's going on down below for people who are paying close attention.

Pro-rating is fine, just make it a month+. So you support with one week left in the billing cycle, they get charged for 5 weeks. You could make it an option that creators set.

I'd really like to see a debit system: I could load up 6 months worth of pledges at once and incur only one fee (whether it's explicitly charged or done 'hidden', I don't care). Creators could choose how often to be paid, as well, so they control what kind of fees they incur as well.

Ah, if that wasn't clear from my post, I think debit/credit is absolutely interesting. Analogous to how Twitch effectively gives you more "bits" to send to streamers as you buy more in bulk.

Creators all run vastly different businesses so pro-rating is always a thought -- one thing it seems incompatible with is in the fulfillment of goods, where the creator takes on a burden of cost. It would be difficult to consider what the patron receives in return if they only pay out a pro-rated portion of a reward tier. Curious how that would play out in your scenario --

You could make your own cryptocurrency. Just a PatronCoin or something that people just buy a bunch of and plop in an account and they get sent to creators who can exchange them for USD if they so choose or sell them to someone for more just like any other coin. This could be an option that provides maybe a fee-less transaction or something. Since the patrons will be buying them from you you could mark them up slightly or something along those lines then you make your money up front instead of waiting for fees to roll in?

> A big problem (that we arguably created ourselves) is a patron could pledge to a creator on the 25th of a month (granting a patron 5-6 days of access), get charged, and then get charged again on the 1st of the next month (granting the patron 30-31 days of patron-access). Pro-rating didn't seem like a right solution for the first charge since becoming a patron unlocked all the content immediately. So imo we should just (like a Netflix or any other subscription) have your first pledge grant you 30 days of access, whenever it was created -- and have this forever be your "anniversary date" of charge -- if you pledge Jan 25th, you'd be charged again Feb 25th.

You could keep your charge date on the 1st of the month, and just not charge for that first month if they are charged up front. Let creators choose the length of their grace period: 3 days, 10 days, two weeks. Creators get paid, don't feel pressured to provide refunds, and patrons get a little sign-on discount.

Creators get paid, don't feel pressured to provide refunds, and patrons get a little sign-on discount.

This seems like the key. Instead of making any decisions about payment scheduling options, design your system to allow for all of them within reason, and give the creators a menu.

If you don't offer your users enough options, you will eventually be out-competed by a patronage service that does. Hardwiring your system to a model that happens to work for you simply ensures that this will happen sooner rather than later.

This is an interesting thought -- one of the other things we see creators do is physical rewards, which means there is a cost to them in fulfilling their pledges. Although I could see compatibility to your suggestion still b/c you just wouldn't get the physical good twice, one good would cover the period of the next month + the original grace period.

I think one essential consideration that's been lost in the discussion is that Patreon can be used in many different ways. For some creators, it may be a periodic direct exchange of money for goods and services. Each month a patron pays $X and receives $X worth of either physical goods or some kind of service. For other creators, Patreon is a tip jar, where the primary product is released elsewhere for free, and fans can use Patreon to give recurring tips/donations, and through Patreon they at most receive some bonus rewards that are generally not themselves worth the $X they paid, since they are mainly paying in order to donate to a project that is external to Patreon. This latter mode of use is the one I, as a patron, am most involved with, since I donate to multiple web comics. I'm sure there are other modes of use as well that I haven't thought of or seen.

The reason this is important is because a one-size-fits-all solution may not be possible. For the case of physical rewards in return for pledges, charge-up-front is completely unnecessary. The creator can just say that the rewards will go out after they receive payment on the first of the month, so someone subscribing in the middle of the month receives nothing until the month ticks over. For the donation case, charge-up-front is useful, but not for the reason you might think. The main point is to prevent freeloaders who pledge, view the exclusive content, and then cancel without paying a cent. In this case, creators would probably be fine with pro-rating the first monthly payment after charge-up-front, or even just skipping it entirely. It's all donations anyway, so the point is not to extract maximum value from each patron, it's to make sure each patron is paying an amount that they are comfortable with, so that they will continue to pledge long-term.

My overall point is that the platform that is Patreon is abstract and flexible enough to support a number of different business models, which may have different needs with regard to charge-up-front and other features. And to get an idea of the full breadth of those possibilities, I think you just need to interview a lot of creators, and I think you'll find that the usage patterns are more varied than you expect. At the same time, I'm sure there's value for patrons in having things like charge-up-front work the same way across different creators' Patreon accounts, so that they know what to expect when pledging.

