NEW Google Ads Experiments Setup

Google Ads Experiments let you test components of your campaigns to see if an ad variant, bidding strategy, or any other number of variables could help you perform better. In this video, we'll walk you through the set up process for an Ad Variant Experiment as well as a Custom Experiment to test new bidding strategies. We'll also show you how to review performance data and determine a winner once your experiment has been running for a while.
0:56 - Quick Recap on the Old Way to Set Up Google Ads Experiments
2:47 - The New Purpose for Drafts in Google Ads
3:37 - Start of New Experiments Flow in Google Ads
3:53 - Experiments Menu in the Left Hand Navigation
4:26 - Google Ads Experiment to Optimize Text Ads Demo
4:53 - Filtering Campaigns and Ads for the Ad Text Experiment
7:00 - How to Create New Ad Variants to Test
8:59 - Quick Overview of Launch Settings for Ad Experiment Test
10:15 - Creating a Custom Experiment in Google Ads
10:29 - Experiment Settings and Filtering for Campaigns
11:49 - Understanding Campaign Structure for a New Custom Experiment Campaign
12:19 - Customizing Your Trial or Experiment Campaign with Custom Experiments
13:16 - Scheduling & Settings for Custom Campaign Experiments
13:43 - Determining Experiment Goals
15:24 - Budget Allocation and Experiment Breakdowns
16:16 - Cookie vs Search Based Experiments
17:28 - Choosing Experiment Date Ranges
18:14 - Enabling Sync Changes Across Base and Trial Campaigns
19:59 - Analyzing Performance of Experiment Campaigns
#googleadsexperiments #googleadscampaignexperiments #googlecampaignexperiments
--
Have a topic you want the Paid Media Pros to breakdown?
Let us know in the comments below!
CHECK OUT OUR AFFILIATES:
Unbounce - bit.ly/pmp-unb...
Instapage - instapage.grsm...
Supermetrics - bit.ly/superme...
TubeBuddy - www.tubebuddy....
Canva - partner.canva....
Find more about the Paid Media Pros here:
Website: www.paidmediap...
Merch: my-store-c4f63...
Facebook: / paidmediapros
Instagram: / paidmediapros
Twitter: / paidmediapros

[@lylam4763]

New experiment flow starts at 3:44. Thanks for the content!

[@PaidMediaPros]

Glad you liked it!

[@darkspace3316]

Good video

[@PaidMediaPros]

Thanks for watching!

[@priya_mailmodo]

I was looking for a good educator. Loved this video, Michelle!!🔥

[@PaidMediaPros]

Glad it was helpful!

[@casparhickman6406]

Thanks for your Detailed walkthrough! and providing a performance analysis section at the end! Thanks,

[@PaidMediaPros]

Glad it was helpful!

[@mfrapp20]

Great video! The overview of custom experiments was just what I needed

[@PaidMediaPros]

That’s awesome Matt!

[@christophelourenco9193]

Super duper helpful. I was looking to test 2 bidding strategies against each other and thanks to you I've been able to within minutes! Cheers

[@PaidMediaPros]

That’s great! Hope the test went well!

[@stantonwindowsandmore6704]

Fantastic video with extreme detail shown, enabling us to immediately benefit from your knowledge. Thank you so much!

[@PaidMediaPros]

Glad it was helpful!

[@saroc7671]

ahhh so helpful, thank you!

[@PaidMediaPros]

Glad it was helpful!

[@maialbaghdady8685]

Thank you very much for the video. I always come here to understand what I don't understand from Google. But I have a question re. custom experiment, would it be beneficial if I make an experiment for totally different campaigns? I have previous campaign and my boss wants to experiment it against totally new setup campaign.

[@yuriykabanets7665]

Thank You!

[@PaidMediaPros]

Thanks for watching!

[@brunovincent1969]

Is there a way to set this up with Google Ad editor desktop? I can't see any settings for it

[@bassbass9175]

This was a great vidoe! Thank you.

[@PaidMediaPros]

Glad you enjoyed it!

[@nikolinatomaskovic]

Thanks for this video! Once the experiment for search campaign started, if sync in enabled, is it better to make no changes and no optimization is needed at all? And what about if sync is off?

[@baszero]

Hi Michelle, let's say you have segment on your site that is a premium user segment and you don't want to show experiments to those. How can I disable experiments in certain situations?

[@deanryanmartin]

An experiment setup is different from split testing, right?

[@pascalgermain4956]

Good content.. Do you have a video explaining what affects the campaign's data when we do a change or update on an ad, conversion or anything ? I mean when we change something into a campaign .. will it affect the performance ? Like this experiment as an example .. when we apply it .. then what will be the affect on passed data collected for optimization ?

[@danielkeebler6205]

Thanks for the video. Does the new flow also mean that experiments for Shopping Ads is no longer available?

[@mutazghanem]

Great video, thank you i have a question my goal at bid strategy is Encpc bid when I choose the metrics at goal i will choose conversion then i choose to increase the second metric is cost and the other box no significant thats right?

[@digitalmarketing2706]

Hi. GREAT video. Question for you. I used the ad variation test to set up and nothing is happening. It says "Active" but also says "You don't have any enabled ads for this ad variation". I called Google and they said it was set up correctly and to give it a week. But I'm thinking the ads aren't enabled. Any advice?

[@zhihaoxu705]

When will it be possible to set up the custom experiment for app campaign? We are advertising for our mobile games and want to evaluate our campaigns by experiment.

[@PaidMediaPros]

As of right now it’s not possible to create experiments for app campaigns since they are so automated. I wouldn’t expect it to come anytime soon or at all.

