This video DOES NOT show proper statistical A/B testing experiment... this is just one ad/campaign with one change compared to an ad/campaign with no change... This really is a problem because you can't confirm statistically that the ad is performing better because you are simply running two ads next to each other at the same time. There are a lot of problems with this that go against proper A/B testing, such as taht the participants (or in this case, the audience for the ads) MUST be randomly assigned to either group A or group B which is NOT the case here and this is a violation of A/B testing. For example, if they are not randomly assigned into these two groups, there could be a person seeing one ad then the second ad too (which would not happen in a proper A/B test), and that first ad that they have seen might influence whether they convert on the second ad.
Hey, so does just duplicating like this ensure the traffic gets split 50/50? This seems like one would get more traffic eventually, not making the test as relevant as possible? Or am I thinking about this wrong.
yeah i thought the same. how i see it is that you pay double the spending and you indirectly bid up the cpc by competing with yourself. or did i misunderstand him and he pauses the old campaign vs the new campaign? hope someone can respond
so if i ama correct you leave both the campaings on for split testing correct and that is for 2 weeks? not longer? doesnt they work against each other and you even pay more because you bid against yourself? please explain?
Straight to the point and very informative, thanks very much this was a great video and very helpful. So many videos blab on or are confusing. Great job
Hey great video appreciate you putting this together. Had a question about competing keywords though. If I wanted to run an A/B test for if website clicks objective performed better than leads, wouldn’t I effectively be bidding against myself and driving costs up? Is there a better way to test this method
Can anyone answer if/how to set up an A/B test if we want to test different ad types against each other (not creative or creative element, though the 2 ad types would have different creative/creative elements)
If you are changing multiple things (ex different ad types and also different creatives) it's not a true A/B test. Make sure you are only changing one of them
Can we A/B Testing the Headlines & Descriptions? The rest of the settings remain same but we differ the headlines & descriptions to know which ad copy attracts more clicks/leads
The idea is to operate them briefly, for a limited duration of, let's say, two weeks, in order to assess the test's effectiveness in deciding between the two options.
Yes it can be done on either the campaign or ad group level. But if done on the campaign level, you're forcing the budget to be evenly split 50/50. If done on the ad group level, google will tend to spend more on the broad match keywords making the test less effective since one variation is getting less ad spend
If you duplicate the campaign, the new one no longer has six months of data, though, right? Would it use conversions from the whole account to optimize right away?
The account data will help but its not copying campaign level data. You can duplicate the campaign and let it run for a while before making adjustments so it's a more accurate split test. Then make changes and utilize date filters to compare. Alternatively use experiments which run into the same issue (why Google why). Consider always having an experiment running 50 50 for each campaign once they stabilize make the changes.in the experiment. Makes it a little easier to upload the changes to the original campaign. But even if not using experiments always having a full duplicate campaign to manage the budgets seems to help.
Great video tnx! Question: what if you want to test the effects on Cost per Conversion because increasing campain budget? So how to A/B test budget changes? = campaign budget changes. example: a campaign has a daily budget of €10 and I want to find out the effect on increasing the campaign budget to €20 without disrupting the original campaign. For instance, Pmax.
Funny I tried this last night and Google flagged it as a duplicate Ad and disapproved. One Ad was a duplicate of another, only thing different is the URL. Both point to the same domain but different landing page address. Weird.
Experiments very ofen don't push 50% of the cost into your variation. Doing it this way forces the budget to be split perfectly even. As for the cost, so yes, in some examples where you are duplicating on the campaign level, you should cut the budget in 1/2 before duplicating so then once theres two of them the total ad spend remains the same :)
@@axelrunner5662 so if I was to create a split test at the ad group level with different landing pages then you can edit the site link extensions as well correct? Also since they are targeting the same keywords would they be competing with each other? Would it be better to duplicate it at the campaign level and cut the budget of the original in half?
Any advice on how to split test a different domain in the URL? I am selling a book and we want to split-test the URL and send one to our website domain and another to the Amazon page listing. I know you cannot add different domains in an ad group. Any advice on how to test this would be much appreciated.
Could you please make a video on how to decide for bidding price on keywords and get more leads. How we can decide how much price we need to increase and decrease for a keyword and when. How to get relevant clicks and leads.
This video DOES NOT show proper statistical A/B testing experiment... this is just one ad/campaign with one change compared to an ad/campaign with no change... This really is a problem because you can't confirm statistically that the ad is performing better because you are simply running two ads next to each other at the same time. There are a lot of problems with this that go against proper A/B testing, such as taht the participants (or in this case, the audience for the ads) MUST be randomly assigned to either group A or group B which is NOT the case here and this is a violation of A/B testing. For example, if they are not randomly assigned into these two groups, there could be a person seeing one ad then the second ad too (which would not happen in a proper A/B test), and that first ad that they have seen might influence whether they convert on the second ad.
