NONE: Re: ONLINE-ADS>> Auditing Discussion Lists
Re: ONLINE-ADS>> Auditing Discussion Lists
Sean Pfister (seanp_at_cnet.com)
Mon, 20 Jan 1997 13:07:24 -0800 (PST)
Cliff Kurtzman writes:
>I'd agree that a mailing list impression is not the same thing as a web
>page impression, but I do think that mailing lists can still be mapped by
>use of impressions.
>
>... There are also browsers that download all pages
>hyperlinked from the currently viewed page to save the viewer time if they
>should actually click the link. In either case no person necessarily views
>the page even though it shows up as a hit in my log file. So no true ad
>impression on a human is necessarily made.
>
>This problem is magnified on the web when trying to measure
>"click-throughs." ...
I don't want to blithely dismiss these issues you raise, but, as you
explain, there are server-based means of measuring. (For example, I can
count the ads that are cached vs. the ads that are directly delivered from
our server) That's how we're able to discuss the difficulties in the first
place. From a marketing perspective, one of the appeals of the Web when
compared with print is that one knows to a much greater extent that the
"magazine" is actually "opened" rather than lying on someone's desk.
>With an electronic mailing list, I don't believe that a substantial number
>of subscribers will stay on a list they are not reading. Of the 40-ish
>lists that I subscribe to, there are only 2 of which I do not read on a
>regular basis.
I tend to agree--it makes sense, but somehow it would have to be proven.
(to the satisfaction of advertisers). I would do exactly what magazines do:
take a random sample of the maillist and email a survey to that sample.
Have a questionnaire that probes for readership habits. (If the population
to be studied is small, there is a technique called finite population
correction to deal with small universes.) Such a survey would provide
projectable audience data, which is what makes advertisers salivate...
>So I don't see any problem in multiplying the number of list subscribers by
>the number of messages sent to a list with a banner to calculate the total
>number of impressions. While some of the messages won't be read, the
>problem pertains to all lists with probably a similar frequency, so
>relevant exposure comparisons can be made.
I don't see how one can assume that all list subscribers behave in like
fashions. Better to study the issue and determine behavior patterns.
> Giving someone permission to dump mail in one's mailbox is not
>something that people tend to due unless they are _really_ interested in
>the subject matter.
I agree completely. Advertisers probably would too.
>One factor that is quite relevant when gauging list banner impressions is
>whether or not the list owner keeps the list "clean" from bouncing e-mail
>addresses.
Yes, if I were contemplating an ad program using mailing lists, I would be
very interested in the churn rate among subscribers. That and some kind of
survey would help me make a decision.
____________________________________________________
Sean Pfister http://www.cnet.com
Director of Research and Analysis http://www.news.com
seanp_at_cnet.com http://www.search.com
Tel: 415/395-7805 x1529 http://www.shareware.com
=46AX: 415/395-7820 http://www.download.com
150 Chestnut St., SF, CA 94111 http://www.gamecenter.com
____________________________________________________
-------------------------------------------------------------------------
This week's Online Advertising Discussion List sponsor:
Free Subscription to Larry Chase's Web Digest For Marketers. A bi-weekly
email report, containing 25 brief reviews of marketing oriented sites.
http://wdfm.com
-------------------------------------------------------------------------
Online Advertising Discussion List To Unsubscribe send UNSUBSCRIBE
http://www.o-a.com/ to online-ads-request_at_o-a.com
