Micropayments

1
Micropayments

• Presented by Anthony Wood
• To CRAB, March 2, 2004

2
Some Quotes

• Micropayment technology in and of itself is
  about as interesting as new and improved dish
  soap. Mike Gaynor, founder of RedPaper
• Users should be willing to pay, say, one cent
  per Web page in return for getting quality
  content and an optimal user experience with less
  intrusive ads. Jakob Nielsen
• In many ways, theft is the unspoken
  inspiration for micropayment systems. Clay Shirky

3
Outline

• Micropayments
• A Littered Landscape
• Current Schemes
• Coin-flipping
• Rhetoric

4
Micropayments

• Means of making money (or at least recovering
  costs) from online content
• Low prices (lt 1), high volume (hopefully)
• Alternative to subscriptions, aggregations
• Allows customer to quickly purchase only what she
  is interested in
• Moves content creators closer to consumers

5
Motivation

• Even free online content costs something
• Currently funded by
• Advertisements (banners, popups, sponsorships)
• Subsidies
• Good-will of content creators
• Using CCs invokes high transaction costs
• Prevent Spam? DoS attack?

6
History

• Digital Silk Road (Hardy, Tribble, 1993)
• Include coins in packets
• Millicent, NetBill, NetCard, PayWord, MicroMint,
  DigiCash, E-Money, E-Coin
• Biggest losers to date venture capitalists
• Continuum from tiny, per-packet schemes to
  normal online banking (EFT, credit-cards, etc)
• Current focus is on smallnot tinybank-mediated
  access control for WWW

7
Design Considerations

• Double-spending or forgery of coin
• Mediated online transactions, or offline cash
  alternative
• Availability of online principals
• Converting coin to real money
• Overhead of each purchase
• Necessity of custom software
• Try before you buy
• Handling fraud, chargebacks
• Does user buy content, or only viewing rights?

8
Current Schemes

• Peppercoin
• Bills customers CC periodically
• Paystone
• Like bank account
• BitPass
• Like debit card
• Anonymous

9
Peppercoin

• Founded by Rivest and Micali
• Merchant fees 5-10
• Merchant uses PepperMill (Java app) to
  encrypt/enclose content in PepperBoxes
• Customers install PepperPanel
• Purchaser downloads PepperBox, pays using
  PepperPanel, decrypts contents
• Peppercoin pays merchant probabilistically on
  some transactions
• Peppercoin charges customers CC monthly

10
Paystone

• Merchant fees
• 10 cents 5 under 5 (25 cents minimum)
• 30 cents 3 over 5
• Customers load account with money
• Can walk into Bank of America branch
• Merchants create paylinks to PayStone,
  including encrypted return URL
• Purchaser follows paylink, enters email/password,
  redirected back to content
• No software installed

11
BitPass

• Founded by 2 Stanford grad students
• Merchant Fees
• 15 under 5 (1 cent minimum)
• 50 cents 5 over 5
• Customer loads account by PayPal, CC
• Follow link to BitPass, authorize payment,
  redirected back to seller
• Seller installs gateway to control access
• PHP, Perl CGI, mod_perl, ASP.net

12
Coin-flipping

• Probabilistic payment scheme
• E.g., with probability 1/200, user pays larger
  amount otherwise, access is free
• Expected payment is therefore small negotiated
  amount
• Principals User, Vendor, Bank
• Goals efficiency, fairness, authentication
• Process
• Pre-processing
• Coin-flip rounds

13
Pre-processing Stage

• Vendor creates one-way chain
• y f(f(f( (x))))
• Vendor sends y and proof of x to User
• User verifies proof of x
• User creates its own one-way chain
• y f(f(f( (x))))
• User sends (y, y) and signature to Vendor
• User sends proof of x to Vendor
• Vendor verifies proof of x and signature

14
Pre-processing Stage

• User has
• Own chain y, , x2, x1, x
• Vendor pre-image y
• Vendor has
• Own chain y, , x2, x1, x
• User pre-image y
• User signature of (y, y)

15
Coin-Flip Round

• User reveals next pre-image in y chain
• Vendor reveals next pre-image in y chain
• XOR of pre-image bits defines coin-flip
• If User refuses to pay, Vendor takes (y, y),
  signature, and transaction record to third-party

16
Coin-Flip Round

• U -gt V x4
• Vendor now knows flip-result
• If User knew, she could abort protocol and not
  pay (if bad flip)
• V-gt U x4
• Both can verify that xi1 f(xi)
• Flip x4 XOR x4 (suitably biased)

17
Arguments For

• Those who pay, control content
• Micropayments align content providers interests
  with consumers
• Electricity and long-distance both meter usage
  and are successful
• For small amounts, users dont stress about cost
• Subscriptions force all/nothing decisions in
  advance of viewing

18
Arguments For

• Subscriptions wall-off from linking, browsing,
  and spidering
• Art is not a commodity than can be replaced by
  free alternatives
• (Buy McClouds comic strip)
• iTunes and PayPal are not far from micropayments

19
Arguments Against

• To continue, click here to contribute
• 0.50 to the presenter

20
Arguments Against

• Littered landscape of failed companies
• Flat-fee schemes are more successful
• Metered charging is successful only for
  monopolies
• Very small transactions can be very hard to value
• How much is half of one Wired article worth?
• Existing payment infrastructure is becoming more
  flexible

21
Arguments Against

• Aggregation smooths variability in content, and
  is more efficient
• Mental transaction costs make deciding to buy
  more expensive than item being purchased
• Viewers can always find something else for free

22
Questions

• How much is this presentation worth? The entire
  seminar? A UVA education?
• Have you purchased an item for less than 1
  online?
• If user does not have to explicitly decide when
  to purchase, how does he prevent fraud?
• Are micropayments a solution to a non-existent
  problem?