Google Ads | Experiments

Recourses:

Videos:

https://www.youtube.com/watch?v=w3iDO2d09qw

 Articles:

https://support.google.com/google-ads/answer/6318732?hl=en

https://www.searchscientists.com/google-ads-experiments/

https://www.seerinteractive.com/blog/google-campaign-experiments/

  

How to set up an experiment:

To create an experiment, it is a 2 part process. First you need to create a draft of the campaign you want to run an experiment on.

Part 1 The Draft

Once you have created a draft of the campaign you have essentially made a carbon copy so you will now make the edits to the draft.

  • If you want to test out a different landing page for ads you would go into the ads and do a find and replace for all the URL's. There will be a checkmark next to all the ads that have successful had the URL changed.

  • If you want to test a bidding strategy you go into settings and change the setting.

  • You are able to test all sorts of things but these have been the main 2 used recently.

 

Part 2 The Experiment:

Now that you have made your changes in your Draft you now will want to start the experiment.

  • Go to drafts & experiments and create a new experiment and apply your draft to it.

  • Name it something similar to the original campaign so you know what campaign it is and what you are testing.

  • Pick the Start and End date, it usually defaults to starting a day after you are working on creating it.

  • Select if you want the experiments to run based on searches or cookies.

  • Then start the experiment.

  

Use Cases:

  • Bidding - Google does not suggest this due to the learning curve for automated bidding. It will need the 2 week learning period in the experiment and then again if you choose to apply it to your campaign.  It does keep your campaign safe from a new bidding strategy that does not work well though.

  • Landing page testing - if you want to split traffic to two different pages and Google optimize or instapage a/b testing is not an option or not a good fit.

  • Bid modifiers - can test how different bid adjustments to things such as dayparting, mobile bid adjustments and more can affect the overall performance of the campaign.