| Home >
Traffick Articles
> Internet Marketing
By Andrew Goodman, 7/25/2005
Recently, Google introduced what they’re calling “simplified
keyword states” and “quality-based minimums”
as a way of injecting more flexibility into its AdWords online
advertising platform. Advertisers will no longer have to grapple
with the former clickthrough rate (CTR) “cutoff”
of “0.5% normalized for ad position.” Now, there
is no cutoff, and low CTR’s will not cause a campaign
to be slowed. However, for low-CTR keywords, advertisers may
be forced to bid higher than the current minimum bid of .05
in order to keep their keywords active.
On the surface, that’s a fundamental change, because
that 0.5% CTR threshold has been a much-touted feature of
the AdWords program from Day One. Now, with a flourish and
a promise that a new cutoff-free regime will be live in “a
matter of weeks, not months,” it’s gone. Google
cared so much about CTR’s -- or more to the point, their
image of being obsessed with relevance and search quality
-- that they once claimed to be doubling the threshold to
1.0% “under certain conditions for some keywords.”
(That initiative quietly fizzled out.)
In spite of the appearance of dramatic change, the new approach
might be just another way of arriving at the same answer using
a different set of calculations. Leave it to mathematically-inclined
Googlers to figure out ways of expressing essentially the
same formula in different ways.
I recently spoke with Richard Holden, director of product
management at Google, and he confirmed a few points. The main
purpose behind the changes is to help advertisers who found
the old keyword states (such as ‘in trial,’ ‘on
hold,’ and the former ‘at risk,’ etc.) “confusing,”
and who found it “frustrating when they couldn’t
keep keywords active.”
What’s a ‘Quality Score’?
Here’s where it gets slippery, though. One of Google’s
bullet points says that “Ad Rank will continue to
be based on maximum CPC and quality (now called the Quality
Score).” [Emphasis added.] However, most advertisers
are still under the impression that ad rank is based on CPC
multiplied by CTR. Now, “quality” is judged on
other factors as well, including the historical performance
(whatever this means) of keywords as they trigger specific
ads, or ads with certain characteristics (Google doesn’t
say, exactly). To some advertisers, this smells like a covert
way of increasing prices. To someone worried about the continued
legitimacy of Google Search as a usable research tool, though,
this selective price increase is welcome, as a way of forcing
bad ads off the page if they really do come across as clutter
to the end user. Advertisers who want to clutter up the page
with “low quality” (as Google defines it) messages
will be forced to pay more.
Sound surprising? It isn’t if you have followed Google’s
history.
It feels like Google has now succumbed nearly fully to the
temptation to view ads not as ads at all, but as extensions
of organic search. Just as search results are constantly tested
for quality, so now will ads. Similar to Google’s formulas
for ranking organic results, the algorithm for ad placement
won’t be transparent. The difference: as an advertiser,
you’re paying to participate in this scheme, and your
ads might be judged in ways you don’t quite comprehend.
If you aren’t attuned to the concept of “search
quality,” your campaigns will start to cost more than
they should.
WebmasterWorld.com forum moderator eWhisper spotted the emphasis
on Quality Scores in Google’s documentation, and noted:
“I hope this will be a published score.” Don’t
bet on it.
The “quality score” of a keyword appears to be
related to the predictive system Google brought in some time
ago that would force some keywords to go “in trial”
(slowed delivery, pending eventual disabling or activation)
while a particular advertiser’s performance on a keyword
was measured. The “quality score” could include
the historical performance of keywords in various industries
against different kinds of ad creative. So in that regard,
Google’s new system -- like the previous confusing iteration
-- is somewhat predictive in nature, at least until a particular
advertiser has developed a strong track record on a keyword.
Advertisers might see a high “minimum” that they’ll
need to bid to be shown on a keyword with a low “quality
score” that results from the historical performance
of other people’s ads. Eventually, Holden maintains,
an advertiser could find themselves doing better than other
advertisers. In that case, their higher quality score would
eventually allow them to bid less without having the keyword
declared inactive.
It would be great if we could hear more detail about how
this works, exactly, but Google typically doesn’t disclose
the exact details of its often-complex formulas. The makeup
of the performance history in an advertiser’s keyword
account, as it relates to “quality scores” that
will decide whether a high minimum bid will be enforced, is
still murky. We don’t know how much protection you build
up against blips in performance if you build a long track
record of great performance on a given keyword. If anyone
not wearing a Google ID badge tells you they know the exact
formula, get in your car and drive away as quickly as possible.
Even junior Google employees (read: anyone who has not literally
overseen or designed the AdWords platform) have been known
to trail off in their attempts to describe the rules for ad
quality, using phrases such as “…or whatever.”
We do know that under the new system, there is no floor for
bids. If advertisers have a sufficiently high quality score
on a given keyword in their account, they could bid as low
as one cent (or the local equivalent). What a fun way Google
chose to announce a summer price drop! (They sort of slipped
it in there as an extra “oh, by the way” tidbit.
Pretty cleverly done, given that something like this a couple
of years ago would have generated headlines reading “Google
Drops Prices!” Of course, when you both raise and drop
prices at the same time, no one knows which headline to write,
so the stock price is safe.)
This change -- at the end of the day, probably an incremental
one -- does offer an opportunity to look a bit more closely
at this by-now-longstanding “holy grail” of paid
search auctions: is CTR really so important? Is Google admitting
that it isn’t quite as important as they once claimed?
In short, I think CTR is as important as ever in the new
regime. The change is more one of smoothing advertisers’
experience in dealing with essentially the same challenges.
1 | 2
| 3 | 4
| Next
|