I think there is a real benefit for creators of physical rewards when they aren't obligated to fulfill or refund pledges that come in at the very end of the month. They can begin and end shipping earlier, and apply late pledges to next month's rewards.

And as a patron, I appreciate receiving my December rewards near the end of December rather than mid-January.

This doesn't seem like a great idea. What if I just joined for the first month, grabbed all of the content, then dropped out? I mean, I love this idea for the Creators that offer premium content that I want but don't want to pay for but it seems like they'd get ripped off a lot.

The problem that charging up-front was created to address was people pledging money, reaping rewards, then dropping out before paying any money at all. With this system, if someone signs up for one month, then drops, the creator still gets paid for that month.

Worst case: a patron could game the system and only pay half as much as other patrons for the same content. That would involve pledging and canceling repeatedly every other month. That is not a likely scenario, and still infinitely better than the previous worst case: patrons paying $0 for that premium content.

One solution I’ve seen suggested: Pro-rate the SECOND transaction. The one on the first of the next month. Full charge for the first one.

You have ONE solo transaction, the rest are able to be bundled as normal, and the $1 pledges that make up half of my Patreon income (and, I’m extrapolating, a similar fraction of everyone's Patreon income) aren’t eaten up by that 35¢ fee.

> A big problem (that we arguably created ourselves) is a patron could pledge to a creator on the 25th of a month (granting a patron 5-6 days of access), get charged, and then get charged again on the 1st of the next month (granting the patron 30-31 days of patron-access).

A simple solution to the "patron pledges near the end of the month" problem would be to have two batching dates, one at the end of the month and the other at the middle, and chose the one farthest from the current date, so the initial charge of a pledge would always be for at least a couple of weeks. Though personally, I would find that solution a bit annoying; having everything aggregate into a single charge per month is very convenient for patrons.

> This is a remarkably direct and appropriate apology [...] many people lost a significant fraction of their support because of this change, and that won't necessarily come back

They could have restored all cancelled pledges and covered their costs for a few months. That would give creators money to pay the rent, and probably go a long way towards restoring relationship with patrons.

Clearly, they're not that sorry.

I don't understand why this .37 applies to every transaction. Wouldn't it make sense to charge the overhead once a month, with the actual dollar(s) being distributed as the user wishes? They should use the power of their platform (obviously nearing or having passed critical mass) to make it easier for prolific supporters while also maximising their paetron's income.

I think they had good intentions, but got greedy and lost sight of their overall goal

Successful/appropriate company apologies seem so rare these days!

Anyone have examples of ones done right?

Off topic, but is there a good solution like Patreon for registered charities?

>Many of you lost patrons, and you lost income.

I think Jack at Patreon fundamentally misunderstood the nature of the relationship between patron and creator. A lot of patrons feel like their payment is a donation instead of paying a subscription or water bill. Yes, some patrons+creators have a quid-pro-quo arrangement where payments unlocks content or extra services. However, many payments are just "appreciation" type of money. That is a fragile relationship and it was wrong to tamper with it by charging extra fees to the patron.

Jack was/is a creator himself so he should have known this dynamic and therefore predicted the bad outcome ahead of time.

Ideally, when you start a multi-sided platform, you want to get the economics correct from the very beginning so you can leave it unchanged. (E.g. Apple iTunes charges 30% since 2008 and it's stayed that way.) However, if you have to readjust the percentages, it's preferable to take it out on the sellers' side and not the buyers' side. When ebay started, the fees for sellers were ~3.25%. Over the last 10 years, it has crept up in increments to ~10%. All of those price increases were absorbed by the sellers.

If Patreon needs more than 5% to make the numbers work for a sustainable business, they need to take it from the creators and not the patrons. Patreon was riding on the "good will" of patrons making voluntary payments. It was a terrible miscalculation to destroy that good will and nickel & dime patrons with extra fees.

> Jack was/is a creator himself so he should have known this dynamic and therefore predicted the bad outcome ahead of time.

I thank that might exactly be the reason why it happened. From what I recall as a fan of their artistic work, he (& his wife) never asked for straight up donations, but either offered something in direct exchange (via Bandcamp) or the money went towards a specific project (Kickstarter).

> I think Jack at Patreon fundamentally misunderstood the nature of the relationship between patron and creator. A lot of patrons feel like their payment is a donation instead of a subscription or water bill. Yes, some patrons+creators have a quid-pro-quo arrangement where payments unlocks content or extra services. However, many payments are just "appreciation" type of money.

