That page describes web traffic, not app traffic, and this was around a release and not a holiday.

We just updated an app from 1.2 to 4.0, and saw the number of unique visitors immediately spike at release. I'm trying to figure out why. We weren't setting the visitor id before, and aren't (intentionally) now either. To make sure I did a string search on the codebase and didn't find any references...

Aha! No, offlineEnabled is false. I'll give that a shot. 

Using offlineEnabled plus a timeout of 10 didn't help, but that plus a timeout of 60 did. Unfortunately, due to the need to continue tracking un-upgraded versions of the app, we can't turn on offline tracking on the report suite. Apparently that'd cause the server to drop all of those tracking reque...

Hi Hunter, thanks for looking at this. My concern is that the longer that timeout is the more likely it is that another tracking event will occur before that lifecycle hit gets sent. That's why I'm calling it a race condition. There is no way (that I can find) to detect when that InstallEvent hit wa...

Given that our problem is a race condition where other tracking gets sent before the acquisition hit and thus absorbs the campaign data (why does that happen?) I don't think waiting even longer to send the acquisition hit would make the problem less frequent. The information is definitely present an...

I've implemented Mobile App Acquisition as instructed here: https://marketing.adobe.com/resources/help/en_US/mobile/android/acquisition.html I've run into the situation where sometimes the InstallEvent request fails to pick up any of the campaign information. It is always coming through, this is app...

The migration documentation does not cover how resume or launch were handled in 2.x. 

In the 4.x SDK, if the app launches from the background less than five minutes from the previous launch, it counts as a resume, but past five minutes (by default) it counts as a new launch. Was it the same way in 2.x? Was there a time distinction between resume and launch, or some analogue of the li...