How to avoid false hits on original URL in split testing

When running a Split URL experiment in Mida, all website traffic is initially directed to the original URL before Mida's script processes the traffic and re-routes it to a variation URL, if that is the designated action for a particular visitor.

From your analytics software, such as GA, it might appear that all the traffic is being received by the original URL, even when variation URLs are accurately recording their respective shares. 

This discrepancy can produce the misleading perception of increased page views, as each variant retains the correct percentage, yet the original URL shows total accumulation in addition to its own percentage, which matches or nearly corresponds to the total.

To fix this, follow these steps:

1. Check the Order of Tracking Codes: Make sure the Mida tracking code is placed on the page before your analytics software's tracking code. This could solve the issue in most cases.

2. Wrap up Analytics Tracking Code: To make sure the original URL doesn't record traffic when a different version is shown, you need to wrap up your analytics software's tracking code like shown below:

<script>
window.mfunc = window.mfunc || [];
function onMidaReady(callback) {
  if (window.mida) callback();
  else window.mfunc.push(callback);
}

onMidaReady(() => {
  if (!mida.isRedirect) {
    
    // Insert GTM/GA4 Code Here
    
    console.log(`Execute GA4 tracking`);
  }
});
</script>