online targeting

From: <TLee_at_webcmo.com>
Date: Wed, 13 Jan 1999 13:30:20 -0600 (CST)

Online targeting is a hot topic. There are many numbers
showing the benefits of such targeted banners (usually
higher click through rates). But is it worthwhile to pay for
such targeted banners?

The basic logic is that high quality hits is better than low
quality hits if it does not bring you any additional cost.
However, if gaining higher quality hits involves additional
cost, you need to determine whether it is optimal to gain
the quality of the hits. Or in other words, where is your
optimal quality of the hits?

I would like to use an example to do some calculation.

Infoseek has developed a targeting technique-Ultramatch. It
uses state of the art statistical technique-Neural
Network-to target online audience one to one on its web
site. There is a report written by Peter Rip, VP and GM of
Infoseek Network.
(http://info.infoseek.com/doc/ultramatch/ultramatch.html)
This is one of very few real research reports I have seen on
the Internet.

Peter Rip reported: "The results of the experiment with 3%
of the traffic demonstrate a click-through rate some 20-25%
higher than achieved through Infoseek's standard targeting."
Infoseek's standard targeting is defined as: "such as
purchased keywords or topics or random general banner
rotation." Among them, "Keyword targeting accounted for
approximately one-fifth of the ads on Infoseek."

So let's assume the average CTR of Infoseek's non keyword
targeted banners is 1% (This number does not affect the
result of our calculation because it will be canceled out on
both sides of the equation later.) and the CTR of keyword
targeting is N times of the average CTR of non keyword
targeted banners, then the average CTR is (0.8x1%+0.2xNx1%)
and the average CTR of Ultramatch is 1.25x(0.8x1%+0.2xNx1%).
To make the average CTR of Ultramatch be higher than the
average CTR of the keyword targeting:

1.25x(0.8x1%+0.2xNx1%) > Nx1%

N must be smaller than 1.33. It means that if the CTR of
keyword targeting is 1.33 times higher than non keyword
targeted banners, Ultralmatch is less effective than keyword
targeting.

(This is the only data available for us to understand the
effectiveness of using advanced statistical technique to
target customers on the Internet. If you happen to know
other such data, please let me know.)

As I said before, targeting by keyword and content is more
cost efficient than other targeting method. Our research on
advertisers' preferences has also shown that in general,
advertisers prefer content and keyword targeting to other
targeting methods.
(http://www.webcmo.com/report/uap/report2.htm) However, even
if you go for content or keyword targeting, you still need
to calculate: is the gain of the quality of hits able to
cover the additional cost?

Online targeting is an old marketing issue but needs new
theoretical study. We have been working on this issue for
months and trying to find a way to make online targeting
more effective and less costly.

Best,

Tim Lee
Director of Research
http://www.WebCMO.com
A site dedicated to web marketing research.
Come and see our risk free & satisfaction guaranteed web survey service.


========================================================================
------------------------------------------------------------------------
       This week's Online Advertising Discussion List sponsors:
         ValueClick, Enterprise Interactive and eGroups.com

 ValueClick has cost-effective ad space available for *Smarter Buyers*
 -> Our Cost-per-Click model means risk-free Reach to over 10k websites!
 ValueClick Ad Network | 805.965.0543 | http://www.valueclick.com

                                ---

   CUSTOM CONTENT FOR YOUR SITE - articles, features, daily tips by
   Enterprise Interactive's team of writers, editors and marketers.
   peter@enterprise1.com, http://www.enterprise1.com, 203.838.9383

                                ---

 eGroups.com hosts e-mail discussion groups with over 2 million users.
 And you can precisely target them with e-mail advertising that really
 works. To reach opt-in, interested users, contact sales_at_egroups.com

------------------------------------------------------------------------
========================================================================
Online Advertising Discussion List To Unsubscribe send UNSUBSCRIBE
http://www.o-a.com/ to online-ads-request@o-a.com



Received on Wed Jan 13 1999 - 15:03:24 CST