AdWords
1.7K members online now
1.7K members online now
Improve your AdWords performance and boost your ROI, CTR, and Quality Score
Guide Me
star_border
Reply

Experimenting with duplicated Ads

[ Edited ]
Visitor ✭ ✭ ✭
# 1
Visitor ✭ ✭ ✭

A month ago I decided to duplicated 2 Ads (Text and Mobile versions) to check if there were significant differences. I am using CPC on this Campaign.

 

After a month, I have got very similar CTRs (3.05% vs 3.25% in the text ads, and 4.63% vs 4.44% in the Mobile ads). I guess that now I should pause the ads with the worst CTR.

 

However, the Text Ad with lower CTR got 9 times more conversions than the other one (45 vs 5 conversions), and the cost/converted click is lower too. Since clearly it's not a copy issue, should I keep the one with higher CTR for Quality Score, or would the one with higher conversions be ranked higher?

 

Thank you,
Pol

3 Expert replyverified_user

Re: Experimenting with duplicated Ads

Explorer ✭ ✭ ✭
# 2
Explorer ✭ ✭ ✭
Hi Pol,

You should look at the conversion rate as well and keep the ad which have better conversion rate. Quality Score and Conversions are good metrics but in this case you should look for conversion rate.

The one with better conversion rate is good as compare to the one with lower conversion rate. Even if lower conversion rate ad have more conversions.

Cheers,
Nouman

Re: Experimenting with duplicated Ads

Badged Google Partner
# 3
Badged Google Partner
Pol,

At the end of the day we advertise to make money, we spend money to make money. ROI (return on investment) is the Major factor for all decisions.

Anything that converts is a PLUS regardless of CTR, Quality Score etc. So always stick with what converts and makes you money.

Re: Experimenting with duplicated Ads

Rising Star
# 4
Rising Star
Hi Pol,

I think the answer to your question is very simple. If you are looking to drive the number of conversions, you should obviously go for the ad copy with the higher number of conversions.

Now regarding the lower CTR on the better converting ad copy, you are facing this scenario because your ad copies speak of something not every user will relate to. Basically since only the user that relate with your ad copy are clicking on your ads there is a high chance that they are better qualified leads.

Quality leads are always good when it comes to lead generation and I would also suggest that you emulate a similar ad messaging (as the one that is converting for you) through out your campaigns. It could help drive the conversion numbers.

Hope this helps!

Cheers,

Shashank
Cheers,
Shashank Singh, AdWords Rising Star |
Did you find any helpful responses or answers to your query? If yes, please hit "Like" or "Accept as Solution"
Marked as Best Answer.
Solution
Accepted by topic author Pol Smith
September 2015

Re: Experimenting with duplicated Ads

Top Contributor
# 5
Top Contributor

Hi Pol, can I just clarify what you mean by "duplicated 2 Ads"?  Do you mean you created different versions of your text Ads, with different text, or simply duplicated them exactly?

 

If the former (which is probably the case), make sure your Campaign settings are set to Rotate Evenly, or you won't get a fair test because AdWords will "fixate" on one Ad rather than evenly testing your variations.  You can check this to some extent by checking the % Served column, when choosing a period where all Ads were active.

 

I'd also encourage you to use more than 2 variations, and try to have 2 Ads that are quite similar at any one time.  The thinking here is that if you only have two Ads, and if they're properly served evenly, seeing one with a greater performance than another doesn't give you much clue as to what element of the text is doing well.  Having more than 2, where 2 or more share similar wording might allow you to have a better guess at which part of the Ad is being found attractive, so you can create further variations using that element and so on.

 

Jon

AdWords Top Contributor Google+ Profile | Partner Profile | AdWords Audits

Re: Experimenting with duplicated Ads

Top Contributor
# 6
Top Contributor
Hi Pol,

The difference in CTR between the ads does not appear to be significant. I wouldn't be concerned with that. Go with the higher converting ad.

Best of Luck!

Pete
petebardo -- Deadhead doing AdWords

Re: Experimenting with duplicated Ads

Visitor ✭ ✭ ✭
# 7
Visitor ✭ ✭ ✭

Hi everyone, thank you for your comments and excuse the late reply,

 

I want to clarify that, as @Cobnut says, I am talking about 100% duplicated ads (same headline, same description, same LP). Image: http://s3.postimg.org/urlyjly4z/duplicated_ad.jpg, see 1st and 3rd ads.

