[8503] in Commercialization & Privatization of the Internet
Online Shopping and Banking? (Pass 2)
daemon@ATHENA.MIT.EDU (Tansin A. Darcos & Company)
Sun Nov 21 09:42:28 1993
Date: Sun, 21 Nov 1993 08:19:55 -0500 (EST)
From: "Tansin A. Darcos & Company" <0005066432@mcimail.com>
Reply-To: "Tansin A. Darcos & Company" <0005066432@mcimail.com>
To: bobk@cyberspace.com, Everyone Else Lurking on Com-Priv <com-priv@psi.com>
>From: Paul Robinson <TDARCOS@MCIMAIL.COM>
Organization: Tansin A. Darcos & Company, Silver Spring, MD USA
-----
This is a further addition to my earlier remarks.
> The options to shop and do banking transactions electronically
> have proven to be a highly desirable services on other
> commercial networks. The Wall Street Journal recently had an
> article that said PCFlowers on Prodigy has become one of the
> top five FTD sellers of flowers. Microsoft announced today
> that it is getting into the home electronic banking business
> with US West and US Bankcorp. I'm sure that the Internet will
> soon have such services.
Perhaps. We need to fix two important issues to cover: (1) availability
of the service and (2) price to cost as opposed to price to value.
Availability of Service.
AT&T has announced you (meaning some businesses) can now get your
telephone bills on diskette (the TV ad showed a 3 1/2" diskette.) The
local telephone company will supply customers with magnettic media, but
only if they pay something like $500 a year for the service, and only in
1600 bpi mag tape. (My other office gets its phone bills that way.)
What it means is that for the average person, detailed billing from their
local telephone company is unavailable. And yet, for those of us who use
computers, it would be cheaper (at 19c for 5 1/4 or 29c for 3 1/2) for the
phone company to send the entire bill (except payment page) on diskette,
because we could read the data ourselves. And probably most companies
would be willing to pay an extra charge ($3 to $5) for the service, which
would probably cost less for the phone company than sending out pounds of
phone bills.
But to think of things like this requires innovation to do so. And local
phone companies are apparently afraid of innovation if it means changing
decades-old methods of handling service.
Quicken has cut a deal with a credit card company to allow you to download
your credit card data so that you can read it into Quicken, or the bank
will mail you a diskette of your transactions for an extra $3 a month.
Price to Cost vs. Price to Value.
Typically, companies that sell information in online forms have this
impression that their data is intrinsically of tremendously higher value
in computerized form than in printed form. For systems that provide
search capabilities, that may be so; but for most systems, the extra
charges are simply soaking people who are willing to pay for it. I see
nothing wrong with charging whatever the traffic will bear, I just wonder
why these companies scream and complain when a cheaper competitor comes in
and steals customers away from them at much reduced prices.
Pricing to cost means that you charge your cost of providing the service
in question plus your usual profit. For, say, a system that charges $150
a year for the printed edition, should probably charge 10 or 15c per page.
Or some number related to the approximate yearly cost of the printed
edition.
Pricing to value means that they will charge online transaction fees plus
per-lookup fees plus anything else they can think of. So that if someone
wants a particular article, they can be looking at $3-5 a page, a
tremendous increase over the cost of the printed edition.
Pricing to cost would be the $3 per month charge by the bank for copying
to diskette, figuring in the extra work involved and the additional media
cost. Pricing to value is the $500 a year that the phone company charges
for mag tape listings. (Of course, $120 of that is covering the cost of
sending the tape by Federal Express once a month. Note that this only
gives the customer the right to use and copy the tape and information in
it; the tape itself must be returned to C&P Telephone within 10 days, or
another $20 will be charged for the tape).
> How far is the Internet from being able to provide reliable avenues
> for online financial transactions? What would it take for reliable
> email-ordering on the Net? How about banking and stock
> transactions? Is the primary roadblock the security issue of
> sending sensitive financial information (such as credit card
> numbers) over the Net?
That may be part of the problem. Public Key Encryption can help, but you
still need to have public registration of people's public keys, a system
that is not in place now. And, of course, the typical nasty intrusions of
the Federal government (witness the examinations into the company selling
a book of encryption techniques) if there was a lot of Public Key
encrypted traffic going across networks.
Currently, one can order stock from an automated system via telephone for
3c per share with a minimum $35 service charge. At some point, someone is
going to ask why, if they are doing all the work of placing the orders
themselves and sending it to the company's computer, is the minimum
commission and per-share charge so high? If all the work is done by
computer, as long as there is nondeniability of ordering, you can cut that
minimum commission down to perhaps $3, since the customer does all the
work; this would encourage the small investor who perhaps wants to just
order 5 or 10 shares. The computers do all the work.