A/B Split Testing your Google Ads made simple! Thanks Axel! 😎🤙
He's the best
Hey, so does just duplicating like this ensure the traffic gets split 50/50? This seems like one would get more traffic eventually, not making the test as relevant as possible? Or am I thinking about this wrong.
yeah i thought the same. how i see it is that you pay double the spending and you indirectly bid up the cpc by competing with yourself. or did i misunderstand him and he pauses the old campaign vs the new campaign? hope someone can respond
so if i ama correct you leave both the campaings on for split testing correct and that is for 2 weeks? not longer? doesnt they work against each other and you even pay more because you bid against yourself? please explain?
Very informative. Great tutorial Axel! 🎉
Indeed!
*Great explanation on split testing!*
Great video! What's the advantage of doing split test for keywords on campaign level over doing it within the same ad group? Thanks!
bro thank you!
Straight to the point and very informative, thanks very much this was a great video and very helpful. So many videos blab on or are confusing. Great job
Thank you man!
Hey great video appreciate you putting this together. Had a question about competing keywords though. If I wanted to run an A/B test for if website clicks objective performed better than leads, wouldn’t I effectively be bidding against myself and driving costs up? Is there a better way to test this method
Can anyone answer if/how to set up an A/B test if we want to test different ad types against each other (not creative or creative element, though the 2 ad types would have different creative/creative elements)
If you are changing multiple things (ex different ad types and also different creatives) it's not a true A/B test. Make sure you are only changing one of them
Thank you!! @@ericbpreston
How to capture leads on those landing pages? and nurture them?
Follow Up Boss ;)
Check out the other reply I left in your other comment :)
Can we A/B Testing the Headlines & Descriptions?
The rest of the settings remain same but we differ the headlines & descriptions to know which ad copy attracts more clicks/leads
Yes, using the Ad variations tool
wouldnt your test campaign compete against your original campaign?
The idea is to operate them briefly, for a limited duration of, let's say, two weeks, in order to assess the test's effectiveness in deciding between the two options.
Keyword match testing can be done at ad group level. Isn't??
Yes it can be done on either the campaign or ad group level. But if done on the campaign level, you're forcing the budget to be evenly split 50/50. If done on the ad group level, google will tend to spend more on the broad match keywords making the test less effective since one variation is getting less ad spend
Thank you 😮 i found a good channel for learning 🎉
Happy to hear that!
thanks a lot, it was so helpful
Glad it helped!
If you duplicate the campaign, the new one no longer has six months of data, though, right? Would it use conversions from the whole account to optimize right away?
The account data will help but its not copying campaign level data. You can duplicate the campaign and let it run for a while before making adjustments so it's a more accurate split test. Then make changes and utilize date filters to compare. Alternatively use experiments which run into the same issue (why Google why). Consider always having an experiment running 50 50 for each campaign once they stabilize make the changes.in the experiment. Makes it a little easier to upload the changes to the original campaign. But even if not using experiments always having a full duplicate campaign to manage the budgets seems to help.
Great video tnx! Question: what if you want to test the effects on Cost per Conversion because increasing campain budget? So how to A/B test budget changes? = campaign budget changes. example: a campaign has a daily budget of €10 and I want to find out the effect on increasing the campaign budget to €20 without disrupting the original campaign. For instance, Pmax.
Funny I tried this last night and Google flagged it as a duplicate Ad and disapproved. One Ad was a duplicate of another, only thing different is the URL. Both point to the same domain but different landing page address. Weird.
Great video. Thanks! I still have an essential doubt. Would url update experiment work if my keywords are linked to their own final url? Best Ricardo
Hi. Great video! Can you update it to the new Ads interface? They changed it and for some reason I'm unable to run the Experiments as expected.
You are in this case doing double bidding and pushing the Cost higher. Why not use experiment that share traffic by 50 50?
Experiments very ofen don't push 50% of the cost into your variation. Doing it this way forces the budget to be split perfectly even. As for the cost, so yes, in some examples where you are duplicating on the campaign level, you should cut the budget in 1/2 before duplicating so then once theres two of them the total ad spend remains the same :)
@@axelrunner5662 so if I was to create a split test at the ad group level with different landing pages then you can edit the site link extensions as well correct? Also since they are targeting the same keywords would they be competing with each other? Would it be better to duplicate it at the campaign level and cut the budget of the original in half?
@@axelrunner5662 but he is duplicating campaigns and on campaign level is the ad buget daily spent? or i am seeing things wrongly?
Thanks
No problem
Any advice on how to split test a different domain in the URL? I am selling a book and we want to split-test the URL and send one to our website domain and another to the Amazon page listing. I know you cannot add different domains in an ad group. Any advice on how to test this would be much appreciated.
He just showed in the first example that you can change the URL
Could you please make a video on how to decide for bidding price on keywords and get more leads. How we can decide how much price we need to increase and decrease for a keyword and when. How to get relevant clicks and leads.