Why Payments Are Hard, Even For Apple And Google | Hacker News Reader

Why Payments Are Hard, Even For Apple And Google

techcrunch.com

52 pointssolipsist41 comments

>In addition, you only get the bank’s confirmation of the payment 3-5 days (in the US) to 3 weeks (in some EU countries) after the payment attempt.

I find it surprising that it's been 15 years since the commercialization of the Internet and (near) instant global data communication and it still takes banks 3-5 days to confirm payment (e.g., the time it takes to send a physical letter, coast-to-coast, via the USPS). This is the timeframe of a standard Electronics Fund Transfer (ETF) for bill pay, bank-to-bank transfers, etc. The only alternative, as far as I know, is a wire transfer, which is a tedious and expensive option.

Does anyone know what's preventing the U.S. banking system from improving its electronic transfer capabilities? Given the importance of the free flow of money and goods (and liquidity) to our economy, it seems like the quick movement of funds would be a significant priority for our overall banking system.

(Note that if banks could confirm electronic transfers within a few seconds (and if we added an extra layer of security beyond simple passwords, such as a challenge/response system), we could significantly reduce the occurrence of fraud. Near instant confirmation would also alleviate, or eliminate, the issue of short-term credit mentioned in this article.)

That's what I like about Canadian banks. They have a system in place they promote as sending money via email. You don't really send it through email. It's simple a system all the major banks in Canada have adopted. I can transfer money from one account to another, quickly and easily, with minimal fuss. It takes 30 minutes, and I believe this is more for security reasons then purely technical reasons. It makes sending money in Canada so incredibly easy.

Yes, I lived in Canada for three years, and had the impression that the electronic transfer capabilities of their banks surpassed what's available in the U.S. Perhaps this is due (in part) to the fragmentation of the banking system in the U.S.? In Canada there are essentially five banks that dominate the banking industry - see http://en.wikipedia.org/wiki/Big_Five_(banks). There is also more regulatory control of banks in Canada, which likely makes it easier to require standard operating procedures, consistent infrastructure, etc.

I remember payments in NZ being <1 min -- possibly seconds -- 10-15 years ago, as there was a single clearing-house for all EFT, nationally.

Same thing in The Netherlands, I can transfer money to my dad and it shows up almost instantly in his account.

Note: Mix of facts and speculation, take with grain of salt.

Banks (and credit card companies) make a lot of money on the float, so it's against their natural incentives unless they can win more business as a result. I think they probably could, but it would take a coordinated roll-out of new products or acquisition/integration with an external company... which is difficult to do because of regulation. So you're in a position where you'll either need more or less regulation to do it - either the government mandates it, or the government deregulates and a competitor starts up who currently doesn't have market share, so has nothing to lose by introducing the new features.

I was a NACHA member for the better part of the past year so I've talked to many of the bankers, consultants and government-types involved with running the ACH network about this.

It's not about the float. It's about the up-front investment in the standard, which millions of bank customers have already implemented. It's also about the risk of allowing instantaneous funds transfers; the current delay allows for at least some mitigation. It's also about the fact that there is no new standard yet. (Based on what I know about NACHA, there won't be for a while.)

What's cute is that much of the time when I pay for something it's instantly charged to my card - my account balance reflects the deducted money, whether it's been actually moved out to the other account or not. But if a merchant gives me a credit/refund, it can take several days for that to show up - some places will say "two billing cycles" (although they have no idea how long my billing cycle is - I presume they mean their own internal cycles).

Money already seems to be moved somewhat instantly, but only ever out of my accounts.

Bingo. This is why it smells like something fishy. Like the banks are trying to earn carry/float interest. And possibly the government is in bed with them, by adding some regulation to impose an artificial delay, to supposedly add safety/caution, but in reality to give the banks cover to earn more profit on this carry/float interest. This is my best thumbnail theory, anyway.

