Micropayment

From CEOpedia
Micropayment
See also


Micropayment - is a very small electronic payment, usually below one dollar (or one euro) and sometimes only a fraction of a cent, made for a single piece of digital content or a single service event such as reading one article, playing one song or making one API call. The engineers who built the first systems in the mid-1990s described the target in plain terms: a protocol "designed to support purchases costing less than a cent"[1].

There is no official threshold. Some authors draw the line at a few cents, others at a dollar, and card networks have used their own internal definitions for "small ticket" transactions. What unites all of them is a problem rather than an amount: at these values the cost of moving the money can be bigger than the money itself. A micropayment system is therefore any arrangement (technical, contractual or both) that brings the cost of processing a payment down far enough that selling something for a few cents makes economic sense.

History of micropayments

The idea is almost as old as the commercial Internet. In the early 1990s most online payment projects (DigiCash, CyberCash, First Virtual, NetBill, Open Market) were built for purchases of roughly 5-10 dollars and up, because each transaction cost several cents plus a percentage. Below about 50 cents, the Millicent team argued, those costs became "a significant or even dominant component of the total purchase price"[1].

Millicent

Millicent was developed at Digital Equipment Corporation and presented at the Fourth International World Wide Web Conference in 1995 by Steve Glassman, Mark Manasse, Martín Abadi, Paul Gauthier and Patrick Sobalvarro. It rested on two ideas. The first was the broker, an intermediary that handled accounts, billing and relationships with vendors so that a customer did not have to open an account with every website. The second was scrip - digital cash valid only at one specific vendor. Because scrip could be checked locally by the vendor's server, without contacting a bank or performing expensive encryption, each payment cost almost nothing to verify[1].

PayWord and MicroMint

In 1996 Ronald Rivest and Adi Shamir, two of the three inventors of RSA, published "PayWord and MicroMint: Two simple micropayment schemes". They said openly that they were inspired by Millicent. Their goal was to cut down the number of public-key operations per payment, replacing them with hash functions, which on the hardware of the time were about 100 times faster than RSA signature verification and about 10,000 times faster than signature generation[2].

PayWord works like this. A user signs one "commitment" to a chain of hash values (paywords), each worth one cent in the authors' example. Every page viewed releases the next payword, and the vendor checks it with a single hash operation. At the end of the day the vendor reports only the last payword received to the broker, who charges the user and credits the vendor. MicroMint went further and dropped public-key cryptography altogether: coins were represented by multi-way hash collisions, cheap for a large broker to mint in bulk but expensive for a small forger to fake. The authors' attitude was refreshingly relaxed - "we don't sweat the small stuff", comparing a lost micropayment to a nickel lost in a candy machine[2].

Nielsen's prediction

On 24 January 1998 usability expert Jakob Nielsen published "The Case for Micropayments". His argument was about control: "those who pay for something control it", and websites funded by advertisers would end up serving advertisers instead of readers. He suggested users should be willing to pay about one cent per page and predicted that most sites not financed by product sales "will move to micropayments in less than two years"[3]. He also attacked subscriptions as a "city wall" that keeps casual readers out and breaks the linking logic of the Web.

It didn't happen. Not in two years, not in ten.

Szabo and mental transaction costs

The most influential critique came from computer scientist and legal scholar Nick Szabo. In his 1999 paper "Micropayments and Mental Transaction Costs" he assumed, for the sake of argument, that the technological cost of a payment system was zero - and showed that micropayments would still struggle. The obstacle was the customer's brain. Every priced decision forces a person to think: Is this worth it? How much have I spent today? Did I get what I paid for? Szabo argued that these mental transaction costs (coming from uncertain cash flows, costly observation of product attributes and costly decision making) set the real lower bound on how finely things can be priced, and that this bound sits far above "a few cents or even fractions of a cent"[4]. His conclusion favoured flat pricing and bundles. It also connects neatly to transaction cost theory, where costs of searching, deciding and monitoring matter as much as fees.

Andrew Odlyzko of the University of Minnesota added the economist's version in 2003, opening with a line that became famous among payment specialists: "Micropayments are the technology of the future, and always will be." His point was that the barriers were rooted "in economics, sociology, and psychology", not in technology, and that cards were themselves getting cheaper at handling small sums[5].

The cost problem

Why can't one just use a payment card for 25 cents? Because card pricing is usually built from two parts - a fixed fee per transaction and a percentage of the amount - paid by the merchant through its acquiring bank or payment processor. The percentage scales down with the price. The fixed part does not. And for very small amounts it is the fixed part that dominates.

