Rich-text Reply

Safari - Redirect Test Issues

bjsvejda 07-19-16

Safari - Redirect Test Issues

I have a simple redirect test setup on a free account on desktop only.  It redirects to a page that I then target 100% of the visitors and style it.  The test itself is to test one version of a lead gen form vs another.  The redirect test is setup correctly and I don't believe the additional experiment that is used to style the page we land on is causing any issues.  However, our results are greatly skewed for safari browsers.  I am seeing 98% of the visitors entering through a safari browser are shown as being bucketed into the original and only 1 visitor has been bucketed into the variation.  To top it off, the original says it is converting at 92% which we know not to be true based on what we see in GA and historically speaking.  All of the other browsers look normal with low traffic at this point.  Has anyone run into similar issues with a redirect test on safari desktop browser?  Does anyone have any recommendations as to what the issue could be?  I am going to try and remove the additional setup I have that simply styles parts of the page it lands on, but I honestly don't think that is the problem from what I can see using developer tools.

Level 1

Reneevl 07-21-16
 

Re: Safari - Redirect Test Issues

Hi there,

 

When you are running an experiment with Optimizely it takes a few hundred visitors for the traffic split to match the actual traffic split on the Result page.

This is because the actual traffic split expresses a probability and won't match the allocation on your Results page exactly. As more visitors enter the experiment, the actual split will start to match your allocation.

This might be the reason why there is a discrepancy between your traffic allocation settings and the actual traffic split on your result page.

 

I hope this helps!

Renée van Leijen
Optimizely