That's because the plastic card interchange works in real time and ACH does not. Even your charge cards have a delay, though. There's a difference between authorizing a purchase and capturing the funds. Once funds are captured there's still a batch settlement process that happens each day. So it's not really that big of a conspiracy; the players aren't even the same.

Agreed on auth/capture/settle - it's possible it could be different though, we just have an entrenched group of players that prefer the status quo.

It definitely could be different. That's why I'm working on FaceCash.

The banks earn by investing the money while the money is "in transit", so they want to keep it there as long as possible. Not all banks do this. I can transfer money from my account to friends' accounts and they get the money in seconds.

Agreed. Any software engineer knows it shouldn't take much more than 1 second, tops for a number to be incremented in one computer and decremented on another, separated across a network. Roughly speaking. Now add in fudge factor to account for a long and possibly unreliable network path (the Internet cloud, or similar private banking inter-network) and for the need to perform an atomic and 100% bulletproof and auditable transaction, and we can envision a scenario where the total latency between when the request is made and when the deal is done and put to bed as say 1 day, tops. And that's being very generous on implementation complexity assumptions. So when I hear things like 3-5 days, a week, 2 weeks, that's getting into insanity land. I smell either really old skool legacy systems in terms of process and hardware capabilities, or, something fishy like a sneaky way to allow the middle-men to earn "carry interest", or whatever, during that bloated time gap. Possibly also some government regulation which adds an enforced, artificial delay. But in terms of what are the actual minimum technical requirements to carry out that task, a few seconds would be generous, an hour would be padded and a day's wait would be bloated to hell. I could even say a day if each bank wanted to run all transaction requests as a single batch performed nightly, with human auditors being able to look at reports both the day before and after to "sanity check" results. But again, that should cause a 1 day delay at worst.

I am not a banking programmer, so take with grain of salt. Perhaps there's some evil super-complex-and-unavoidable issue at play, but I doubt it.

if banks could confirm electronic transfers within a few seconds

Banks do this millions of times each day on the global ATM network. I can swipe my bank card at most any ATM in the world and within seconds the funds are removed from my account and dispensed into my hand.

In that case, the card is indirectly linked to my bank account. It's only when you know the exact bank account and routing number from which you want to remove the funds that you're forced to use the archaic ACH system.

Clearly the technology for instantaneous confirmation of funds transfers already exists, but it has yet to be expanded beyond its original purpose.

I have been a banking programmer for some large European banks, so whilst I can't speak for the US system I can describe the way ATM transactions work in the UK, Germany & Switzerland.

Making a withdrawal from an ATM/ePOS does not debit your bank account immediately, but it does reduce the limit (i.e. available funds) on your card immediately. Checking your "balance" at an ATM shows the card's total available funds, not your bank account balance.

ATM operators are often not the same as your bank, even though they may show your bank's logo at the top. There are fees at every stage of the transaction so ATM operators and ePOS transaction acquirers can decide whether to ask your bank/VISA etc whether you have the funds. For small, low risk transactions they will often wait, preferring to batch a bunch of debits to a single institution at a quiet (cheaper) time of day.

Your card is reconciled with your bank account daily (usually daily). A payment to a third party takes longer - the payee's bank needs to complete reconciliation too and, as other commenters have pointed out, the network breaks sometimes.

Banks can improve the time it takes to transfer funds to third parties and have done so in many countries but it has taken government intervention. The float does play a part, but getting competing banks to agree to a standard is almost impossible.

Side note: Your card has many, many limits most of which you will never encounter. Examples: total cash per ATM per day; number of cardholder-not-present transactions per day; total spent in gambling establishments per month.

The author suggests that Amazon hasn't dealt with third parties at major scale. That seems counter to my experience, where almost 50% of my purchases on Amazon are fulfilled by a third party. For companies at the scale of Google, Amazon, and Apple - I'd suggest that hiring the talent and developing the risk management processes aren't as large a hurdle as he would suggest.