And at least for serials the "quid-pro-quo" arrangement is usually earlier access to content (e.g. bonus/patron chapters), there's no exclusive/sale type arrangement.

>And at least for serials the "quid-pro-quo" arrangement is usually earlier access to content

Yes, when I typed out "quid-pro-quo", I was thinking back to the recent HN article about webcam women making money on Patreon. Patrons would pay $10 a month and for that, the women would do private chats, role play, etc. For that particular type of patron who needs to see a particular woman take her clothes off, the 2.9% fee plus 35 cents would be more of a "nuisance" fee and they'd just go ahead and pay it. Let's call this group, the SUBSCRIBERS.

However, a lot of Patreon relationships are not quid-pro-quo and it's a totally different psychology when patrons pay a $1 donation to "support an artist" with no direct benefit other than some good feelings or minor things like "early access to content" like you pointed out. In this scenario, Patreon is the online version of throwing money into a hat to show appreciation for the musician at a coffee shop. Inflicting extra fees on patrons damages this type of relationship. (E.g. "Twitter was filled with screenshots of people removing their $1 pledges.") Let's call this tip jar group, the DONATORS.

If SUBSCRIBERS are the majority and dominated the platform over the DONATORS, Patreon could have stayed the course to charge extra fees to patrons on December 18.

However, it's telling that Patreon aborted that plan which signals to me that the DONATORS are a very large and influential part of the Patreon membership. Jack said they did a bunch of spreadsheet models and surveys before announcing the fee changes but obviously, their assumptions about their members were way off somewhere.

(I'm oversimplifying with categories of SUBSCRIBERS vs DONATORS to illustrate a point. It's certainly possible for some patrons to be a blend of both psychologies based on how their pledges are split to various creators.)

This episode is a good demonstration of how the structure of production works: prices begin with what the consumer is willing to pay and work backward to higher-order goods or factors of production.

It makes intuitive sense that if there are costs generated by the arrangement, you go to the ones benefiting from the arrangement (creators) before you go to the ones already providing the funding (patrons). Maybe Jack's being a creator himself is what blinded him to this.

Exactly - I'm thrilled to donate $30-50/mo to the YouTuber FlightChops because he does some amazing, high production value work and I want him to be able to keep doing it. Even still, I only watch 1/3-1/2 of his videos simply because it's rare that I have 25-30 minutes to sit down and watch one. I'm happy to donate the money to him because he's doing good work and I spend a lot more money on much more pointless things. I don't view myself as buying anything from him. I wouldn't be upset if he went 2 months without putting anything out just like I wouldn't be upset if he put out 10 videos in a month (I have a cap on my donation to him as I think most Patrons do).

> ebay started, the fees for sellers were ~3.25%. Over the last 10 years, it has crept up in increments to ~10%. All of those price increases were absorbed by the sellers.

Isn't this a false dichotomy? It's just a matter of how you phrase it, no?

If the increase in seller fees is reflected in real pricepoints on eBay's market, it could work out as it really being the buyers paying the difference.

>If the increase in seller fees is reflected in real pricepoints on eBay's market, it could work out as it really being the buyers paying the difference.

That would be true but that didn't happen. Final ebay auction values for stable items (e.g. books) did not rise by +7%. The sellers really did end up paying most of the fee increases.

Yes, some sellers tried to claw back the extra fee commission by adding extra "handling charges" to the total. That had limited effect because it made prices higher than (1) other ebay sellers that didn't add handling fees and (2) competitive Amazon Marketplace prices.

Also, Ebay lets bidders sort search results by "total price including shipping & handling" so adding handling fees just pushes your items down to the bottom of the list.

I don't get this. Eventually, it is just taking a cut from a transaction. What does even mean the seller is paying for it or the buyer?

Let's say the final price is 100$ and cut is 10%, so seller gets 90$ and buyer pays 100$. You could say the buyer is paying for it since the seller was willing to sell for 90$ and the buyer had to add 10$ to get the item. Or you could say the seller pays for it, since the buyer was willing to buy for 100$ and the seller has to give up on 10$ out of those 100$.

I really don't see the difference. It seems just marketing in terms of how you phrase it. I even saw companies saying things like: seller is paying 5% and buyer 5%, which seems just a way to make it look lower for both sides.

The difference is when the cut increases, but the overall price does not. Buyer still pays $100, but seller only get $80

What the gp is saying is that before, the buyer would pay $100 and the seller would get $97. So in effect, the seller has borne the extra fees

The distinct difference is that if the increase is absorbed by the sellers, they either leave their prices the same or raise them, but the buyer is always paying the amount listed. If the increase is absorbed by the buyer, they can see the cost and buy it, and then be changed the extra amount at the end.

