<<< Date Index >>>     <<< Thread Index >>>

[IP] more on Click Fraud, Google, and Telepathy





Begin forwarded message:

From: Brad Templeton <btm@xxxxxxxxxxxxxx>
Date: October 22, 2006 10:39:30 PM EDT
To: David Farber <dave@xxxxxxxxxx>
Cc: ip@xxxxxxxxxxxxxx
Subject: Re: [IP] Click Fraud, Google, and Telepathy

On Sun, Oct 22, 2006 at 07:18:36PM -0400, David Farber wrote:
Google says that less than 10% of ad clicks through their system are
illegitimate.  Taking that at face value, it can still add up to an
awful lot of fraud given the total amount of clicking going on out
on the Web these days.  And of course, when someone says that fraud
is at such and such a level, what they really mean is *detected*
fraud.  Undetected fraud, like other proverbial "perfect crimes,"
doesn't show up in your stats since you don't know about it in the
first place.


Google already has the answer underway -- pay for performance
advertising.  And only they, because of their dominant market
position, can pull it off.

Advertising in general, and web advertising in particular has always
had two forces pulling at it.

Publishers (ad sellers including Google) really care about only one
thing -- total ad revenue, and the normal driver of that is CPM -- revenue
per thousand pageviews.    If publishers sell pay per click or pay
for performance it's really just a means to the real end -- CPM.

That's because as a publisher you're going to put up 1000 pageviews
and you want the most money for them.  You care a bit about how you
get it of course -- dancing flash ads may scare away users -- but
you really have no desire to be in partnership with the advertiser's
business.   With pay per click, if they don't put up an appealing ad,
you don't make money.  With Pay for performance, if nobody buys their
product, you don't make money.   This is not something you have the
time to worry about -- how good a company they are or how good they
are at composing ads.

Google is #1 in part because they figured an answer to this.  They
sold PPC, but they really demanded what all publishers really want,
namely CPM.  If your ad didn't get good clickthrough (and thus generate
sufficient CPM) they had the audacity to actually refuse your ad.

In the days when most other publishers were being total whores, taking
any ad with a hope of payment, even if it did pop-ups, music and
dancing, Google was strong enough to say, "No, we only take ads that
people actually click on.  And we only take them as plain text."
They said to their real customers (the advertisers) that they were
not always right.   In many ways I think that's what "don't be evil"
actually meant.

Going against the grain like this made them billions.

CPC however, runs the risk of click fraud as described.  But
Google, as it encroaches on monopoly, and pull this off again.
They can sell pay for performance while really, internally, caring
about CPM.  They can say, "You can run pay for performance, but if
it doesn't perform, you not only don't get sales, we refuse your ad
in favour of some other ads that do better."  Only by being large
can they see the wisdom in refusing business, because they have other
business waiting in the wings.

There can't really be clickfraud in pay for performance, not as we
think of it.

All this thinking does leave certain types of ads in the lurch, however.
Those for whom "performance" is hard to define.   This includes
"image" ads which are just promoting brand awareness with no immediate
plan for a sale (certainly not a sale on the web.)  It also includes
products with long sales cycles, political ads and the like.

For them, the answer may be to go back to CPM. CPM got abandoned because
if you're an advertiser, and you can get pay per click, or pay for
performance, you would much rather have those.  But if you can't have
them, you have to do your homework and figure out which ads really
work when you're paying CPM.  Of course, you can still see the clicks
and the followthrough as much as the web allows.  Fraud is possible on
CPM ads, though it's harder.




-------------------------------------
You are subscribed as roessler@xxxxxxxxxxxxxxxxxx
To manage your subscription, go to
 http://v2.listbox.com/member/?listname=ip

Archives at: http://www.interesting-people.org/archives/interesting-people/