I have been running some ad text versioning on an account that I manage and noticed some unusual activity. I ran 3 types of ad text versions with impressions rotated evenly: 1. advertising low quantity pricing 2. advertising high quantity pricing 3. no quantity/price mention at all There was a clear winner but when I paused the other 2 ad text versions I concluded that order volume and total revenue would go up across the account since more people are seeing our best performing ad. Even though orders did go up for the winning ad text once the others were paused I actually saw total orders go down for the account with less ad text diversity. My question is this: Even if you have a clear ad text winner, does it still make sense to run all three ad text versions because what might appeal to one searcher might not appeal to all searchers and that by eliminating the losers you are actually killing of a percentage of searches who were compelled to purchase from the lower performing ad text? Does a typical user search the same keyword several times and eventually see one of the three ad texts that makes them click through and buy? Also, has anyone else seen a similar behavior? I may end up running a week over week comparison for a few months to try to get to the bottom of this and determine if running only my best performing ad text is actually hurting my bottom line. Any thoughts and insights would be greatly appreciated. Thanks.
Just because an ad is best CTR-wise, doesn't mean it's best conversion-wise. Yes, your ad can affect your conversion. Use the ad that's best at both. Multiply both values, highest one wins. Or you may just be experiencing normal variation or seasonal effect.
Sorry, when I said that there was a clear cut winner, I meant to say that there was an ad text that was the winner in terms of orders and conversion rate. Not by CTR which was only slightly better. My question is has anyone seen a trend where you paused all but your best converting ad text and seen total orders actually drop because you lost the ad text diversity that appealed to more peope than just your top ad text. I feel like some discussions on ad text a/b versioning are somewhat oversimplified in that if you test version 1 vs. version 2, and version 1 ends up having a better conversion rate, if you then run version 1 by itself then your overall campaign will perform better. My guess is that if you run different ad text that appeals to different types of shoppers (say bargain shoppers vs quality shoppers) for the same keyword you might get more sales overall because when the searcher refreshes the screen or runs the search again they might have the chance to see both of your ads and you would then appeal to a larger demographic. I'd love to see a versioning system that runs your best ad text first in a searchers session and then ran the lower converting ad text in order of best to worst conversion rate. You'd then ideally put your best foot forward and then keep trying again with different ad text. Unfortunately, I don't think google has a way of rotating ads like that at this point. I hope that clarifies my question some. Thanks.
Maybe only some keywords in your group are responsible for your high conversion rate with a certain ad while the others do better with another ad. You might have just picked the wrong ad for the overall group. If that's the case, split those keywords into a separate group using the high performing ad for them. That's what A/B testing is all about. Google won't do it for you. You have to figure it out yourself.
This is why you should put 1 keyword to it's own Adgroup if it generates enough impressions and clicks. Then, you aren't having to worry about if Ad A performed better for some keywords whilst Ad B performed better for the rest. I only ever put more than 1 keyword into the same Adgroup if they produce minimal impressions/clicks (relative to my account) and they are closely themed.