The purpose for the statistics was to generate some demographic data
about the list activities for the purpose of garnering corporate
sponsorships for the list. I had previously considered starting a list
for-profit. NTBugtraq is not-for-profit. Gathering the statistics I
suggested would give me additional information (to my web archive
statistics) that would allow me to better visualize the overhead and
traffic patterns for my membership (whom I'd hope to market a future
for-profit list).
Being somewhat marketing oriented I've previously been involved in
efforts to quantify market potential vs. cost of ownership. Some of the
statistics address market potential (subscribe vs. remove over time
could represent forecasted membership at some future date), obviously
number of posts/queries represents cost of ownership (overhead). Etc...
As a statistics pro, you can probably appreciate the value that
statistics can yeild in creating the answer you desire...;-]...and in a
situation where you are trying to put a business case together, creating
desirable statistics can be useful (and detrimental, obviously). Until
you see the statistics its often not possible to accurately determine
whether or not they'll be valuable. Part of the value of statistics is
that they may disclose previously unrealized aspects of their subject
matter. The number of "info refcard" commands being processed might show
its better to have a web page with that information available (and then
the stats on the web page might prove that didn't make a difference).
Windows NT makes performance counters available for most services. These
counters are usually very detailed. Exchange, for example, has hundreds
of counters available which you might use to gather statistical
information. I've found these counters very valuable in assessing
services, and thought something similar in Listserv (and you'll notice I
didn't suggest NT functionality) would be equally valuable.
Like most things, its not necessarily of value to eveyone. I didn't
believe that the performance hit would be significant, nor that it would
increase administrative overhead. I did believe it would give a list
owner the ability to get summarized information about list activities
that could be useful.
I should also point out that I honestly believe that for-fee list
management is a hugely under-tapped market. With push technologies
rapidly becoming popular (of which many already offer email delivery)
the potential growth of fee-based mailing lists is very good. I say this
as a member of MCI's Internet Solutions Center which provides hosting
services to customers for a variety of servers (web, sql, commerce,
etc...). IMO, one of the reasons this remains untapped is because lists
are seen as huge consumers of infrastructural resources (bandwidth in
particular). By showing the list in a different light it may be easier
to convince companies to use them as a marketing device.
So if the consensus is that most lists are run because they have to run,
or because someone wants to run them, and not to profit someone, then
the information is probably trivial and of little value.
If a list is run to make money, impact market share (say, of an
advertiser/sponsor), or deliver marketing information, then I think the
information is very valuable.
Cheers,
Russ
R.C. Consulting, Inc. - NT/Internet Security
|
