NONE: Re: ONLINE-ADS>> Online marketing research dilemmas
Re: ONLINE-ADS>> Online marketing research dilemmas
Karl Zetmeir (kzetmeir_at_gvi.net)
Tue, 21 Jul 1998 08:16:08 -0500
Tim Lee <TLee_at_webcmo.com> wrote:
>>>First you see a tree, then a forest, then a tree in a
>forest. As a marketing researcher, first you have to
>master the analytical techniques, then you have to forget
>them before you use them.<<
I'm not the statistician you are, Tim ...
However, there is a very basic point ... and a grievous
error ... the so-called market research firms have made
in their approach to Internet analysis that, in my
opinion, invalidates the entireity of their work.
How they've been able to sell the results of their work is
totally beyond me. In an article from MacNews ...
apparently many Internet advertisers are bringing them to
task also.
http://www.zdnet.com/macweek/1223/in_webstats.html
Here's the gist of the error ... but first here's the
research firms' current state of the art methodology as I
understand it. They have compiled lists of Internet users
through the use of RANDOM DIALING telephone systems. By
using this panel selection technique it's impossible to
have a clue as to the relevancy (pardon the intended pun,
Relevant Knowledge!) of the selected sample to the whole
of the audience universe.
To use your metaphor, it is NOT ENOUGH to see a tree ...
the question is WHAT KIND OF TREE IS IT ... AND WHAT
PERCENTAGE OF TREES IN THE FOREST ARE SIMILAR? If the
tree we saw was an oak and not a maple ... if that is the
only oak tree we saw ... was that oak the ONLY oak in the
forest, thus being 100% of universe ... or one of many
oaks thereby being a very tiny representation of all oaks?
In other words, the question is one of MARKET PENETRATION!
If you don't know what percentage of the population you
have measured, then you have only measured the sample and
can NOT extrapolate that sample to the universe as a
whole!
Therein lies the problem for Nielsen ... as competent
market statisticians ... they understand the problem ...
others who think that you can place software on ANY
Internet user's PC and get reliable information, simply
don't understand statistics.
For example, if a sample taken by randomly calling
particpants was made in Florida, we might conclude there's
an extremely high percentage of people over 60 using the
Internet as we extapolate that survey supplied age to our
results. In fact, since the universe in FL over 60 is so
large, our sample might actually represent a miniscule
penetration of the real universe segment for Florida and
certainly is not applicable to the entire population.
That's why sample selection is critical to Market
Research. Random calling simply doesn't cut it! The
objective is to slice and dice the entire universe, i.e.
the entire population of the U.S. (sorry, rest of the
world ... this is only an example! <G>) into
geo-demographic segments the research firm believs to be
relevant. Then, samples need to be taken from those
segments and measurements taken from those users so
transparently, they virtually forget they're being
measured.
Then, and only then, can we make conclusions by
extrapolating the sample size responses regarding Internet
usage, to the larger universe as a whole and as larger
segments.
The Market Research firm that ultimately accomplishes this
task will have truly contributed to the body of useful
knowledge needed and demanded by Internet marketers. Until
then the total output of these so-called research firms,
regardless of multi-variate regression analysis, etc. etc.
... is pure kaka!
Regards,
Karl Zetmeir, Pres.
Interactive Innovations, Inc.
mailto:kzetmeir_at_gvi.net
========================================================================
------------------------------------------------------------------------
This week's Online Advertising Discussion List sponsor:
24/7 Media
"We chose 24/7 Media because of their proven
success selling online advertising."
-- Howard Lefkowitz, VP Business Development, EarthLink Network
24/7 Media -- Always Serving You
http://www.247media.com/join.html
------------------------------------------------------------------------
========================================================================
Online Advertising Discussion List To Unsubscribe send UNSUBSCRIBE
http://www.o-a.com/ to online-ads-request_at_o-a.com