[@rahulmukherjee5199]

Is it possible to test variations of Pmax campaign? Is there anything different in the setup while testing a landing page?

[@D0naxa]

If I have successful experiment with another bidding strategy and I apply bidding changes to original campaign, will all my algorithmical learning be applied there as well? Or i need to create a separate campaign so all learning will be kept?

[@mei33r]

Thanks for the informative video!! But can you elaborate on what happens after clicking "end experiment"? Does it mean the trial campaign would just be deleted for good?

[@PaidMediaPros]

Hi, when you click "End Experiment", that means Google will no longer split the traffic between your control and experiment campaigns. If you want to make the changes from your experiment to your live campaigns, you'll have to click Apply, otherwise things will go back to the way they were before the test. Hope that helps!

[@goinghigher7349]

How do I turn off the experiment?

[@whatsupolya]

Hi! I have a question, can you go back to the experiment if it is still in draft stage and change the start date? Thank you!

[@prathimareddy336]

Great video. Can we create experiments for app campaigns as well? I am unable to create. Pls let me know

[@PaidMediaPros]

Experiments are not available for app or shopping campaigns. support.google.com/google-ads/answer/6261395?hl=en

[@prathimareddy336]

@@PaidMediaPros Thanks

[@rezasemnani7421]

Many thanks for the great content. After we set up our experiment campaign (schedule section) and create experiment, how can we edit the "Experiment Goal", "Budget Edit" or "Enable Sync"? I can't find them in the setting. Appreciate your help.

[@PaidMediaPros]

Hi Reza, I believe some of those things are locked in once you have launched your experiment. The Experiment Goal and Enable Sync, I believe, aren't changeable after you start the test. For the Budget, you should be able to change the budget on the Control version of your campaign and it will be shared equally across the Control and Experiment. Hope that helps!

[@rezasemnani7421]

@@PaidMediaPros Many thanks for your reply! It's crystal clear now.

[@tomcha75]

Has this flow or setup changed again in the last year? It's not what I'm seeing when trying to create an experiment.

[@PaidMediaPros]

Hi there, thanks for the note! For the most part this process is pretty much the same, but it might be time for us to create an updated version of this video to cover some of the nuanced options.

[@mediatixgmbh]

Hi - very helpful video. One question looking at the "cookie vs. search-based" experiment option: Would you still prefer the cookie-based vs. the search-based option in times of user consentent (GDPR / CCPA ...) world?

[@PaidMediaPros]

Hi Philippe, I would still prefer cookie based as much as possible as most people currently just accept all the cookies that they're presented with. That said, this might cause some tests to have a lower data set, so if you feel like the test data is too low compared to your normal traffic, then it might make sense to shift to search based. Hope that helps!

[@MattM-df5sf]

How do you remove an experiment? Mine ended months ago and wherever I try to remove it, it says "You can't modify the status, date, budget, or name of an experiment campaign". Really?!

[@DanielADeYoung]

With Experiments... Can I use a CPA strategy? Can I use an existing portfolio strategy? Cannot seem to get that to work???

[@PaidMediaPros]

Hi Daniel, you should be able to test a CPA strategy, but I don't believe experiments will work with a bidding portfolio. I think they have to be separate bidding strategies to test 1:1 rather than 1:an existing portfolio. Hope that helps!

[@alexandruvida8813]

Hi. I have a question. Right now I have 1 shopping campaign with 3 ad groups (rings, bracelets, necklaces). If I want to add products daily or every 2 days, should I add the new products inside the ad groups? So, if I want to add a new ring, I’ll put it in the rings ad group within the same campaign? Thank you!

[@TopTurfLawnCare]

Hello, I work for Top Turf and our marketing department. But I also run ads for various different companies on the side. I would say if you are just adding new products like your rings put the ads inside the "rings ad group" within the same campaign. It depends on what you're looking to accomplish. Are you looking to do a comparison.. also depends on the comparison. May I ask what rating do you get for ad strength?

[@AndrewScarella]

Hi. I have a question about an experiment I am running where I am comparing results between ads that drive traffic to a landing page versus the home page of a website. If the homepage contains more keywords, would my keyword relevancy score go up in the experiment campaign? Or does it just split the traffic 50/50 and not modify any of the keywords?

[@PaidMediaPros]

Hi Andrew, the experiment will only split the traffic and won't augment the keywords in any way. There could be a small chance that in separate campaigns the keywords have different relevancy scores based on the landing page, but in my experience, it would be a pretty minimal impact. Hope that helps!

[@benjamingoldman9988]

When setting up an experiment to perform a LP A/B test, using ad variations, you cannot segment by conversion action which is a concern. Therefore, using the custom experiment route, when selecting the 2 goals, will that bias Google in any way, or are those just for me?

[@PaidMediaPros]

Hi Benjamin, that would depend on what type of bidding you're using. If you're using automated bidding, then the conversion actions will influence how Google is seeing "success" and will likely impact the test beyond just the variables you're testing, but if you're using manual bidding, then that should level the playing field a little. It's not a perfect system, but it should get you closer to an answer on which set up performs better. Good luck!

[@LuisLondonG]

Skip to 3:40 to see the campaign set-up

[@earlytoriseab]

I wanted to change the destination of display ads (HTML), but I couldn't quite get it to work. It doesn't seem possible to bulk update display ads, only search and call ads, so I ended up changing all the ads individually, but that also changed to the destination of the base campaign. Is not not possible to change the destination of an HTML-campaign in experiments?

[@earlytoriseab]

Been chatting with Google Support and it doesn't seem possible to run an experiment with a different destination using HTML ads.

[@PaidMediaPros]

Yeah HTML ads are different. We honestly have rarely used them because it is more difficult to make some simple updates.