We're A/B testing some landing pages, showing some users one version, the rest another.
We split the traffic by using windos.location.replace function, as suggested by Kissmetrics and others. So a user clicking on a ad with destination URL of /landing-page will be redirected to either /landing-page/1 or /landing-page/2
Will this have an impact on our QS? We're seeing for nearly all ads that our 'Landing page experience' is 'below average'. Despite the fact our engagement on our landing pages is high, around 60% of users clicking the call to action button.
Could this be a factor, will Google think we are cloaking the page we are sending users to?
June 2012 - last edited June 2012
Yes, There will be an impact in the QS if the LP change. I suggest you to use Adwords Experimet feature to do A/B Test. This is really a good feature from Adwords to A/B Test.
Here is the link
Instead of windos.location.replace why dont you use Google Analytics content experiments?
@Raghu: Using Adwords Experiment is nice, but makes only sense if your traffic is entirely from adwords. Otherwise you would be testing on a subset of the users that visit your page which is maybe unwanted.
So @10JH, if all your traffic sfrom adwords, use this option, as it will be the safest.
@theniks: I do not see any difference in using Analytics site experiments to using a custom crafted solution. If analytics redirects to different pages it would need to use the same mechanism.
Howver, QS will most likely not be affected. But you might run in another trap. Google states in its Analytics policys:
"Google does not view the ethical use of testing tools such as Content Experiments to constitute cloaking."
BUT later it says:
"We reserve the right to take action on any site that violates Google's Webmaster Guidelines, or if we feel that visitors are being deceived or abused. For example, if a site's original page is loaded with keywords that don't relate to the combinations being shown to visitors, we may remove that site from our index."
Since the idea behind that is, that the user might be seeing sites with lower relevance, Google will most likely not lower your Keyword's score, but will give you a penalty for cloaking.
If your testing pages are alike (lets say red button vs blue button), I see no difference in QS and no problems
If your pages are very different, lets say Imprint vs. Shopping Basket Page you should better not use window.location or analytics experiments, but go for Adwords Experiment.