Re: Ad Click EMetrics - Filtering out spiders and robots

From: Cliff Kurtzman <moderator_at_o-a.com>
Date: Wed 10 May 2006 15:10:26 -0500

James Raposa wrote:

>In response to "Ad Click EMetrics - Filtering out spiders and robots":
>Maybe it used to be possible to filter out possible fraudulent
>clicks by simply
>looking for visits coming from the same IP address over a short period of
>time, but that's not something that can be done with any degree of accuracy
>today. And here's why. I do consulting work at a medium sized university
>in New England. Their network is set up in such a way that EVERY web
>surfer eminating from behind their equipment appears as THE SAME IP
>address. Their router takes care of parceling out the information to its
>proper recipient upon return (via someting called a MAC address.)
>

I agree with James to some extent. He is correct that with some
networks (such as ours at ADASTRO) every click from every user will
come from the same IP address even if originating with a different
computer and human being. There are also some networks out there
(like AOL's) for which the originating IP will change with every
click and file accessed. Both of these factors complicate accurate
reporting, particularly by automated tools.

Yet in a more practical sense, one can sometimes deal with these issues
effectively. For a web site with on the order of 100,000 visitors a
month, unless something very unusual is going on, the chance that two
people from the same company or same university will be following the
same click path over the same time interval is quite small in the
greater scheme of things, and if a few true unique human clicks are
eliminated in such a filtering process it seems to me an acceptable
loss in the interest of providing fair reporting.

With ads in an email newsletter that is delivered to a lot of
recipients simultaneously, the chances of this happening are greater,
but with email distribution using current technology it is fairly
easy to tag the hyperlinks with an extension that uniquely identifies
the user that received the email. Unless it is a situation where the
email being sent to the recipient is in turn being mass distributed
to a second distribution list within the receiving organization, it
is easy to identify unique user clicks on email ads, even if all
users appear to have the same IP address.


--Cliff

Clifford R. Kurtzman
CEO | Moderator
ADASTRO Incorporated | The Online Advertising Discussion List
http://www.adastro.com | http://www.o-a.com
(281) 480-6300






Received on Wed May 10 2006 - 15:10:26 CDT