But to have this capability, you need three things: (1) that the company
accepts the order and provides nondeniable proof of understanding and
acceptance; and (2) that the means to place the order provide nondeniable
proof that the order was placed by the customer, and of the stock being
ordered.
This eliminates DKing of stock transactions. ("DK" is when the customer
Denies Knowledge, e.g. repudiates a transation). This eliminates the
broker claiming he never got the order. This eliminates a customer using
blue skying to screw the broker. (A customer places an order for a stock
he expects to go up in price dramatically; the broker takes the order,
then the stock falls, and the customer refuses to pay the settlement
because the stock is blue skied (not registered) in his state, forcing the
stockbroker to eat the cost). Since each stock can be identified as to
what states it can be sold in, the broker can refuse to trade stock that
is blue skied.
We also need to change the regulations so that ordering can be done by
computer without requiring a stockbroker to do so, e.g. that direct orders
can be placed without human intervention, if this is not already possible. We
also need better checking models so that if someone orders lots of stock,
that it is an order and not their computer misapplying a computer trading
program they are using either accidentally or on purpose.
> In addition to the mechanics of how to implement such services,
> I'd like to hear some thoughts and predictions on the evolution of
> this important and probably inevitable development and its effect on
> the character and nature of the Internet as we now know it.
Well, as one place put it, they get so many requests for WWW service that
just to handle the traffic they might need to put a Cray out to handle the
load. Someone else has mentioned that he could put up a correctly tuned
Sun for $50,000 and be able to take 1,000 simultaneous FTP transfers of
the 4.11 X Sources, and the system would still be usable for all the other
usual purposes (except Internet, for obvious reasons).
If there are a lot of people doing transfers, it means we need serious
increases in bandwidth. In order to get reasonable costs, it may be
necessary to totally deregulate the communications industry - which means
allow anyone to run wire anywhere someone wants to pay for it - and make
no considerations over what they run on that wire, be it data, facsimile
or voice.
The United States has 160 area codes and has so badly run out it is
changing the more than 30-year-old Numbering Plan Area system to allow
any number from 200 to 999 to be an area code. This implies that there
are hundreds of millions of people making phone calls at some not very
expensive cost but not zero. So what happens if someone figures out a way
to put up a 128K line and stat mux it to run a data connection and
compressed voice at the same time, so that someone can call over the
Internet to anyone else. This brings new meaning to the term "voice mail".
If someone says he has to check his voice mail, and goes to his computer,
it's going to make people wonder.
With the nominal cost for connections over the internet to be zero, it's
only a matter of time until someone figures out how to reduce those huge
phone bills by encouraging more places to put at least a 56K line in and
run compressed voice circuits over the Internet. We've already got
facsimile going via the TPC.INT domain; how long before that capability
includes delivery of voice messages? (Or even more interesting, ability
to return responses.)
If there is so much competition that a T1 can be gotten for $100 a month,
just figure at 8c a minute, how many call minutes have to be taken off
the long distance network before the T1 connection pays for itself in
reduction of long distance telephone calls. If it reduces the number of
long distance voice calls by 60 minutes a day, then all the data traffic on
that T1 rides free!
A T1 is the equivalent of 23 telephone trunks, if memory serves me. If
that T1 can substitute for say, 12 56K data feeds and 11 voice trunks
during business hours, and a full T1 data feed the other 12 hours a day,
then to make this happen, all we need is the hardware to digitize voice
into something that will fit into a 56K data stream (not hard, I think
most encoding systems are only using 8-16K a second) and to decode it,
plus the ability to tell a PBX system that for some accessible numbers,
the least-cost routing is over the Internet.
We also need "Internet Dial Tone" or the ability to indicate the system is
ready to accept a call and a means to route traffic; "Busy" and "Ring" as
well as "Answer" or "No Answer" indications.
Most educational institutions have the hardware installed to handle a
voice message delivered in real time; the problem is figuring out how to
do switched connections to someone else's computer. And we need to find
out how to compress and decompress the message. Many of the newer
computers have the capability with Soundblaster boards (or the
equivalent). All they need now is a microphone and an internet
connection, plus the APIs to allow Internet Dial Tone Service.
Anyone interested in trying this as an experiment?
---
Note: All mail is read/responded every day. If a message is sent to this
account, and you expect a reply, if one is not received within 24 hours,
resend your message; some systems do not send mail to MCI Mail correctly.
Paul Robinson - TDARCOS@MCIMAIL.COM
Voted "Largest Polluter of the (IETF) list" by Randy Bush <randy@psg.com>
-----
The following Automatic Fortune Cookie was selected only for this message:
You might have mail
