In-Reply-To: <199607012301.TAA16002@social.com>
Message-ID: 
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII



	Rex Ballard - Director of Electronic Distribution
	Standard & Poor's/McGraw-Hill
	Opinions expressed do not necessarily reflect
	the Management of the McGraw-Hill Companies.
	http://cnj.digex.net/~rballard


On Mon, 1 Jul 1996, Web post wrote:

> My understanding,from an interview I did with Paul Grand, 
> chairman and founder of third-party Web measurement firm 
> NetCount, is that to really measure "click through" you 
> need software on both sites, and that's what NetCount sets 
> up. Then you know everything -- especially, you know if the 
> user that clicked on the ad banner actually made it over to 
> the intended site (i.e., a transfer made, not merely requested). 

There are really two things that matter.  One is, did the advertizer's
page get read.  The other is, who made the referral to that advertizer's
page.

Remember, the goal of the ad is to increase demand and ultimately produce
qualified customers (people who will actually buy the product being
advertized).

If FORD puts an ad on a Television sports event, it is unlikely that he
will get the customer to go out to his dealer and buy the Bronco with one
ad for the hocky game.  On the other hand, the Miller Beer ad may produce
immediate results (dad gets up and gets another can of brew).

If I'm shopping for a truck, wouldn't it make sense to get the Cosumer's
reports, and perhaps some web content that rates trucks?  I might use
Lycos to search for "Trucks".   Now, if I happen to see the link to a 
an ad telling me about the Bronco, especially if the link directly related
to the "Hot Point" I've been examining (like repair history), I might be
willing to explore the intire Ford site more thoroughly.  The result is a
highly qualified customer.  I might even be willing to click the box that
will have a salesman drive to my door in the model I have chosen, and
take it for a test drive.  WOW :-).

> This is obviously important to an advertiser who pays the 
> publisher only on the basis of how many pairs of eyeballs 
> actually landed on and saw the advertiser's site. As to the 
> other questions, the only way to "count" (i.e., as opposed 
> to "estimate") unique visitors is by registering and 
> authenticating every one of them.

What matters is that qualified customers hit an appropriate entry point.
If I put an ad in the SuperBowl, for "Technogeek stuff", and don't produce
even one customer...I've flushed $1 Million down the proverbial toilet.

If I only get 10 hits/day, but 3 of them lead to sales of $200,000 homes,
I've just made $18,000 in commissions - PER DAY).  WHO CARES that I
haven't been visited by 100,000 little old ladies who are shopping for a
nursing home, and 200,000 prebubescent boys hoping for a gawk at Kathy
Irish.

> Some services and counting 
> software try to measure "visits," determined by looking at an
> IP addressess, the occurrence of an IP address in a log file.

An IP address can be very misleading.  My browser comes from behind a
fire-wall that serves 5,000 people.  Most of them use their web browsers
to gather information required to make corporate purchasing decisions.
Often, a supplier who would otherwise have been ignored is identified as a
result of either linkage to a highly relevant page, or as a result of a
high score from a search engine.

> But a "visit" does not equal a "visitor," because if the IP 
> address disappears for a 30-minute period and then returns, the
> Web site calls it a new "visit."  One person can create hundreds
> of "visits" in a day on the same Web site.

Look at the flip side.  One "visit" can result in a tour of hundreds of
privately managed pages.  Because the advertizer can just put up his site,
your link to his home page (or better yet a targetted relevant page) can
result in a transfer that becomes a complex qualification "Dialogue" with
the customer.

If your site is generating $200,000 /week in housing sales commissions,
and $50,000 / week in car sales commissions, who cares if you are only
"producing" 200 hits/week.

You will know you have become a gold mine when you threaten to cancel the
referring article.

> Then consider all the
> spider search engines crawling through Web sites to index the
> Web.

Yet another form of gold.  That one hit, which results in the advertizer
showing up in search-engine queries, becomes a source of substantial
revenue for the advertizer.  Again, can the advertizer afford to lose your
linkage?

> So you could say that counting unique hosts does NOT give 
> a close estimate of visitors.

The critical information the advertizer needs to know is who referred the
user to that page.  The "Apache" server supports a "referred by" function
which the advertizer can use to acknowledge you as the source.

My RESUME (see http://cnj.digex.net/~rballard) contains links to previous
employers along with AD blurbs.  One company (softronics) is actually
getting referrals because relevant information shows up on my content.

I'm afraid to take it off (I have a new job now, thanks in part to the
internet/web site) because I don't want them to lose that referral.

> If you'd like, I'd be happy to 
> e-mail you the two interviews I did with Paul Grand, which 
> were intended to clear up a lot of the issues and confusion
> in Web measurement.
> 
> Alan Coon
> Editor, Interactive PR (newsletter)
> editor@interactivepr.com   
> 
> +---------------------------------------------------------------------------+
> This message was posted to ONLINE-NEWS. http://www.planetarynews.com/o-n.html
> 


From rballard@cnj.digex.net Fri Jul  5 16:14:11 1996
Status: O
X-Status: 
Newsgroups: comp.os.linux.advocacy,comp.os.os2.advocacy,comp.sys.mac.advocacy