i would suggest to play with the new functionalities of virtual report suites: Virtual Report Suites
just create a new VRS on the desired Report Suite with no segment applied and play with the session timeout: Context-Aware Sessions
this way you might find out if it is really a problem of session timeout or not ...
hope that helps
If you have visits that start with your payment processor, that has nothing to do with the quality of your implementation (generally speaking, at least).
Adobe Analytics determines a session break by 30+ minutes between image requests. This means someone could be sitting on the checkout page, about to confirm a purchase, get distracted or forget about it for an hour, then make the purchase. The site could very well be open the entire time, but since Adobe Analytics determines sessions based on the timestamp of image requests, you'll always get new visits that are really the continuation of an long-stalled session.
I agree this can happen, but should always be the minority. The majority is expected to have a straight journey, otherwise something is poor.
But I tested it myself and found that the session doesn't break or end, it is rather wrongly attributed. Entries of direct traffic are strangely attributed to the first encounterd referrer.
So if I come directly to website and conduct a payment, in the reports my entry is marked as coming from this referrer instead of typed/bookmarked. Still have to find out why, though. I believe it has something to do that we always send the referrer variable, even if it is empty, then we just send an emtpy string "". Maybe this conflicts...
No, because it is part of the user journey and we want to see outgoing traffic as well. But maybe we'll consider this option if nothing else helps