AdWords is now Google Ads. Our new name reflects the full range of advertising options we offer across Search, Display, YouTube, and more. Learn more

Ads
2.5K members online now
2.5K members online now
Improve your Google Ads performance and boost your ROI, CTR, and Quality Score
Guide Me
star_border
Reply

Testing Automated Bid Strategies

Visitor ✭ ✭ ✭
# 1
Visitor ✭ ✭ ✭

'Test your automated bid strategies' (https://support.google.com/adwords/answer/6167141?hl=en&ref_topic=6167089)

Informs me of how to setup drafts and experiments to test automated bid strategies. Then, when navigating to 'About campaign drafts and experiments' 

(https://support.google.com/adwords/answer/6318732?hl=en) it states that you cannot test automated bid strategies, and in a test account when you try to set the strategy to this in a draft it fails. 

 

To be further confusing, under features that aren't supported by experiments campaign experiments itself is listed. Is this simply stating that one cannot create an experiment within an experiment? 

 

Is there anyway to test the automated bidding strategies? The conversions are tracked account wide, the strategy can only be turned on and off. 

 

Can I upload multiple conversions, check different single conversions that are to be 'Included in "Conversions"' turn on Target ROAS for that conversion and watch? Then at a later time uncheck that conversion and check another conversion and return to Target ROAS to test how it does with different conversion metrics?

 

There is documentation about 'testing' all over google adwords, but it seems the strategies that I would really want to test are completely ignored by this infrastructure.

 

Any knowledge on how I could implement testing with Target ROAS would be appreciated.

1 Expert replyverified_user

Re: Testing Automated Bid Strategies

Top Contributor
# 2
Top Contributor

@James D

Since you have no control on the setting of automated bids (they are set by a machine algorithm) you technically cannot set a draft experiment:  you simply cannot set a value to the bid.

Generally, when  you have no control on the variables (e.g the bids) I don't see how you can test. For a test to be successful you must control the variable (the bid) not the result (i.e. the conversion)

 

Moshe, AdWords Top Contributor , Twitter | Linkedin | Community Profile | Ad-Globe
Did you find any helpful responses or answers to your query? If yes, please mark it as the ‘Best Answer’

Re: Testing Automated Bid Strategies

[ Edited ]
Visitor ✭ ✭ ✭
# 3
Visitor ✭ ✭ ✭

@MosheTLV Thanks for your response! Hopefully I can make it more clear what I am trying to achieve. 

 

My variables in this case would be the automating bidding itself and the conversion value that it is bidding towards.

 

I.e. 

  • Manual CPC vs. Automated Bidding towards Conversion Metric 1
  • Manual CPC vs. Automated Bidding towards Conversion Metric 2
  • Automated Bidding towards Conversion Metric 1 vs. Automated Bidding towards Conversion Metric 2

    Is automated bidding better than manual CPC? I can't test this. I can only test this in time so I can't discern if automated bidding is better or if it was just better over a period of time.

    Next, I cannot change the variable that I am feeding to the automated bidding strategy. Maybe it's better for my client to track towards phone calls for some campaigns and towards offline conversions in another campaign.

    The experiment infrastructure is exactly what I want for this. I want for a single campaign to drive 50% of traffic towards bids set by the automated bidding strategy and 50% traffic towards manual bids.

    I could then A/B compare them after the experiment.

    Lets step back, if I were to upload 5 different conversions into my account. Could I binary select a single conversion to be "included in conversions" and all my automated bidding strategies would bid towards that conversion?

Re: Testing Automated Bid Strategies

Top Contributor
# 4
Top Contributor

@James D;

The issue here is statistics.  Not AdWords. In statistics when testing  a variable you must have the ability to set its value. Only in manual bidding you can set the value yourself. In auto-bidding you cannot set a value - it is  is calculated by the AdWords algorithm, hence it cannot participate in a statistical test.

 

Moshe, AdWords Top Contributor , Twitter | Linkedin | Community Profile | Ad-Globe
Did you find any helpful responses or answers to your query? If yes, please mark it as the ‘Best Answer’

Re: Testing Automated Bid Strategies

Visitor ✭ ✭ ✭
# 5
Visitor ✭ ✭ ✭
Auto-bidding requires that you set the variable, the conversion metric, to calculate bids. It is clear that setting different variables will provide different results.

Re: Testing Automated Bid Strategies

Top Contributor
# 6
Top Contributor

Partially correct; The conversion value is the result / the outcome, and  not the variable. Since AdWords in its core is still a CPC  bidding system - the variable (to be tested)  is the bid. And this you can set only in manual bidding. All bidding strategies in AdWords are "converted" back to a bid. When you set the conversion value you lose the  "degree of freedom"  to manually set the bid.

 

Moshe, AdWords Top Contributor , Twitter | Linkedin | Community Profile | Ad-Globe
Did you find any helpful responses or answers to your query? If yes, please mark it as the ‘Best Answer’