The major challenge is convincing people (vendors and sellers) as to why they should use your system.

Ohad's background is in risk so he's inclined to view the hurdles as being more related to what he's familiar with. I agree with you though, there are much more important factors at play.

>The major challenge is convincing people (vendors and sellers) as to why they should use your system.

See, that is the /first/ challenge; but as you start overcoming that barrier, you start hitting the problem the author starts talking about. E-gold, for instance, did pretty well solving the "convince people to use us" problem, but choosing to ignore the fraud issue killed them (and almost landed them in jail)

I'm going to (respectfully) take issue with your claim here, though I would happy to be proven wrong. I am clearly an early adopter (as pretty much 95% of people on HN likely are) when it comes to technology, particularly online technology. I routinely use paypal (+bump), buy almost everything online, haven't touched a paper book or newspaper in over two years. I even have a SquareUp dongle and use it from time to time. I have over 200 transactions in Paypal and over 500 purchases from Amazon.com. I won't even try to calculate the number of iTunes purchases (content+Apps) I've made.

I've heard of E-Gold, and may have created an account- but I've never used them. I just IM'd three of my friends - one had heard of it, two hadn't. None had ever used it. They all frequently use both Paypal and Amazon (one is a prime member).

I realize that anecdote is not the singular of data, but, a quick glance at Wikipedia, shows that at their peak, E-gold only had 5 million accounts - 2.5% of what Apple has today. (And, in Apple's case, these accounts are connected to an actual Credit Card customer)

So - two things to be aware of:

o E-Gold was relatively small. When you are small, you can't afford to hire the high-level risk management talent. These people are rare, and very expensive. Therefore, one of the critical tasks of _small_ payment processors (not the large ones, Apple/Google/Amazon can afford their salaries and departmental budgets) is to survive long enough and not get shut down by making these mistakes. I'm certain Paypal, when it was smaller, also came close to the abyss several times before they got to scale.

o Of course, the way you get to scale is you make yourself attractive, _or_ you subsidized your payment processing with the rest of your business. (Think about Microsoft buying their way into Console Entertainment)

I'll agree that fraud and risk management are important, but I do not believe that they are barriers to success for the companies the author had identified, they are more important to the smaller players (E-Gold).

For the larger players the question they have to answer for sellers is "Why should we use you? Which consumers are using your system that would make it worthwhile to going to the hassle to adding you as a payment option?" For the consumers it is "What do you do better than my existing Credit Card/Paypal Account which has served me well as a consumer?"

PayPal spent $300M learning how to deal with fraud[0]. This money is in terms of investment in development and outright monetary losses. The problem is you can't get a large number of users while you are also being taken apart by fraud. When dealing with payments, finding users is not your hardest problem.

[0] Source: One of the investors in my company was involved fairly early in PayPal. We've had many conversations about this matter, especially regarding another company I was working with at the time that was building an "anonymous" wallet.

And all of this because the credit card model is fundamentally flawed. When buying something it's like sending the seller a trivially copyable key to your vault and letting him get the money he wants.

While I agree that credit cards could be a whole lot more secure (using the same token for identification and authentication is just plain stupid) fixing that still leaves you a long ways away from solving the fraud problem.

E-gold used username/password pairs and seemed to be making pretty good progress towards solving the 'using someone else's account' kind of fraud.

The kind of fraud that killed e-gold (and I think, the harder kind of fraud to solve) is "he sold me a defective whatsit" or "after I paid him, he never sent my thingamajig"

If you don't solve that harder problem, you quickly become known as the payment method of choice for criminals.

That isn't a harder problem. It's a much much easier problem. YOU can decide where to buy from. It's a much saner system from both the customer and the seller's perspective.

And it works right now. In the Netherlands there is a system called iDEAL that works like this:

1. You go to a seller's site e.g. bol.com (which is like amazon). 2. You fill your cart and click checkout. 3. You get sent to your bank's site with a page that displays the amount and a button "pay". 4. The bank then transfers the money to bol.com.