The general formula for the processing cost of a single payment is:

\[ C = F + r \times P \]

where:

  • C - total processing cost of the payment,
  • F - fixed fee per transaction,
  • r - percentage (ad valorem) fee,
  • P - payment amount.

The effective cost as a share of the price is:

\[ \frac{C}{P} = \frac{F}{P} + r \]

As P falls toward zero, the term F/P grows without limit. That single fraction explains thirty years of micropayment history.

Example of the cost problem (hypothetical)

Assume an online publisher sells single articles for $0.25 and pays a processor a fee structure of $0.30 per transaction plus 2.9% of the amount - a common pattern for online card payments, though actual rates differ by country, provider and contract. The figures below are illustrative.

Price (P) Fee (C = 0.30 + 0.029P) Fee as % of price Merchant receives
$0.25 $0.307 122.9% -$0.057
$1.00 $0.329 32.9% $0.671
$5.00 $0.445 8.9% $4.555
$10.00 $0.590 5.9% $9.410
$50.00 $1.750 3.5% $48.250

For the 25-cent article the calculation is:

\[ C = 0.30 + 0.029 \times 0.25 = 0.30 + 0.00725 = 0.30725 \]

The publisher loses almost 6 cents on every sale. Selling more makes things worse.

Now suppose the same reader tops up a prepaid wallet with $10.00 and then reads 40 articles at $0.25 each. Only one card transaction takes place: $0.59, or 5.9% of the value. Spread over 40 articles, that is less than 1.5 cents per article. Same reader, same content, same card - but the processing cost per article falls from about 31 cents to under 2 cents. This is the logic behind almost every commercially successful micropayment model.

Business models

Since the cost problem cannot be solved by charging tiny amounts one by one through card rails, real-world systems group, pre-fund or hide the individual payments.

Aggregation

Many small charges are recorded on an internal ledger and settled as one larger payment - daily, weekly or monthly. PayWord is an early example: the vendor submits only the final payword of the day[2]. Telecom operators have long used the same principle with carrier billing, where ringtones or app purchases appear on the monthly phone bill. Odlyzko pointed out that such accounted systems are not micropayments in the strict sense, since value doesn't change hands at the moment of purchase[5]. For a business, though, the distinction hardly matters.

Prepaid wallets and stored value

The customer pays in advance (a prepayment) and the provider deducts small amounts from the balance. Transit cards, game currencies, coffee-chain apps and chip-based electronic purses all follow this pattern. The provider also earns from "breakage", the part of the balance that is never spent, and from the float on money held before it is used.

In-app purchases

Mobile platforms solved the problem by sitting in the middle. The user's card is already stored with the app store; a 99-cent purchase is approved with a fingerprint and billed in aggregate. The platform charges developers a commission (historically around 30% for many sales) - expensive by banking standards, but the developer gets an audience that can pay in one tap. Games built entirely on small purchases of virtual items are probably the largest micropayment market that exists today.

Per-article payments and content unbundling

News publishers have repeatedly tested selling single articles for small amounts instead of (or alongside) subscriptions. Results have been mixed. Readers who visit rarely like the option; frequent readers prefer a flat subscription. Economic theory explains why sellers often prefer bundles: price bundling lets them capture more of the variation in what different customers would pay. Odlyzko noted that mixed strategies - offering both bundles and single items, with single items priced relatively higher - can still leave room for per-item sales[5]. In practice micropayments tend to work as an entry point that later converts readers into subscribers.

Tipping and pay-what-you-want

Streaming platforms, blogging tools and creator sites allow viewers to send small voluntary amounts. The decision is emotional, not analytical, which partly sidesteps Szabo's mental transaction cost problem.

Technologies

Stored-value and token systems

Classic stored-value systems keep a balance either on a card chip or in a central account. Earlier digital cash schemes tried to reproduce the anonymity of coins using cryptographic tokens. Millicent's vendor-specific scrip belongs to the same family[1].

Payment channels and the Lightning Network

Public blockchains such as Bitcoin record every transaction on a shared ledger, which makes tiny payments slow and expensive when the network is busy. The Lightning Network, proposed by Joseph Poon and Thaddeus Dryja in a 2016 draft paper, moves most activity off the main chain into a "network of micropayment channels". Two parties lock funds in a channel, exchange as many signed balance updates as they want, and settle only the final result on the blockchain[6]. Conceptually it is the PayWord idea again: many payments, one settlement. Adoption has grown, but volatility of the underlying currency and the need to manage channel liquidity still limit mainstream use.