 

Since duplicated ads are competing with each other, I'm thinking maybe if I leave one alone it will "absorb" the rest of the clicks of the other. Therefore, since I understand #Conversions and Conversion Rate would be added up all in one ad, I thought picking the one with a higher CTR would make me achieve a higher Quality Score, and therefore a lower CPC. Does that make sense?

 

@Nouman T: Just to make sure, did you reply considering they are duplicated ads? (same headline, description and LP). Why would you go with the one with higher Conversion Rate?

 

@Eric G: I totally agree we should focus on ROI. I guess when rotating evenly 2 duplicated ads I should always go with the one with the higher #Conversions, which will most likely be the one with a higher CR.

 

@ShashankSingh: Actually, both ad copies are just the same. I really like your explanation on why a low CTR can actually be a good factor, as long as it is converting well. Will consider it on the rest of the campaigns.

 

@Cobnut : I hadn't set even rotation, I'll do that to fair test. And thank you for the ad variations suggestion, I think that is what I am going to do with my ads, therefore I'll mark your post as the solution.

 

@petebardo : Just to make sure, would you recommend the same if considering completely duplicated ads?

 

In general, @Nouman was the only one that recommended me to pick the ad with higher CR instead of  the one with higher #Conversions. I think the reason is because @Nouman was considering the ads to be simply/completely duplicated, while the rest considered a slight variation. Is that correct?

 

Thank you,

Pol

Re: Experimenting with duplicated Ads

Top Contributor
# 8
Top Contributor

Hi Pol, I'm not sure what you expect the outcome of this test to be.  If the Ads are exactly the same as each other, there's no benefit in the test; any variations in conversion rate or other metrics will be down to random variations in who converted after clicking which Ad, not anything proven by method.

 

"Since duplicated ads are competing with each other, I'm thinking maybe if I leave one alone it will "absorb" the rest of the clicks of the other."

 

Ads only "compete" with each other when the Ad rotation settings are one of the optimisation choices (Clicks or conversions) and these can be useful when the Ads are not the same.  However, if the Ads are identical, there is no reason for one to perform better than another, it's a meaningless experiment.

 

Neither, I'll admit, am I sure what you mean here by "absorbing the clicks of the other".  I suspect there may be some misunderstanding here somewhere so if you can clarify what you mean by the above line, we may be able to find it!

 

Jon

AdWords Top Contributor Google+ Profile | Partner Profile | AdWords Audits

Re: Experimenting with duplicated Ads

Visitor ✭ ✭ ✭
# 9
Visitor ✭ ✭ ✭
Hi Jon,

The reason I used duplicated text ads was to find out how important the "random" factor is. I found out CTR percentages are quite similar.

Since this experiment of mine has finished, I would like to remove one of the text ads. The thing is which one should I remove.

By "absorbing the clicks of the other" I mean that, when I remove one of the ads, since text ads are completely duplicated I am guessing that all the impressions & clicks of the removed text ad would go for the remaining one.

Please let me know if I made myself clear.

Thank you,
Pol

Re: Experimenting with duplicated Ads

Top Contributor
# 10
Top Contributor
Hi Pol,

First, ads don't compete with each other if the campaign is set for even rotation. That's not what I saw in the image you posted, one ad is getting far more impressions than the other. That could be that the date range includes a period before the duplicate ad was created. Ads are selected according to the rotation settings.

I ran your numbers for the two ads on this site: http://getdatadriven.com/ab-significance-test. The report indicates that while one ad does outperform the other in terms of conversions of visits. The ad that has the most clicks (and impressions) does convert better. But, as I suspected, the results are not statistically significant, with only a 76% confidence factor.

I often run duplicate ads in an AdGroup, along with at least one test. I leave the rotation setting to "rotate evenly forever". I do this to reduce the number of impressions for the test ad, and to limit the risk. What I mean by this is I have an ad that produces clicks and conversions at a satisfactory rate. If I add a test ad that has a lower that performance, the "control" ad (which has more than one identical ad) still gets a higher number of impressions while allowing the test to run. In general, I have 3 identical ads in a group, and a test. That means the control gets 75% of the impressions and the test gets only 25%. If the test ad performs well, I'll expand the test to get more impressions by either creating a duplicate of the test or by pausing (or deleting) one or more of the control ads.

I have noticed there is a different CTR and conversion rate for duplicate ads within the same AdGroup, but it is rarely statistically significant. If you want to delete one of the duplicated ads, I would probably keep the one with the most impressions. If your rotate settings are either Optimize for clicks or Optimize for conversions, Google is preferring that ad over the one with fewer impressions. That might improve your QS, but I really don't think that's likely.

Best of Luck!

Pete
petebardo -- Deadhead doing AdWords