That's almost exactly what e-gold did.

They stated as a design goal that you shouldn't be able to back out of a transaction once it's gone through. The problem was that the system became popular with sellers of fraudulent goods, and this does, eventually, start to cause problems.

I suspect that iDEAL has some way of letting you report fraud and get a refund from your bank if you get ripped off. they probably have some form of chargbacks, etc.

It works like wire transferring money to the seller, you get the same rights and protections. In fact it is wire transferring money to the seller.

Charging back is not easy, but it is not a problem in the least. It's only a problem when you buy from people you don't trust (but then I wouldn't want to give them my credit card number at all!). This does happen on sites like ebay. The risk is moved from the seller to the buyer. For sites like ebay you need a middle man to be safe, only now you need it as a buyer instead of as a seller.

I wonder if Google's smartphone Google Authenticator app and three factor auth will prove to be a competitive advantage for them in the payments space?

Given the choice between a Paypal and a Google Checkout option, I think I'd be significantly more likely to choose to use a 3 factor authorised Google service over an emailaddress/password authorised Paypal one (especially if I was using an untrusted network)

I'm not sure Apple would be able to ramp up a cross platform 3 factor auth system particularly quickly... (seeing iTunes on Windows makes me suspect market takeup of "iThirdPartyAuth on Android app(tm)" isn't going to be world-changingly rapid...)

I wonder if technically a bank wouldn't be a in a better position than a tech company to disrupt the online payment scene.

Sure, banks are not universally known for their innovative powers, but they usually have the infrastructure and the customer support necessary for such a system (so they are more where the author sees Apple)

Ohad knows his stuff, but he didn't cover everything here. There are other reasons payments are hard for everyone. You can't really have this discussion without talking about point of sale systems. The existing retail point of sale infrastructure in the United States is unbelievably fragmented and old. Apple and Google are in a good position to make devices that replace registers, but adoption is still just as much an issue for them as for everyone else.

This is good for payment startups (such as mine), but you have to have a lot of tenacity to untangle such a giant knot.

What do you think of Square?

I think it's the most gorgeous eight-track player ever made. They have enough money to shift, so for their sake, I hope they shift soon. For my sake I hope they don't.

I'll bite ... what is it about their reader / product line that makes it equivalent to a "gorgeous eight-track"?

I've thought the reader looked a bit fragile (and wouldn't hold up to heavy use), but that doesn't make it obsolete right out of the gate.

The next generation of payment technology will not be plastic card-based or involve a magnetic strip. They've created a really nice magnetic strip reader for plastic cards. It's kind of making the most amazing tape deck ever right before the mass distribution of Audio CD. Certainly it's an impressive feat of engineering and design, but it's not what I would pour my time into as an entrepreneur or my money into as an investor.

Personally I think my company's technology does make Square's obsolete right out of the gate. We just face a much different adoption hurdle because we replace the entire system from end to end.

And that's probably why they'll win. This generation of payment technology isn't going away in the next 12-18 months. I bet we won't see the next generation surpass this generation (in usage) for at least five years. During that time, someone like Square can build tremendous momentum and relationships that will allow it to much more easily transition to the next generation, because it has a strong brand, strong relationships with the marketplace, and corporate competency in the space.

For a fantastic example of what I'm talking about in another space, see Netflix.

Maybe. But relationships with whom? Why would a large retailer want to use Square? Why would a competing point of sale vendor with an enormous installed base want to work with Square?

Netflix isn't a very good analogy. The retail payment space involves a three-sided market: consumers, merchants and POS vendors. Netflix didn't even have to struggle with the complexity of a two-sided market. Though it's done well for itself, it also hasn't completed rendered the previous standard useless. Plenty of people use DVD and Blu-Ray today.

