CPM

From: Avram Grossman <AvramG_at_e-novative.com>
Date: Tue 10 Jul 2001 10:59:20 -0500

An article by Stewart Alsop back in January 2001
addressed a topic I'd like to see a debate on: Cost
Per Thousand (CPM) versus some kind of alternative
"Value Price" model.

The Nickel Backgrounder:
In traditional circulation-based distribution media,
a CPM of course makes sense. This I believe is also
the origin of the CPM models used today and dates
back to the invention of the press. In TV (and radio)
advertising, the stations live and die by their
Nielsen (or Orbitron) statistical projections to
derive at an estimated circulation figure for which
the CPM is then based. The advertiser is then charged
that calculated CPM figure.

Using the Internet, banner ads became yet another
variation of the CPM theme, where at least
differentiated from TV/Radio, one knew exactly how
many impressions where made, and how many click-through
hits were counted. In both impression-counting and
click-through counting, the effectiveness and "value"
of the ad are neither measurable nor qualified. There
must be a reason why banner advertising went from $.20
per impression in 1995 to less that $.02 per impression
today. Every variation to measure, qualify and project
banner advertising pricing I've come across seems only
an attempt to explain or justify the CPM mode - and
it isn't working.

Other pricing models might take on some "value pricing"
theme, most likely influenced to some extent by the
number of "eyeballs" that pass through the site. As an
example, our local swap meet charges by the size of booth
space one rents for the day, with some variations on
price attributed to positioning, and I'm sure influence
by popularity of their swap meet, i.e., number of expected
visitors. But for the most part, the "advertiser" is
buying value, not visitors.

OK all you marketing statisticians. One challenge in
this concept might be to derive at an equation that
can approximate a "value price" taking into account
such factors as the "brand recognition" factor of the
advertiser, position on a web page, and of course some
influence of the impression count.

As a spin-off to this, I mentioned "Brand Recognition
factor" above. Is there such a thing and if not, can we
establish this as new way to measure or rate the advertiser.
My thinking is that, by way of example, Joe's Car Detailing
giving away a free car detail places an ad on a highly
viewed web site, the effectiveness of the ad for Joe many
not be the same if, say, Maxell giving away a free blank
tape. The known brand gets a higher ranking factor and
therefore should pay more that Joe (this is a question.)


Avram Grossman / President & CEO (AvramG_at_e-novative.com)
---------------------------------------------------------
e-Novative.com, Inc. http://www.e-novative.com
42 Corporate Park, Suite 250
Irvine, California 92606-3116
Tel: 949-660-0450 Fax: 949-833-9882



Received on Tue Jul 10 2001 - 10:59:20 CDT