AdWords
1.7K members online now
1.7K members online now
Dive into advanced features like Remarketing, Flexible Bid Strategies, AdWords Editor, and AdWords Scripts
Guide Me
star_border
Reply

Ad group structures - multiple tests

Visitor ✭ ✭ ✭
# 1
Visitor ✭ ✭ ✭

After some suggestion from the community I decided to run as many tests as it makes sense. So far I've set up 3 tests with the max. CPC accross all the tests the same for: exact 100%, phase 70% and BMM 50%.

The campaigns are as follow:

 

1.
- Ad group: Phase > Keywords: [KW 1], [KW 2], [KW 3], [KW 4]
Negative: “KW 1”, “KW 2”, “KW 3”, “KW 4”
- Ad group: Exact > Keywords: “KW 1”, “KW 2”, “KW 3”, “KW 4”
Negative: [KW 1], [KW 2], [KW 3], [KW 4]
- Ad group: BMM > Keywords: +KW 1, +KW 2, +KW 3, +KW 4
Negative: [KW 1], [KW 2], [KW 3], [KW 4]
Are my negatives right? Google says that they are all wrong and after running “Keyword Diagnose” test says: “Excluded” beside every single keyword!


2.
- Ad group: KW 1 > Keywords: “KW1”, [KW1], +KW1
- Ad group: KW 2 > Keywords: “KW2”, [KW2], +KW2
- Ad group: KW 3 > Keywords: “KW3”, [KW3], +KW3
- Ad group: KW 4 > Keywords: “KW4”, [KW4], +KW4


3.
- Ad group: KW 1 exact & phase > Keywords: [KW1], “KW1”
- Ad group: KW 1 BMM > Keywords: +KW1
- Ad group: KW 2 exact & phase > Keywords: [KW2], “KW2”
- Ad group: KW 2 BMM > Keywords: +KW2
- Ad group: KW 3 exact & phase > Keywords: [KW3], “KW3”
- Ad group: KW 3 BMM > Keywords: +KW3
Here also got a suggestion to put exact match into one group and phase with BMM into another – not sure what’s difference that’s going to make?


Two questions:
1. Do you have other structure suggestions that I could test or comments how to improve the above to get the best results?
2. In the campaign no 1 why does Google not agree with my negatives?

 

Thanks,

Nata

1 Expert replyverified_user

Re: Ad group structures - multiple tests

[ Edited ]
Top Contributor Alumni
# 2
Top Contributor Alumni

1.A The first example is labeled "Phrase" but your keywords are coded for Exact match. Also, if you're using Exact match keywords, you don't need those negative phrase match ones. Exact is Exact (unless you're opted in to "close variants") and the broad/phrase match search queries won't be matched to these Exact match keywords.

 

1.B The second example is labeled "Exact" but your keywords are coded for Phrase match. Blocking the Exact match traffic, to drive those searches to the Exact match keywords is a valid strategy. 

 

1.C The third example is labeled as Modified Broad match. Blocking the Exact match traffic, to drive those searches to the Exact match keywords is a valid strategy. You might also want to block your chosen Phrase match keyword phrases, to make certain that traffic is driven to those keywords. (I very often find that BMM and Phrase match keywords "compete" with each other.)

 

I see no problem with the Examples in 2 or 3. 

 

I assume these second two are using different keywords than 1 since the software won't serve all of these--it will "choose" one and serve the vast bulk of the traffic to it.

 

As for your closing questions, I'm not even sure why you're running this test, so I can't really comment on whether or not it's a good idea. (To be honest, it sounds like a waste of time to me. If it were me, I'd be spending this time working on ad messaging. I normally divide Ad Groups into closely "themed" keywords and then split out any keywords where performance varies by more than X amount and try custom ad messaging for those keywords. I find that has a much bigger impact on results than playing with mix-and-match type Ad Groups.)

 

The software can often have difficulties "understanding" Ad Group level negative keywords. I would suggest that you closely double-check to insure that  you have the right match types in the right Ad Groups, then give it a day or so to see if any actual data accumulates.

 


Theresa
Google AdWords Top Contributor
*~*~*~*~*~*~*~*

Re: Ad group structures - multiple tests

Badged Google Partner
# 3
Badged Google Partner

I agree with Theresa as is usually (always) the case. Smiley Happy

 

Where do these KW terms come from? Are they initial seed, or suggested terms from the KW tool, or underlying search query terms that you know help you acheive your advertising goals? 

 

No matter what kind of account, campaign, or adgroup structure you use, if that structure is not helping you acheive your goals - Then that's a lot of time and effort spent which could perhaps be more wisely spent in other areas. Smiley Happy

Tom

Re: Ad group structures - multiple tests

Visitor ✭ ✭ ✭
# 4
Visitor ✭ ✭ ✭

Hi Theresa,

 

Thank you for your toughts and suggestions.

The reason behind running these tests is to see what works best in my account as we all know there is no golden rule for every account.

The tests won't be that time consuming as I took only few keywords with the highest search volumes and created these experiments. I'm not running them at the same time, I was thinking to collect a reasonable sample to compare how each variation performs.

As you said messaging is VERY important, but I believe that I can't create a great account without good roots. So in order for the right ads (messaging) to appear for the most relevant keyword I think that the account structure has to be done right firstly.

 

Thanks

Re: Ad group structures - multiple tests

Top Contributor Alumni
# 5
Top Contributor Alumni

I understand and don't deny that all aspects of a campaign are important--and that organization can be very important.

 

If you plan to run these sequentially, remember to watch for seasonality. (By which I mean that different industries not only perform differently in different seasons of the year, but some of them perform very differently in different days of the week or weeks of the month. You may need to choose your start & stop dates very carefully in order to make sure that your different experiments are running in an "apples-to-apples" environment.

 

Come back and tell us what you're learning, okay?

 

Good luck!


Theresa
Google AdWords Top Contributor
*~*~*~*~*~*~*~*