Option to reset stats only on one variant


#1

It would be very very useful if it was possible to reset stats only on one variant. When we add another AB test page, we often need to publish the page to see if the form works well, and directs the visitor to the right download page, etc. And it unfortunatly counts as a visit. So, from the start I have wrong data, since they are not true visits, but my test of the form. 

So, adding ‘reset stats on variant’ would help when launching a new page variant. This would reduce having misleading insights. It would be much appreciated, and I am pretty sure many other people are experiencing such issue !

Hopefully, it’ll come soon!

Thanks ! 


#2

Hi Jean-Luc! 
Thanks for the suggestion here. Just a suggestion, but if you wanted to clear the stats on one variant you could simply select ‘duplicate this variant’ to get another copy of the variant with all the stats zeroed out. You could also just create a new page from the variant if you want to re-start with absolutely zero stats.

Also, be sure to block out your own IP address so you’re not messing with your stats during testing. 

Hope this helps in the meantime, Jean-Luc! Be sure to let us know if you have any other ideas for making Unbounce even better. :) 


#3

I came here for the same reason and just realized Justin’s suggestion would do exactly what I needed it to do HAH! Although, Justin, the reason for wanting to reset the stats on this variant is because somehow my “goals” box was unchecked so it significantly brought down my overall page stats. It would be nice to be able to reset just that variant AND have the stats from that variant removed from the overall page stats. But this is definitely one step in the right direction for me.


#4

Justin’s solutions are good, but I will also warn against A/B testing two variants over different time ranges. It introduces more variables (time of day, day of week, ad creative changes, market conditions, etc.) that could muddy your results. Resetting only one variant and testing against a control page might not yield apples-to-apples results if the date ranges don’t match up.