# Medium for Sessions vastly different between PP and GA3

**URL:** https://community.piwik.pro/t/medium-for-sessions-vastly-different-between-pp-and-ga3/2345
**Category:** Reports and working with data
**Created:** [June 9, 2023, 12:58am UTC](https://community.piwik.pro/t/medium-for-sessions-vastly-different-between-pp-and-ga3/2345 "2023-06-09T00:58:01Z")
**Posts on this page:** 3
**Page:** 1

<div class="post-metadata">

### Author: ![David\_Culbertson](https://dub1.discourse-cdn.com/flex005/user_avatar/community.piwik.pro/david_culbertson/32/1479_2.png) [@David\_Culbertson](https://community.piwik.pro/u/David_Culbertson)
#### Post date: [June 9, 2023, 12:58am UTC](https://community.piwik.pro/t/medium-for-sessions-vastly-different-between-pp-and-ga3/2345/1 "2023-06-09T00:58:01Z")

</div>

Hello,

I had asked a similar question about a month for big differences in channel distribution of sessions between GA3 and PP even though total sessions is roughly similar. Now, I am taking a deeper look, focusing on sessions by medium. For penzone.piwik.pro for 1 June through 7 June:

GA3, total sessions = 13,562

Direct - 6,031  
Email - 582  
Organic/Local (Local based on UTM codes) - 6,744  
Other mediums - 205

PP, total sessions = 14,022:  
Direct - 8,414  
Email - 327  
Organic/Local (Local based on UTM codes) - 5,070  
Other mediums - 211

The medium counts that are vastly different are Direct and Organic/Local, almost as if something is happening in how mediums are assigned. Are there differences between GA3 and PP for assigning the medium that could account for the huge mismatches?

---

<div class="post-metadata">

### Author: ![Adam](https://dub1.discourse-cdn.com/flex005/user_avatar/community.piwik.pro/adam/32/1494_2.png) [@Adam](https://community.piwik.pro/u/Adam)
#### Post date: [June 9, 2023, 9:00am UTC](https://community.piwik.pro/t/medium-for-sessions-vastly-different-between-pp-and-ga3/2345/2 "2023-06-09T09:00:40Z")

</div>

Hi David,

One of the reasons could be the method Google uses to attribute conversion for sessions with direct/none referrer.  
Please take a look at graph available on [Campaigns and traffic sources - Analytics Help](https://support.google.com/analytics/answer/6205762?hl=en). Please note that for campaign data, Google Analytics takes the last known traffic source even if the source of the visit was in fact direct. It can lead to significant differences since we don’t do that.

---

<div class="post-metadata">

### Author: ![David\_Culbertson](https://dub1.discourse-cdn.com/flex005/user_avatar/community.piwik.pro/david_culbertson/32/1479_2.png) [@David\_Culbertson](https://community.piwik.pro/u/David_Culbertson)
#### Post date: [June 9, 2023, 8:53pm UTC](https://community.piwik.pro/t/medium-for-sessions-vastly-different-between-pp-and-ga3/2345/3 "2023-06-09T20:53:53Z")

</div>

Thank you, Adam. I took another look using new visitors only over the past 7 days. The Organic/Local numbers are closer but the Direct medium number is still vastly different. Weird.
