NONE: Re: ONLINE-ADS>> Auditing Discussion Lists
Re: ONLINE-ADS>> Auditing Discussion Lists
Cliff Kurtzman (cliff.kurtzman_at_tenagra.com)
Sat, 18 Jan 1997 18:31:28 -0600
Sean Pfister <seanp_at_cnet.com> writes:
>
>
>I don't see how impressions can map to a listserv. You have no way of
>knowing what percent of the email has been opened.
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.
When I look at web page accesses in my access log, some of them are from
robots or people using software that downloads the site for possible
offline viewing. 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've found that the number of "false" click throughs is
often significant enough to materially affect the apparent click-through
percentages that are reported to the advertiser. This stems not only from
robots and the like, but also from people that have clicked the link twice
because the new page didn't load fast enough the first time. Once you
correct for these factors, you may find your ad is being clicked upon by a
significantly lower number of true human eyeballs than you had thought.
This problem is not isolated to the web. I never read some of the dozens
of magazines I'm sent each month. Still, the magazine's advertisers are
told I'm part of their circulation. If I do a direct mailing by postal
mail, some of them will be circular filed without the recipient opening the
mail, but still a direct mailer service will charge you for the address.
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.
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. Of course, when sponsoring a
discussion list for a period of time, the advertiser knows that the banner
is being delivered to exactly the same audience over and over again, which
is rather different than what happens with a web banner. It is probably
also relevant in many cases that a subscriber to a mailing list or
discussion list on a particular topic is generally a much more qualified
and interested audience than the average browser of a web site on the same
topic. 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.
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. On lists I run and that I mail to on a monthly basis, I know
with some certainty that each month that about 2-3 percent of my
subscribers will have closed the account from which they subscribed (and
usually failed to unsubscribe from the list). If a list owner does not
keep their list clean, then it will quickly become filled with bad
addresses, so a subscriber count will materially misrepresent the true
number of deliveries. (How does C|Net deal with this with 650,000
subscribers? You must easily have 20,000 bounce messages a month!). In
addition, advertisers should also take into account that with each mailing
to a monthly list, 1 - 2 percent of the addresses will bounce due to
temporary problems at the receiving site. All this should be taken into
account by mailing or discussion list advertisers, but does not
fundamentally alter the use of impressions to count ad exposure.
--Cliff
Cliff Kurtzman
The Tenagra Corporation
281/480-6300
sponsor of The Tenagra Awards for Internet Marketing Excellence
http://awards.tenagra.com/ Application deadline: January 31, 1997
-------------------------------------------------------------------------
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
http://www.o-a.com/