Web Monetization

Web Monetization is a proposed browser standard developed in a W3C community group. A website adds a link element with the keyword "monetization" pointing to a wallet address; a visitor's browser, through the visitor's chosen monetization provider, can then send payments to the site automatically while the page is open. The provider relies on the Interledger family of protocols (including Open Payments) to move value between different payment networks[7]. The model removes the decision from each page view - the user sets a budget once - which is exactly the kind of fix Szabo recommended for lowering mental costs. The specification describes itself as a work in progress.

Advantages of micropayments

  • New pricing granularity - content and services that are worth only cents can finally be sold instead of given away or bundled. This widens the pricing strategy toolbox and allows finer price discrimination.
  • Alternative to advertising - Nielsen's main argument: when readers pay, sites design for readers. It also reduces dependence on tracking[3]. (Compare with the advantages of online advertising.)
  • Low entry threshold - a reader unwilling to sign up for a 10-dollar subscription to an unknown publisher may spend a few cents on one interesting article[1].
  • Pay-per-use fairness - occasional users pay little, heavy users pay more, which suits APIs, cloud computing, and metered digital services.
  • Support for small creators - independent writers, musicians and developers can earn without building a subscription base first.
  • Machine-to-machine commerce - software agents and connected devices can pay each other per request, a use case that grows with automated services.

Limitations of micropayments

  • Fixed processing costs - as the example shows, the fixed fee per card transaction can exceed the price. Workarounds (aggregation, prepaid balances) bring their own costs and lock-in.
  • Mental transaction costs - each priced decision costs attention. Szabo argued that this cost, not technology, sets the floor for price granularity[4]. The topic is closely tied to the behavioral aspects of decision-making and customer price sensitivity.
  • Preference for flat rates - many consumers will pay more for a flat fee just to avoid the anxiety of a running meter. Nielsen himself noted how people ration long-distance calls but never think about the cost of a light bulb[3].
  • Network effects and the chicken-and-egg problem - users won't load a wallet unless many sites accept it; sites won't integrate a wallet that few users have. Strong network effects favour a few large platforms.
  • Competition from established payment methods - card networks and mobile wallets have steadily lowered the cost of small payments, shrinking the niche that only dedicated micropayment systems could fill[5].
  • Slow change of payment habits - Odlyzko showed that payment systems diffuse over decades, not "Internet time"[5].
  • Fraud and dispute handling - at a few cents per transaction nobody can afford manual investigation of complaints. Systems accept some losses by design, as Rivest and Shamir openly did[2].
  • Regulation - stored balances may fall under electronic money and anti-money-laundering rules, which raises compliance costs for small providers.

Related approaches

Micropayments compete and coexist with several other ways of monetising digital goods:

  • Subscriptions - flat fee for unlimited or bundled access; low mental cost, high entry barrier.
  • Freemium - basic version free, premium features paid.
  • Advertising - user pays with attention and data instead of money.
  • Metered paywalls - a set number of free articles, then a subscription offer.
  • Donations and memberships - voluntary support without direct exchange.

Choosing between them is a question of business model innovation, audience habits and the economies of scale a publisher can reach. Pure micropayment models have rarely won on their own. Hybrids - a wallet inside a platform, a tip button next to a subscription, a per-article option for occasional readers - are where they usually survive.

References

Footnotes

  1. ↑ Glassman S., Manasse M., Abadi M., Gauthier P., Sobalvarro P., The Millicent Protocol for Inexpensive Electronic Commerce, Fourth International World Wide Web Conference, 1995, Abstract and section "Electronic Commerce Background"
  2. ↑ Rivest R.L., Shamir A., PayWord and MicroMint: Two Simple Micropayment Schemes, Security Protocols Workshop, 1996, sections 1-2
  3. ↑ Nielsen J., The Case for Micropayments, Nielsen Norman Group, 24 January 1998
  4. ↑ Szabo N., Micropayments and Mental Transaction Costs, 1999, Introduction
  5. ↑ Odlyzko A., The Case Against Micropayments, Financial Cryptography 2003, sections 1-3 and "Bundling"
  6. ↑ Poon J., Dryja T., The Bitcoin Lightning Network: Scalable Off-Chain Instant Payments, 2016, Abstract
  7. ↑ Web Monetization Community Group, Web Monetization Specification, section "Model"
Micropayment — recommended articles
Digital cash — Electronic purse — E-commerce — Transaction cost theory — Price bundling — Fee structure — Pricing strategy
Retrieved from CEOpedia