But at least I mention lollipops a lot.
First, a warning: I'm definitely not qualified to have this discussion. There are lots of people who know a lot more about the intricacies of the existing systems, the implementations behind them and the technical issues they've dealt with so far. That being said, I have done some stuff. Back in late 1995 at theglobe.com, I implemented the initial version of the credit card processing system. The plan was to charge for VIP membership to the site, but back then there were few-to-no existing web-based systems for charging users. In fact, SSL was a relatively new concept that was only starting to get into the browsers. Through kludging together a few different systems, I got the system working, and learned a little about the pieces of existing processing systems in the meantime. In addition to that, I've had in-depth discussions, that taught me much more, with a friend who worked at govWorks prior to its demise. He was actually one of the people doing work there, instead of working on being filmed for documentaries. He was in charge of dealing with massive amounts of payments, processing them through disparate systems, and making sure that the payments ended up going to the right people. Basically, the sorts of stuff I was really happy I didn't have to do.
To start with, let's look at how a micropayment system that works.
I'm walking down the street, I see a man selling lollipops for a quarter. I give him a quarter; he give me a lollipop. He's happy with his quarter; I'm happy with the sweet goodness of a lollipop. Millions of these transactions happen daily, and they tend to work.
Now he decides to sell his lollipops online, using PayPal for his lollipop sales.
Uh oh.
According to PayPal:
A single variable rate of 2.2% plus a flat fee of 30ยข will apply to all payments, regardless of payment size or funding source.So he'll make -$0.0555 per lollipop he sells.
Clearly, this doesn't work, but pricing structures similar to this are in place for pretty much all credit cards, and any other processing solution you can think of.
We're faced with two options: Either there's a vast amount of money to be made in micropayments that every payment processor out there is too stupid to try and take advantage of, or the economies of it simply don't work.
As there's little that can be done to either prove or disprove the first option, let's focus on the second.
Why would the economies not scale? Why isn't this like getting a 5% discount on lollipops when you buy them in bulk? Why would there be a fixed overhead per transaction?
Simply put, because this is tough stuff.
When you buy something through a middleman, using a virtual currency like a credit card, there's a lot that has to go on for that middle man. It isn't a case of here's my money, give it to lollipop man. What happens if something goes wrong? Do you want your money to go to him if you never got your lollipop? What if he never finds out that your money got taken? What if you only had 25 cents in your account, but try to buy two lollipops simultaneously from two different vendors? Both of those transactions can't go through; does the processor want to get stuck holding the bill? Do they want to bother tracking you down for just 25 cents? What if, after 1 week, you decide that the lollipop isn't quite as grapey as you would have liked, and try to return it? Now not only do all those transaction elements get multiplied by 2, but the middle man better have held onto a record of that transaction. How long should they hold onto the (assumedly) thousands of micropayment transactions that you're making?
Now, stop being selfish. Multiply, again, all the steps and problems above by the (assumedly) millions of people who would be using these micropayment systems.
This is one of those things that requires not only massive amounts of hardware and storage, but also requires a bug-free system (so the programmers had better be decent), and you'd better make sure that everything gets very adequately tested, too. The massive hardware isn't just for the storage space for the billions of transactions you get in X amount of time either, but for making sure that the machines work fast enough so those thousands of vendors and millions of users aren't left waiting too long for their payments to clear.
Using Scott McCloud's example, whoever manages all this is supposed to make it work on 10% of a $0.25 transaction. This means that all that work gets done for $0.025, or 2.5 cents. They get $750 per month for handling 30,000 transactions, and all the headaches that go along with them. If we lower the initial amount to the 1-10 cent range he uses later, it becomes even more ludicrous. The idea of getting a tenth of a cent to handle that amount of processing just doesn't make sense.
When PayPal or other processors ask for a minimum flat fee per transaction, its only partly greed. The other part is to dissuade you from putting that kind of load on their systems. Simply reiterated, this stuff is tough. The economies don't scale because the technology doesn't.
I used PayPal as an arbitrary example, but the same holds true for Visa, MasterCard, Amex, and the real people who handle credit cards, which are the 2 or 3 massive processing houses behind them all. These people already have systems so massive, it boggles the mind. The amount of data they hold, and the number of transactions they clear in any given period of time, is staggering. If I was a real reporter, I might even have some numbers here to back up that claim. But to ask these people or a new start-up to develop, out of the blue, a system that is as secure and reliable as what they have now while simultaneously being able to handle an order-of-magnitude more transactions, is just infeasible.
Or, I should say, infeasible with current technology. We'll have something close to micropayments eventually. They won't be what we currently think of as micropayments, but they'll function in a similar way, in the end. This will happen because computers get bigger and faster at a greater rate than the rate that the number of transactions increases. In addition, competition will eventually lower the fees and rates charged to merchants to a level that makes us think of them as micropayments. Not too many years ago, it wasn't uncommon to have merchants not take credit cards for payments less than $20, whereas now I can use my card in the supermarket downstairs for a $2.00 purchase. Combine that with things like single-use credit cards, so-called tamper-proof smart chips embedded in the cards allowing them to authorize limited disbursal of funds without immediately contacting the central processor, and other advances and changes, and eventually these things combine into a useable system.
Eventually.
