RE: A reply to Elsevier [Re: Usage Statistics for Web Editions

["David Goodman" <David.Goodman@liu.edu>]

Dear JIm, 

I think you misunderstand in the exchange below: If a company produces
reports using a report generator for the use by different customers of its
journal server, then it is just as easy to produce the reports for 200
customers as it is for 150: one just runs the program the necessary
additional number of times with the additional ip address ranges. It's
just like serving ejournals in the first place--the incremental cost per
customer is not exactly zero, but it is very small as compared with paper.

As you say, the analysis of the logs is the hard part (I've done it also);
but if you have the program to do it, and are not doing it by hand; and if
you're storing the results on the computer and not on paper printouts;
then it is not a significant impact. It is certainly not an impact if the
company is generating reports on all its customers in any case for its own
use, and simply not making them available to some of the customers.  (The
company was indeed apparently producing such reports internally, since it
did send them to those customers who convinced them they would otherwise
cancel multiple titles.)

>> The argument that usage statistics is "costly to provide" doesn't make 
>> much sense once the basic infrastructure is in place, especially if 
>> they already have to provide statistics for a subset of fully licensed 
>> titles so that the account is already set up. 

>Most likely you're wrong on this count.  From my experience, it is very
>costly in terms of infrastructure support to generate these kinds of
>statistics.  I would suggest it takes a lot of dedicated resources to
>generate COUNTER compliant reports, and if one can "drop on the floor" a
>particular set of logs because the statistics aren't going to be provided,
>then it may in fact make a significant impact on the resources needed to
>generate the reports.

In other matters, the company concerned is fully aware of the details of
present and forthcoming releases, as it has a representative on the
committee that is writing them.

The other list to which you are referring is presumably "reedelscustomers"
http://www.lib.utexas.edu/admin/cird/reedelsevier.html, 
maintained by Douglas Dillon of the University of Texas Library, where
extensive discussions of these matters have taken place, in addition to
the few that have appeared on liblicense. Reedelscustomers is independent
of the company, but company representatives often use it for authoritative
clarifications of policy.

David Goodman
dgoodman@liu.edu