Payments as an answer to spam

[bear]

On Wed, 14 May 2003, Ian Grigg wrote:
>
>Imagine a world dominated by hashcash.  I am
>a spammer.  I can either run my servers incurring
>a cost of - say - a penny each mail going out,
>or I can offer to pay the recipient of my outgoing
>mail that same penny to read my mail.
>
>I think in the end the latter would win out.  That's
>what economists mean by "more efficient use of
>resources."  Those pennies are re-used, which
>makes them much more powerful than the pennies
>that were burnt up.


That's a good point, but remember TargetFirst?  They were
the first movers on the "get paid to surf the web" business
model.  Basically they paid people for the right to put
advertising on their desktops, and their business died
a horrible death.  I know because I scored some good
deals on laptops at their liquidation sale.

What happened to them, as I understand it, is that the
per-ad revenue got driven down to the point where it was
more expensive to process the payments than the payments
were actually worth.

Whenever anyone writes a check, it costs at least a nickel
for the merchant to clear it.  Clearly you don't ordinarily
use checks for such tiny purchases because (a), almost
nothing can be sold for less than a nickel, and (b), it
would be more time and bother for you to write the check
than you probably consider a nickel's worth of trouble.

If you posit a system for making very small payments,
you've got basically three problems: (1), consumers
will never make micropayments because a decision about
whether to spend money on something involves stress -
in this case stress that's out of all proportion to
the payment itself - and will be resented.  (2), Each
and every consumer system will need configuration and
a relationship with a "bank" somewhere to cash such
tokens made to them - and that requires the attention of
the consumer again to maintain a business relationship
which he won't bother to do unless there's a significant
(more than a few dollars a month) revenue stream. (3),
the cost of processing payments must be kept small
relative to the payments themselves.

I've seen half-a-dozen papers in the last year touting
the ability to keep per-transaction costs on micropayments
down to less than 20% of gross.  All of them did so at the
expense of precision and accountability.  These "solutions"
are non-starters because insofar as they operate without
human attention at any part of the system, if they are not
precise and accountable then they are subject to fraud.
And insofar as they require human attention, people will
ignore them in droves because the payments are too small
to be worth the hassle.

No paper, so far as I know, addresses the basic consumer
psychology issue, that payments of less than ~$1.00 or
revenue streams of less than ~$10.00/month are seen as
too annoying to bother with, especially if they come with
business attachments, contracts, advertising, emails
from the bank about additional services, monthly statements
that incur storage overhead and mental bandwidth, etc...

Every bit of that stuff is a teeny bit of stress.  And
humans will forego revenue streams that come with more
stress than they are worth.

			Bear


---------------------------------------------------------------------
The Cryptography Mailing List
Unsubscribe by sending "unsubscribe cryptography" to majordomo at metzdowd.com