I think your view is possible but I'm not sure I understand the details. In contrast, I can answer both of those questions given the model for my company's technology.

I agree, Square has the best possible strategy -- they catch the laggard businesses who still don't have merchant accounts, individual people who have never had merchant accounts (i.e. everyone) and let them get away from PayPal, AND get a platform deployed which can handle NFC or some new payment/loyalty credential for early adopters on the merchant side.

On top of that, Keith Rabois is one of the top operating executives anywhere, and has extensive paypal experience.

NFC has been "the next thing" for years. I worked for an NFC reader company in ... 2003, which had been around for years. (ViVoTech). We don't even use chip and pin cards in the USA -- the magstripe is going to remain a major part of the market for the rest of the decade.

Time for Bitcoin to take over the world.

It's worth pointing articles like this whenever the Apple haters berate Apple's "greedy" 30% cut. Particularly when you only spend, say, $1-3 at a time, the fees take up most of that.

The only way you could have a more efficient system without cutting the credit card companies entirely (which is a huge proposition with enormous barrier to entry) is to buy prepaid credit in larger blocks (say $20 minimum) rather than charging each transaction (or day's transactions) in one small lump.

Apple partially has this with it's retail cards but they probably lose 20% of that to the retailer.

Payments seems to me to be one of those areas like music (and really all digital content) where there are players that have cooperate who basically have no interest in cooperating. In music and digital content it's the labels, publishers and studios, all of whom are investing heavily into turning the clock back to 1997 (thank you, 30 Rock). In payments, it's banks and other financial institutions (including credit card companies).

One player that has huge potential in this space in coming years are the mobile telecommunication providers. They have the network for POS system, mobile payments is an area growing in leaps and bounds and they have a payments infrastructure already.

Carriers seem intent on fighting progress too (as really does any large incumbent). Apple totally changed this industry with one product release [1], something the Apple haters seem to conveniently forget. If it wasn't for the iPhone, none of what we currently take for granted (even on other platforms) would be possible.

Carriers (and other distributors like cable companies) are terrified of becoming dumb data pipes for which the only differentiators are price and service area. But that is (IMHO) their inescapable fate. What carriers could be however is the infrastructure for mobile payments, which has enormous potential.

But currently they're ceding that market to Apple and Google.

EDIT: let me be clear, I'm not advocating Apple's position on mandatory use of their system. Far from it. I think the move is arrogant and short-sighted beyond belief. But criticisms of them being "greedy" predate that move by years. Personally I think Apple's payment infrastructure makes sense if you're small and not if you're large (much like hosting actually). As such it should be voluntary. Let it stand on it's own merits. I for one am going to be extremely pissed off if Kindle disappears from my iPad.

[1]: http://cdixon.org/2010/06/06/steve-jobs-single-handedly-rest...

> It's worth pointing articles like this whenever the Apple haters berate Apple's "greedy" 30% cut. Particularly when you only spend, say, $1-3 at a time, the fees take up most of that.

I think the greedy part is more based on Apple's requirement that you use their system (and thus pay their 30% vig). If it was voluntary I don't think people would care too much about how big of a cut they took because you could always try and do better on your own.

At least for me, this is absolutely the case. I'm perfectly hapy to pay 30% to a payment processor for small-ticket items, and consider it a good deal.

But the problem comes in when I don't have a choice to run big-ticket items through Braintree, and there's no way for another company to provide a competing small-ticket gateway. In the long-run, that looks too much like a distributor model, which never worked out well for musicians.

"They have the network for POS system"

Carriers have no advantage with regard to POS systems just because they own the network.

Carriers also have enormous regulatory problems. Currently most of them are regulated by the FCC, but when payments start showing up on phone bills the Federal Reserve and Treasury Department start getting curious. This leads to two dynamics that are bad for everyone: delay between each carrier and the government, and delay between each agency within the government as they fight out who gets to be in charge.

Reply on news.ycombinator.com