[59] in Public-Access_Computer_Systems_Forum
Re: Network License Agreements
daemon@ATHENA.MIT.EDU (popmoi%uu2.psi.com@RICEVM1.RICE.ED)
Thu Apr 23 12:27:57 1992
Date: Thu, 23 Apr 1992 11:22:57 CDT
Reply-To: orens%moi.com@RICEVM1.RICE.EDU
From: popmoi%uu2.psi.com@RICEVM1.RICE.EDU
To: Multiple recipients of list PACS-L <PACS-L@UHUPVM1.BITNET>
----------------------------Original message----------------------------
I've been following this whole discussion of network license agreements
with some interest. It seems to me that the topic under discussion is not
really license agreements, but information pricing, which is currently a
"hot topic" around most commercial information companies. I think we all
know that producing information is very expensive and labor intensive.
The new technologies offer the tantalizing prospect of enhanced revenue and
less costly distribution, but publishers and distributors are understandably
nervous about the implementation of these technologies. The goal is to
provide useful information products at affordable prices while generating
enough money to make it profitable for people to develop and continue to
provide these products.
Last year I developed a hypothetical pricing scenario for a presentation I gave
at the ALA annual meeting. It assumes a hypothetical database which
occupies 5 gigabytes of disk space and cost us (the distributor) about
$20,000 in combined processing and labor time to implement on our system
(a not untypical figure). It factors in our disk storage costs and takes
into account a royalty to the database producer of 50% of all revenue
(again, not untypical). The last major factor is an anticapated payback
period of 2.5 to 3 years.
I developed three different pricing models for this database:
1. Uses current online pricing models of high connect time charge plus a
nominal record viewing charge. Assuming an average of 40 hours of use per
month and 1200 prints per month, we would have to charge $90 per connect
hour and 80 cents per item viewed to make the payback period.
2. Uses a high charge per item viewed and a nominal connect time charge.
Assuming the same usage as #1 (which ignores the fact that pricing
changes will affect user behavior), we could make the same break-even point
by charging $10 per connect hour and $3.50 per item viewed.
3. Uses a distributed model to assume that users at five different sites
will load the most recent 1/3 of the database locally, and link to our
online system for the backfile use. For argument's sake, I assumed that
2/3 of all use will be on this most recent 1/3 of the database. Under
this scenario, holding the online pricing for the backfile as in either of the
two preceding models, we would have to charge each of the individual sites
about $5,000 per year to make the same break-even point.
This is by no means a complete cost/pricing model, but I hope it serves
to put some typical "real-world" sort of numbers around the philosophical
discussions. The whole issue of pricing for electronic information is
an ongoing experiment in which we are all participants. It may be
traumatic, but it ain't boring!
Oren Sreebny
Database Development
Maxwell Online
orens@moi.com
