Re: ONLINE-ADS>> cache counting standards

Matthew Donald (mdonald_at_sofresimr.com)
Tue, 16 Jun 1998 20:36:50 +1000

• Messages sorted by: [ date ][ thread ][ subject ][ author ]
• Next message: Hairong Li: "RE: ONLINE-ADS>> Online marketing research dilemmas"
• Archive Home Page
• Search
• Previous message: Glenn Fleishman: "Re: ONLINE-ADS>> MatchLogic's Statement on TrueCount"
• Maybe in reply to: lperyer_at_doubleclick.net: "ONLINE-ADS>> cache counting standards"
• Next in thread: Bob Stewart: "Re: ONLINE-ADS>> cache counting standards"

> Regarding cache counting technologies such as referenced by
> Ulrich Voss on June 6 regarding Matchlogic. There is a
> company in Australia that adds a supposedly light weight
> attachment to a file that lies dormant until "woken" by the
> end users browser. The reason most companies, using this
> style of technology, don't reference the specifics is that
> the "light weight" code is usually Javascript. I am not sure
> about Matchlogic but certainly its equivalent downunder uses
> a Java applet and script to count activity.

I presume that you are referring to our products, WebMeasure and
AdMeasure.

We use a either a java applet or a cgi (for non-java enabled browsers)
to count page impressions. Java is an excellent choice since it is:

* not impacted by caching (whether in the browser memory or disk
caches, or by network caches)

* invoked whenever a page is viewed (including through the "back"
button)

* the only technique which allows the page view duration to be
determined

* accurately determine visitors to a site without having to use
cookies

* (for banner ads) the only technique which can determine the
banner the user actually viewed rather than the banner which
was requested (which gets around caching of banner ads).

IMR chose browser-based measurement after examining the failings of
server log analysis. It has the following problems:

* Cacheing. It doesn't count requests which were satisfied by
caches. All major internet providers have (very large)
transparent caches in place or under construction

* Visits. Relating a number of requests to determine the number
of visitors, or the audience is not possible soley using
server logs. IP addresses are not useful since many users
are behind firewalls. Mostly, cookies are used to track
users. This has the problem that many users have cookies
turned off on their browsers

* Duration. There is no possible way to measure the duration of
a page view using server logs.

Log analysis is an execellent tool for managing server activity - it is
simply not very good at audience measurement.

As far as I know, IMR developed the idea of browser-based measurement -
at least I can not find references to it which pre-date Jan. 1997.
Maybe we should patent it ;-)

> The problem as we see it at ABC Australia is that many
> firewalls prevent Java due to its ability to get "inside"
> your network and send information out that you would prefer
> secured, but more importantly it can load other unwanted
> software onto your computer. If we all remember the over
> reaction to "cookies" when they first arrived it seems
> strange that it has taken so long to understand the greater
> dangers associated with Java if used maliciously.

What dangers? The architecture of java applets is designed to be
inherently secure. Its sandbox model of security allows the applet to
only play within its own sandbox, and prevents it from accessing other
resources and data contained in its host (the browser).

It is impossible for a java applet to 'load other unwanted software onto
your computer'.

While some firewalls do prevent loading java applets, they are a very
small minority (less than 1% of all users and far less than 1% of all
page views). There is empirical evidence that the blocking of java
applets is decreasing as confidence in java security increases.

It is possible to estimate the effect that firewalls blocking java
applets have on page view measurement by including a test cgi (which
displays a one bit gif image) on a number of pages and comparing these
figures with those measured using the applet.

> Therefore, this method undercounts activity from many secure
> networks. Secondly, browser based activity has yet to be
> manipulated to exclude "back" button requests where the
> viewer is simply using this technique to return to say the
> search engine of his/her choice. I would suggest that
> browser based activity reports are the desired way to go but
> only where clear guidelines are set and this seems some way off yet.

WebMeasure and AdMeasure typically show that 60%-80% more page views
have occurred when compared to log file analysis. I have seen special
cases where there have been upto 250% more page views (static home pages
in high traffic sites).

The fact that page views resulting from "back" button requests is a
deliberate feature. The user has seen these pages, so they should be
counted. If at some time in the future, the industry decided they
should not be counted, it would be trivial to exclude them. Since both
WebMeasure and Admeasure measure the duration of the page view, it would
be simple to filter out pages which were viewed for less than, say, 5
seconds.

I would suggest that the lack of clear guidelines for web measurement
would equally proclude the log file based methodology used by ABC
Australia.

--
____________________________________________________________________
Matthew Donald                                    Ph +61 3 9864 0706
Technical Director                             mdonald_at_sofresimr.com
IMR Worldwide                               http://www.sofresimr.com


========================================================================
------------------------------------------------------------------------
       This week's Online Advertising Discussion List sponsors:
            ICONOCAST, Larry Chase and ChannelSeven.com

 "This continues to be the 'coolest' and most informative e-mail
  newsletter I get." Randy Haldeman, Director of Marketing, TEN-TV
            To subscribe, visit http://www.iconocast.com

			        ---
                      Free sample chapter to
     "Essential Business Tactics for the Net," by Larry Chase
                      http://LarryChase.com

			        ---

 Subscribe to the NEW ChannelSeven Newsletter - News, Ad Case Studies,
 Mrkt Research, Turbo-Ad Resources, Event Info & People-On-The-Move...
 EMAIL: ch7-news-request_at_list.channelseven.com with "SUBSCRIBE" in subj.

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

• Next message: Hairong Li: "RE: ONLINE-ADS>> Online marketing research dilemmas"
• Archive Home Page
• Search
• Previous message: Glenn Fleishman: "Re: ONLINE-ADS>> MatchLogic's Statement on TrueCount"
• Maybe in reply to: lperyer_at_doubleclick.net: "ONLINE-ADS>> cache counting standards"
• Next in thread: Bob Stewart: "Re: ONLINE-ADS>> cache counting standards"

With an archive of more than 14,000 postings, since 1996 the Online Advertising Discussion List has been the Internet's leading forum focused on professional discussion of online advertising and online media buying and selling strategies, results, studies, tools, and media coverage. If you wish to join the discussion list, please use this link to sign up on the home page of the Online Advertising Discussion List.