What I mean is that I've got some with 25 impressions and 4 clicks that are "In Trial", alongside "Normal" terms that have 6,986 impressions and 120 clicks ... alongside "Disabled" terms that have 1,400 impressions and 100 clicks.
I suspect that there is in fact no program automatically checking on the status of terms, despite G's claims about what happens after 1000 impressions, and that, in fact, they have a roomful of people whose only task it is to go through campaigns and assign status to terms that catch their eye for some reason.
I have a similar complaint about Yahoo's people and their "editorial" work when it comes to ad reviews. It's painfully clear that different reviewers use different criteria, and there's no consistency in what they are really checking for. It almost seems that if you are unlucky enough to have a reviewer who just broke up with their boyfriend then few of your ads will pass muster. Some reviewers require the term to be both in the title and in a heading on the page ... others are offended by a part of the term where a search on Y produces many PPC ads with the same part, and still others will let just about anything through.
Thoughts?
How could a program miss or bungle the following examples from the same AdGroup which is over 2 years old?
"term", Normal, 0.5% CTR, 12,851 impressions
"term", Disabled, 0.7% CTR, 4,863 impressions
"term", In Trial, 0.0% CTR, 256 impressions
It just doesn't track. Why is the first one "Normal" when the second one is disabled with a higher CTR and 1/3 of the impressions? And why is the third "In Trial" with 1/4 of the impressions required to trigger an analysis?
I have many terms with 1 impression and no clicks, in fact entire AdGroups that fit that description, and they're not "In Trial", so it's not a "didn't get enough impressions" program.
Also, even if I'm very generous in assuming that only the last 1000 impressions are used in these calculations, as G reps have explained over the years, then how the heck could the first example have survived long enough to get to 12,851 impressions? The fact is that it couldn't ... it would have been disabled a long time ago. There are only 75 clicks. I don't care how you do the math ... that term should have been disabled by any program.
I continue to assert that there is no program that does this, that it's at the whim of people who work at Google following an irregularly-applied set of policies.
Of course, they would be in deep doo-doo if people found out that their people weren't holding everyone to the same rules, so they must claim is a program that's doing it and messing it up.
Does anyone challenge that?
I hope I am not missing the point of your post but here goes...
I think it is done by program - just a very complex one. It has been stated that the program looks not only at your own history but also the history of other advertsiers on the same keywords. I believe it also takes your account performance as a whole, and possibly campaign and adgroup performance into account. This is how you can get some "In trials" with very low impressions.
Also is the campaign only on Google? If you have search network activated then the CTRs being displayed (Google and search network) are different to the CTRs used to calculate keyword status (Google only).
It could be, it could be. Perhaps it's so complex as to be relatively useless as a tool for advertisers.
If such a program operates in such a way as to mimic human inconsistencies, then that's quite a feat.
I like their obfuscation techniques, though. Really admirable.
The CTR displayed in your reports includes expanded matches (from the broad keyword matching option) and ad network distribution*. However, because our performance monitor doesn't consider these statistics when evaluating your account, your keywords may still be disabled even if your reports show that you're above the minimum required CTR.
*As we all know, network and Google search CTRs are reported separately.
My point is that there is little consistency in the way the "performance monitor" handles keywords' status, from an advertiser's optimization perspective. If I'm looking at a term with 0.7% CTR, that should need some optimization in any case. If my ad is getting 2.0%, that should count for something.
It's not helpful to see 2.0% CTR terms get disabled while 0.7% terms live on. The system is confusing enough without these "human-ish" reports. If you make the docs vague enough and throw in some "this program is so complex even we can't explain it to you, our advertisers", you can get away with a lot of slop.
(I still maintain that G's "performance monitor" is named Earl and he lives in their basement. ;)
I'm still convinced that there is no programmatic function that determines whether terms live or die. I'm nearly certain that humans go through the terms and "decide" which to put In Trial or On Hold or to Disable... Does anyone challenge that?
I'm 100% certain that it's an algo, and not humans at work StupidScript. ;)
Probably the best place to look for more details on how it works is in the AdWords Help Center. Here are a bunch of links to explore:
What is the AdWords keyword status column?
[adwords.google.com...]
When does a keyword enter normal status?
[adwords.google.com...]
When does a keyword enter in trial status?
[adwords.google.com...]
When does a keyword enter on hold status?
[adwords.google.com...]
Keyword Evaluation Flow
[adwords.google.com...]
Three Keyword Evaluation Examples
[adwords.google.com...]
AWA
I'm sure this is not a forum for posting details about the term status system, if any forum is. It would be nice if the docs were more explicit, though.
AWA, does your company think it is "dangerous" to lay out the specifics of the algo? It's not like the ranking algo, where your competition could use it against you. Rather, explaining exactly what is expected of your advertisers would seem to be a good business move without competitive repercussions.
The docs may as well say "we cook up a stew of mumbo jumbo and you see the results ... just try to keep your ads online" when a formula would be actually useful. I'm trying to keep track of everything, but there's so little information regarding advertiser expectations on your site that it's mind-numbing to keep guessing at why things are happening all the time.
In a magazine environment, for example, an advertiser is told exactly what to provide and exactly what to expect for their ad dollars. There's no PPC system I have seen that does that, although banner/affiliate systems do a pretty good job of it. PPC engines seem to prefer a "secret recipe" approach to communicating with advertisers that makes it very tough to really maximize a campaign.
I've got some with 25 impressions and 4 clicks that are "In Trial", alongside "Normal" terms that have 6,986 impressions and 120 clicks ... alongside "Disabled" terms that have 1,400 impressions and 100 clicks.
Reading other threads in the forum, I can offer a theory on these results. The KW with 25 impressions does not have enough to measure, so the history from all adwords advertisers' experience with that keyword is used, which is less than 0.5%. Of course, you never see this number. When the impressions get up to 1000(?), then it uses your own CTR to determine if it is to go Normal or Disabled.
6,986 impressions with 120 clicks has a CTR of 1.7%, which is enough to be Normal.
If you have search partners turned on, then your ad may be shown on sites with a search engine like shopping.com or download.com. You still get charged per click for these ads, but they don't get included in the CTR used to make your keyword "Normal", "In Trial", etc. However, the CTR you see on the adgroup page does include the partners. Thus, in your example of 1400 impressions and 120 clicks may have been composed of 400 impressions and 116 clicks from a partner search site, leaving 1000 impressions and 4 clicks, or 0.4% from google.com. Thus, it gets disabled even though it looks like it has a great CTR.
This is just my theory. It would be nice if Google count confirm it.
