NONE: Re: ONLINE-ADS>> Are we surveyed out?
Re: ONLINE-ADS>> Are we surveyed out?
Doug Bates (dbates_at_richmond.infi.net)
Wed, 22 Apr 1998 09:05:06 -0400
George Matyjewicz inquired:
> With all the discussion on stats and surveys on the Net, I
> raise the question whether the Net is surveyed out, or are
> the stats not to be believed?
<snip>
>
> Anybody have any thoughts on this?
Why would anyone want to give up 20 minutes of their time to fill out a
survey that gives them nothing? Consumers are vastly less interested in
marketing surveys than marketers are. A long, long time ago, surveys were
novel. People were flattered to have someone ask for their opinion. Now
it's just work.
What's surprising is that so few marketers have yet to figure out that
surveys give highly deceptive results, because the way most marketers (and
professional market researchers) conduct surveys is just junk science.
A couple of personal examples:
When I worked for National Liberty (now Aegon) I was responsible for the
analysis of the implementation results based on a $250,000 consumer survey
conducted for us by M/A/R/C. The implementation results bore absolutely no
statistical similarity to M/A/R/C's survey predictions. (For you
stat-heads, have you ever seen Chi-Square values over 2,000?....) A
quarter million bucks -- into the toilet!
On another survey there we found that purchase behavior was very different
from people who completed the survey (regardless of what they said) versus
people who didn't complete the survey.
Another (should be famous) example: When SRI International was trying to
validate its original VALS 1 instrument they made a valiant attempt to get
100% response through aggressive (and expensive) follow-up efforts (good
science!). In addition to the VALS instrument, subjects were given some
personality tests, including the now-well-know MBTI. The personality types
of the people who responded readily were very different from those
requiring follow-up. The results fit psychological-type theory. The
compliant, detail-oriented IS-Js readily filled out surveys, whereas the
E--Ps ignored them.
George, IMO your worries are misplaced. It's not the low response rates
that should worry us, it's the mistaken, wide-spread belief that such
surveys will provide anything truthful and of value that should worry us.
Doug Bates
dbates_at_richmond.infi.net
----------------------------------------------------------------------
This week's Online Advertising Discussion List sponsor:
24/7 Media
More brand-name sites. 13 content channels. That's what you get from
the fastest growing online network. Let our media professionals help
you develop a winner plan. 24/7 Media - Always Serving You
http://www.247media.com
-----------------------------------------------------------------------
Online Advertising Discussion List To Unsubscribe send UNSUBSCRIBE
http://www.o-a.com/ to online-ads-request_at_o-a.com
