Welcome to the Official AdWords Community

Reply

Analyzing automatic placements

| ‎24-04-2012 01:43 AM
0

Hi,

 

I am running a campaign that focuses on both automatic and manual placements. The placements are doing fine, the big issue is that I am finding difficult to manage and analyze automatic placements.

 

Is there a way to understand which keywords and what site content triggered a particular placement?

 

The reason I am asking this is that I have "mac" "os x" "ipad" 'iphone" in my negative keywords lists (this is for automatic placements) and I ams till getting click for site such as:

 

http://osxdaily.com/

 

The strange thing is that this site has no content/text related to "Russian" or "Grammar" or "Learning" (my company sells a software product for learning Russian grammar). Why did it get a placed in my campaign. I looked through the keyword list for the ad group that got placed on osxdaily and all of the keywords have the word Russian in them (majority using phrase/exact match, all the broad use the "+russian" modifier).

 

Analytics labels the osxdaily match type as "Content network ads", so I am assuming that our ad was placed on their site due to relevant content and not due to vertical user tracking (where the ads follow a user who could be interested in "learning Russian"

 

I've messed around with the colmn and filter function in the "automatical placements" section in the adwords control panel, I can't figure out how to make adwords display which keywoard triggered a particular placement. Same thing with Analytics, it just says "Content" "Keyword" or some other generic label.

 

Could anyone help me understand how I can get rid of irrelevant placements? I already have a pretty detailed negtive keyword list for my DN campaign and I've blocked domainsponser/parked pages and so on. I manually block placements on websites that have several clicks and a 100% bounce rate.

 

Thanks,

 

Vlad 

Message 1 of 8 (550 Views)
Please use plain text.
Full Reply Editor

Re: Analyzing automatic placements

Reply to grammatica
| ‎24-04-2012 03:23 AM
0

Hi Vlad,

 

To answer your first question - "Is there a way to understand which keywords and what site content triggered a particular placement?"

 

We can find the actual page where the ad was placed as follows:

1.  Please go to "Networks" tab

2.  Click on show details" next to "Automatic Placements"

3.  Select the check box for the placement(s) you want to see the actual URL

4.  Click on "See URL list..." button

 

This would show the URL(s) where the ads have been placed.  Now to find the keyword which triggered the ad, you need to do that manually by analyzing the content and the keyword used along with the tye of matching happened.

 

Reg. your next question about "osxdaily.com", we can't be sure of.  Since I was able to find certain pages in osxdaily.com related to grammer and language settings,  I would recommend you to look at the actual URL of the page where the ad was placed.

 

Apart from adding negative keywords, it would worth to exclude placements, which you find as bad.  This has to be done manually, since you need to make the call on its effectiveness.

 

Let me know if you have any questions.

 

Thanks,

Nimesh.

Message 2 of 8 (542 Views)
Please use plain text.
Full Reply Editor

Re: Analyzing automatic placements

Reply to Nimesh1984
| ‎24-04-2012 03:50 AM
0

Hi Nimesh,

 

Thank for the info. I typically look through the placement domain/URL through analytics, but it's good to learn how to do this through AdWords itself. 

 

However, this method still does not provide any information about which keyword/content triggeres a contextual ad placement. I already use the manual method - the problem is that I often find URL that have no text/content related to my keywords. This is really annoying. 

 

I've done some digging and it seems that this feature is simply not available in analytics. However, it seems that Google is planning to implement this functionality as part of the current (April 2012) mini update to the AdWord control panel.  

 

Regards,

 

Vlad

Message 3 of 8 (537 Views)
Please use plain text.
Full Reply Editor

Re: Analyzing automatic placements

[ Edited ]
Reply to grammatica
| ‎24-04-2012 04:43 AM
Last edited 24-04-2012 04:44 AM by Kim_Clink

grammatica,

Did you know that you can exclude topics and interest categories?

See the networks or display network tab (depends on which one your user interface shows) and go to the bottom of the page and look for exclusions.

Also have a look at the site and category exclusion section and remove yourself from error pages, parked domains, and sites that may have violent or sexually related content...if that makes sense of course, it sounds to me like it would make sense to exclude certain categories in your case :smileyhappy:

 

Here is a link with more about excluding topics http://support.google.com/adwords/bin/answer.py?hl=en&answer=178148

 

More about site and category exclusions: http://support.google.com/adwords/bin/topic.py?hl=en&topic=1713963&from=15911&rd=1 

Kim Clink
Message 4 of 8 (528 Views)
Please use plain text.
Full Reply Editor

Re: Analyzing automatic placements

Reply to Kim_Clink
| ‎24-04-2012 05:09 AM
0

Hi Kim,

 

Thanks for the comment. I initially had a lot of problems with low quality placements, but I've fixed most of the quality problems using the exclude category feature (just as you described in your post) and some other tricks to block unsuitable providers.

 

My current problem is that my ads are being placed on sites that seem to have little to do with my keywords. The quality of the sites is okay.

 

Regards,

 

Vlad 

Message 5 of 8 (523 Views)
Please use plain text.
Full Reply Editor

Re:Contextual Targeting on the Dispaly Network

Reply to grammatica
| ‎24-04-2012 06:13 AM
0

Hello Vlad;

As far as your question that ".this method still does not provide any information about which keyword/content triggers a contextual ad placement.,".

As you correctly pointed out, in contextual targeting all keywords work collectively as a theme, to find the most relevent webpages. With the new Display Network tab in place, you can give a different "weight" to of some keywords in the overall "contextual grading": some KWs would contribute more than others to the evaluation of relevancy, But, still, they all work collectively, as a theme.

 

With the new display tab in place, you can see statistics on individual keywords. Since it is new, I am not quite sure yet, how this individual keywords statistics derives from contextual targeting (of all keywords as a theme).

 

-Moshe

Message 6 of 8 (515 Views)
Please use plain text.
Full Reply Editor

Re: Re:Contextual Targeting on the Dispaly Network

Reply to MosheTLV
| ‎24-04-2012 08:40 AM
0

Hi Moshe,

 

That's rather interesting. Is this weight feature still being rolled out? I can't seem to find it in my AdWords control panel and I can't find any documentation or web discussions about choosing weight for keywords in DN only compaign.

 

Regards,

 

Vlad

Message 7 of 8 (507 Views)
Please use plain text.
Full Reply Editor

Re: Re:Contextual Targeting on the Dispaly Network

[ Edited ]
Reply to grammatica
| ‎24-04-2012 09:43 AM
0
Last edited 24-04-2012 01:24 PM by MosheTLV

Vlad;

Here is the official announcement in AdWords blog. No clear indication how contextual targeting and specific keywords work together.

http://adwords.blogspot.com/2012/03/better-way-to-buy-display.html

 

On another discussion (that may interest you) , I suggested that a Google DN specialist will join the conversation to "put some light" on the new tab... Hopefully Zee will find someone.

 

-Moshe

__________________________________________________________________________

grammatica wrote:

 

Hi Moshe,

 

That's rather interesting. Is this weight feature still being rolled out? I can't seem to find it in my AdWords control panel and I can't find any documentation or web discussions about choosing weight for keywords in DN only compaign.

 

Regards,

 

Vlad


 

Message 8 of 8 (502 Views)
Please use plain text.
Full Reply Editor

Did this Answer Your Question? If not post your question here