While you are correct that one way or another the buyer will be paying more it is how you present it (buyer raises price to counteract the increase vs seller being charged more directly by the service).

Outside of Buy It Now items, sellers don't directly set pricepoints on eBay.

Yes, though buy it now accounts for something like 85% of transactions

It's hard to believe that Patreon didn't ask for feedback on this model initially. I'm sure they did, but my guess is that they didn't expect the blowback from people that depended on $1-2 patrons to have such a large impact on their image.

IMO, before Twitter/Facebook, this apology would've never happened. They would've continued to press on with their fee structure and it would've been yesterday's news.

The real question is - is this still a good move for them? Or should they have dealt with the blowback either way?

> It's hard to believe that Patreon didn't ask for feedback on this model initially.

From their initial blog post announcing this change:

> In preparation for this change, we ran experiments and months and months of research to understand patrons’ potential reactions and we found that many patrons were happy knowing that this change will send more money to creators.

Looks like a somewhat interesting user-testing glitch to me. Perhaps their sample of patrons wasn't representive, or when you sit people down and discuss a new feature 1-1 you can be more convincing than when broadcasting it to the whole world in a blog post.

I could also imagine that they didn't actually show the fee details, and just pitched it at a high-level.

As a patron, I'd be fine if Patreon just added transparency on fees, showing exactly how much reached the creator and how much went to fees, which is equivalent to "passing on" those fees. (They're effectively paid by the patrons either way, the only question is whether it's made clear what fraction of a pledge goes to the creator, what fraction to Patreon, and what fraction to processing fees.)

In my opinion, they made two critical mistakes here, and transparency on fees wasn't one of them. First, they un-aggregated charges, so that a $1 pledge costs $1.37, and ten $1 pledges costs $13.70. Second, they effectively increased all the pledges by individual patrons, rather than just being transparent about fees on the existing pledges. Effectively, they said "we're going to increase everyone's pledges by the amount of the fees (and hey, would you look at that, we get more money now)", rather than "we're going to show you exactly how much of your existing pledge goes where, without changing your existing pledge". Of course that led to a revolt.

I like this idea of showing transparency. They screwed up by not letting the decision remain in the hands of the funder.

> First, they un-aggregated charges, so that a $1 pledge costs $1.37, and ten $1 pledges costs $13.70

According to what I’ve heard, this was requested by credit card processors, because people would revoke the patreon charge, not understand why they’d be charged such a random amount every month.

Instead, they were supposed to charge every patron for every creator separately, and make it clear on the bill for which creator that is.

I fully expect that they will have to keep this unbundling.

> This was requested by credit card processors...

Patreon, to my knowledge, has never said this. The move would be understandable if it were true, but they've instead offered different rationales (like "creators want to be paid out faster on new pledges"). If this were the reason for the new model, I can't think of a reason Patreon wouldn't publicly say so - it would've won a lot more sympathy.

Do you have any evidence this is the case?

> "creators want to be paid out faster on new pledges"

This is exactly the reason. The important follow-up question is: which creators?

Patreon supports two fundamentally different groups. One uses the platform for pure donations, where there are no exclusive perks to becoming a patron. This group doesn't care about faster payout because there's nothing riding on it.

The other group uses Patreon as a paywall platform. They care a lot about payout speed, to prevent people from joining quickly to gain access to the paywalled content, then canceling their payment before start of the next month.

The paywall group is where the money is, because Patreon's model doesn't provide the mutual assurance necessary to convince people to fund public (unrestricted, infinitely-reproducible) goods. And so, inevitably, the platform will cater to the paywall group.

Patreon should use a prepaid wallet for this group. It should continue aggregating for the first.

I'm happy to foot a 50c or so service fee per month. I'm not up for paying 35% service fees! That's outrageous and I'm confident someone else will start the right service and creators will simply move there.

I doubt most of the paywall group cares about payout speed.

They care about payout guarantees. That's not the same thing.

I've seen no evidence or mentions of that particular rationale. They claimed they un-aggregated to support their "pay up front" model and start charging pledges on the anniversary of the pledge date.

They mention it below the original announcement, though they don't give any numbers.

> Not only that, but if you’re a creator on the monthly plan without the benefit of charge up front, you are constantly worried that your patrons are going to delete their pledges before they’ve paid.

