Note: This is a “classic” Interesting Thing of the Day article from over 10 years ago. It has not been edited recently, so it may contain broken links, outdated information, or other infelicities. We plan to eventually update or retire most classic articles, as time permits.

In 1994, when the Web was still young, I attended an internet technology conference. A well-known programmer who had written some internet software for the Macintosh was discussing the commercial potential of the internet, and as an illustration, he told the following story (which I’m reconstructing from memory as best I can, as I’ve never been able to find independent confirmation of the details). Back then, there was a relatively new internet payment service called First Virtual, which employed a clever mechanism whereby one person could send money to another over the internet—even in an email message. A man signed up for the service but was skeptical of how well it would work, so he did an experiment. He posted a message on a newsgroup asking for help testing the service. Everyone who sent him US$1 through First Virtual would receive, for an entire year, a new limerick each day by email. Within three months, over 60,000 people had signed up. This was a mixed blessing—the experiment worked, but he encountered major technological problems in sending all those email messages, and soon discontinued the offer.

Enter Credit Card Information Here

First Virtual went out of business years ago, largely because of the proliferation of banks and services that allow merchants to accept credit cards over the Web. By comparison, First Virtual was awkward and low-tech. And certainly for most purchases one might want to make online, credit cards are a convenient way to do so. But there are a few major difficulties with online credit card purchases. First, they’re fine for paying a merchant, but what if you want to send money to a friend or family member? If the recipient doesn’t have a merchant account, you’re out of luck. Second, lots of people don’t have credit cards, or don’t feel safe using them over the Web. And third, credit cards don’t make sense for very small payments of less than a few dollars or so. The reason for this is that banks charge merchants a fee for every credit card transaction that consists of a fixed amount plus a percentage of the transaction. Because of the way the fees are calculated, a merchant could end up paying only $0.60 for a $10 transaction, but $0.33 for a $1 transaction. That means the $1 purchase is not very profitable, and a purchase of less than $0.33 would actually cause the merchant to lose money.

The term “micropayment” was coined to describe these very small purchases. And there are many good reasons you might want to make a purchase of just a few cents. For example, suppose—hypothetically, of course—that a Web site called “Interesting Thing of the Day” wanted to offer a subscription of some sort for just $1 per year (rather than, say, $10). Using credit cards, this would be a losing proposition. And yet, charging higher prices to offset the merchant fees makes the content less attractive to buyers. Other common examples of micropayments are music downloads, news stories, and Web-based services such as translations and image processing—to say nothing of real-world purchases such as feeding parking meters or buying a cup of coffee.

A Small Problem

In the past several years, numerous systems have been proposed to solve the micropayment problem. They vary tremendously in their details and implementations, some being more promising than others. Unlike credit card transactions, which anyone with a card can make instantly, almost every micropayment system requires users to take some preliminary steps to set up an account or enter billing information. In some cases these steps add just a few minutes to the initial transaction; in others, setting up an account can take a week or more. Of course, no one expects to be able to drop a coin into their computer and have it drop out of someone else’s; a bit of effort is reasonable. But the convenience of a system, coupled with its fee structure, are the main factors in determining a system’s success. I could not hope to do justice to the wide variety of micropayment systems here, but I’d like to describe a representative sampling.

One key concept in micropayment processing is aggregation, which simply means grouping several smaller purchases together and billing the buyer once (perhaps monthly) for the total amount. This takes the edge off the transaction fees for merchants, and it gives consumers the ability to make very small individual purchases—such as those 99¢ tracks at Apple’s iTunes Music Store. Of course, you still need a credit card, and bank transaction fees still frustrate merchants, especially when an individual customer makes very few purchases. If a business doesn’t do enough volume to aggregate its own sales, it can make use of a third-party credit card aggregator, which then also takes a cut of each payment. Another approach to aggregation, used by payment processor BitPass among others, is to sell prepaid cards that can be used to make small purchases at participating vendors. In this case, the aggregation happens up front, requiring less effort for the vendor and the payment processor, but more for the customer.

Good as Gold

A very different approach is taken by several companies that use gold (or other precious metals) as the basis of electronic transactions. Under this system, a company maintains a stash of gold securely in a bank vault somewhere. As a subscriber to the service, you can buy some of this gold, which (for a very tiny fee) the company will store for you. In other words, the gold itself doesn’t move—only the ownership of a portion of it changes. You can purchase gold using cash, checks, credit cards, or whatever form of payment you want. Then at any time, you can log into your account and transfer any portion of your gold to another subscriber. (This can be designated according to value, like $0.02, or mass, like 0.02g.) Because it costs virtually nothing for the company to transfer ownership of gold from one account to another, transaction fees are negligible. And as the owner of any amount of gold, you can request payment from your account at any time, in the currency of your choice.

Gold-based electronic transactions have a number of advantages over aggregation. For one thing, they cost less for the recipient. For another, they can be used to transfer money between individuals, even if neither one is a merchant. And they do not require either party to have a credit card. Gold is even more appealing for international transactions. Most banks charge hefty fees to transfer money from one country to another—US$10 or $20 is not uncommon—making it unreasonable to use a check or credit card for small international payments. But transferring gold ownership internationally involves no currency conversion, so it’s no more expensive than making a domestic transfer. There is a catch, however: you can only transfer gold to someone after you’ve purchased it yourself. So unless you’re willing to maintain a balance of gold in your account, each transaction ends up being a two-step process: first put gold into your account, then transfer it to the other person. And, of course, you can only pay for a purchase with electronic gold if the merchant is set up to accept gold payments.

Saved by Mathematics?

A relatively new idea for micropayments, called Peppercoin, promises to solve many of the problems of aggregation while maintaining credit cards as the means of payment (and thus avoiding any need for account setup on the customer’s part). Under Peppercoin’s patented system, which they call “Intelligent Aggregation,” customers are charged for each transaction, but merchants don’t get paid for each one individually. Instead, a complex mathematical algorithm uses statistics to randomize payments, resulting in a far smaller number of total transactions (and resulting fees). For example, if I as a merchant sold 100 items at $0.10 each, I may receive 100 tokens—each with a value of $10.00, but only a 1 in 100 chance of being “live.” So, on average, I’m still getting paid for the value of goods I delivered, but instead of having to deal with massive numbers of tiny transactions, I only deal with a few. The math on which Peppercoin’s system is based seems to be valid, but it makes the most sense when very large numbers of transactions are involved; if you only expect to sell a few items per month, the logic sort of breaks down.

Unfortunately, for small person-to-person transactions, there’s still no magic bullet. Services like PayPal and Western Union do enable an individual to send money to someone else electronically, but the transaction fees still make it impractical for amounts less than $1. Although many attempts have been made, no one has yet devised an electronic equivalent to putting cash in an envelope. —Joe Kissell

More Information

Because internet payment systems appear, disappear, and mutate on a daily basis, it’s difficult to find up-to-date information. John Scott’s comparison of micropayment systems was last updated in 2001. Also in 2001, Michael Pierce compiled what was then an exhuastive list of internet payment systems. Although these pages are out of date, they still contain a great deal of useful information.

Examples of aggregation services include AllCharge, which uses the appalling term “monetization” to describe their service; BitPass; and Webpay.

Gold-based transactions are available from e-gold, GoldMoney, and Pecunix.

The Peppercoin service is in a state of transition between its “1.0” and “2.0” versions, and temporarily unable to offer new accounts.

Person-to-person payment systems include PayPal and Western Union Money Transfer. The Paystone system also offers person-to-person or commercial micropayments without a credit card, and low transaction fees too. The catch? Like gold-based systems, you must already have transferred money into your Paystone account before you can pay someone else.