Hi all, We use Mixpanel events on our Wordpress/W...
# questions
c
Hi all, We use Mixpanel events on our Wordpress/Woocommerce store to track where our customers are originiating using UTM properties and so on and we are noticing some slightly strange differences in the data between when our caching plugin (Litespeed LSCWP Caching plugin) is switched on and when it is off. The issue is clearly shown by the Mixpanel report attached We have created a
page-view
event that fires on every Page View (as you’d expect). The graph shows page views where the initial referrer is OUR OWN site - on the first visit for
page-view
events. And the peaks are when Litespeed is active, the troughs are when it is off. Our overall daily count of
page-view
events is around 2K so as a percentage, this affects 10-15% of new visits when Litespeed is on and 2% when Litespeed is off. It seems counter intuitive that the Initial Referring domain is our own site on the first visit, because if that was the case, then it wouldn’t be counted as the first visit. But why does this differ so greatly with Litespeed caching? My initial thought was that visitors are arriving at a page and leaving before the
page-view
event fires, their
document.referrer
is set to our site, but the
page-view
event only fires for the first time on the second page. In my Litespeed plugin Page Optimization settings I have turned off Load JS Deferred to try and ensure the
page-view
event fires as early possible, but this has had no effect on the graph. The only thing that works is disabling Litespeed plugin entirely (we still have 2% of visits affected even with Litespeed off, but this is more acceptable than 10-15%) - but obviously I don’t want to turn Litespeed off because of the speed gains. It’s probably a long shot - but if anyone has any thoughts about why this might be happening, I’d appreciate it.