That doesn't say it was requested by credit card processors. As I understand it, the concern there is that they don't want people to pledge, get access to patron-only content, and then revoke their pledge.

> Looks like a somewhat interesting user-testing glitch to me. Perhaps their sample of patrons wasn't representive, or when you sit people down and discuss a new feature 1-1 you can be more convincing than when broadcasting it to the whole world in a blog post.

Or perhaps we're all just a raving mob waiting for the right inputs, and early sentiment by influential individuals actually shapes our thoughts and actions far more than we like to admit.

It's also possible that we only make up 5% of patreons?

But that we are very noisy, because tech people on HN/twitter/etc. really want a micro payment solution for the web, to avoid ads.

The top Patreon patron by far, Chapo Trap House, was adamantly against the change. I do question whether they really did adequate research.

Chapo Trap House has exactly one pledge level at $5+, which means this change would have probably benefited them financially. They'd be unlikely to lose many patrons upset about having to pay $5.50 a month instead of $5, and they'd be getting 3-5% more from each pledge due to no longer having to pay transaction fees on their side.

You'd be surprised how many people would be upset at $5.50. Especially since that $0.50 is going to be multiplied by the number of creators a single patron supports.

For me it was enough to put "just pay an annual lump sum to my favorite creators and drop the rest" on my TODO list; I'm glad Patreon came to their senses.

Do you expect Chapo Trap House to advocate policy positions because it benefits only them? That's, uh, not really their thing, is my point.

The point isn't whether they'd benefit, it's whether they're in favor of it when (possibly) polled.

> patrons were happy knowing that this change will send more money to creators.

That could depend on how they asked patrons.

If I was asked if I were in favor of more money going to creators, I would say yes.

If I was asked if I were in favor of paying more to make this happen, I would say no. If I want to send more money to creators, I would opt for a higher patron tier.

Well, something that's gotten elided in a lot of the outrage is that moving the transaction fees from the creators to the patrons would have given creators 3-5% more money on the same amount of pledges. And while it seems "no duh" to point out that a $5 pledge is literally worth five times what a $1 pledge is, in terms of revenue, that means that many creators could have lost a surprising number of $1 pledges and still come out even (or even slightly ahead). I ran the numbers for several Patreons, big and small, based on what I could scour from Graphtreon, and the amount of money most creators lost over the last few days would have been largely offset by the money they'd get back.

I'm not saying that this makes the change a good idea, or that it was communicated well, or that there aren't some really interesting business dynamics that are likely to come up one way or the other that will piss off creators. (There are only six creators on Patreon who have more than 10,000 patrons as of now; to earn that $450M valuation, Patreon may need creators who have hundreds of thousands of patrons.) I'm just not convinced this would have been the Patreoncalypse.

If patrons on average adjusted their pledges under the new fee structure such that they were paying about as much money as they were before, creators would end up with less money on average, because the new de-aggregated fee structure results in higher fees as a percentage of the total amount paid in almost all cases.

> something that's gotten elided in a lot of the outrage is that moving the transaction fees from the creators to the patrons would have given creators 3-5% more money on the same amount of pledges

Patreon really tried to make this "point". It is not actually valid; "amount of pledges" before the change is not comparable to "amount of pledges" after the change. Measured as a percentage of "money paid by patrons", revenues to creators fell -- they rose when measured as a percentage of "pledges" because the new fees weren't part of "pledges" while the old fees were.

We also need to consider that content creators might crave an audience just as much or more than the donations.

It sounds like they explained the asynchronous collection to the patrons without also explaining the implications for fees?

Yea, could have been a botched user study. I've also seen situations where user studies were carefully crafted (and the questions worded) with a desired conclusion in mind, often in order to justify an executive's already-made decision. Not at all saying this was the case here, but that happens too.

In addition to what others have said, this is almost surely a messaging issue. How you frame the survey and outcome is critical to getting honest feedback. I bet the surveys didn't detail exactly how things would be implemented and how it would be presented to the patrons. Instead, if you frame it as potential solutions that get your favorite creators more money, the response would certainly be positive.

The email announcement also didn't make the fee increase for 1$ pledges obvious.

On my first read I thought the fee was per transaction and that they would keep aggreting pledges...

It would have take the degenerate example to show that... And only a super honest business would show the edge cases.

They ran a poor test and / or the internal politics were always going to prove out a certain result. There is value to qualitative research and I doubt the answer was to do a better job of A/B testing.

It was the disggregation that seemed to piss people off the worst - which I doubt they pitched to people in their test groups.

before twitter/fb patreon would've never happened


Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact