2.2K members online now
2.2K members online now
Learn to identify referral spam traffic, as well as best practices to reduce referral spam in your Google Analytics account
Guide Me

Referral Exclusion and Campaign Timeout

Visitor ✭ ✭ ✭
# 1
Visitor ✭ ✭ ✭

Hi there


I have taken over an account with a large number of self-referrals. Clearly there was an issue with the referral exclusion list but this seems to have been resolved at some time in the past. When i go back month by month, i can see that the last time there was a very large percentage of NEW users tagged in this way was March 2016, 14 months ago. Since then the vast majority of this traffic is EXISTING users, which leads me to conclude that that was the month that they fixed the referral exclusion list.


I understand that after fixing up the referral exclusion list, existing users remain tagged incorrectly until their campaign expires (or they click on an ad/link/search and start a new campaign that way), and therefore after fixing the referral exclusion list you will continue to see a lot of existing users incorrectly tagged but no new users. But the thing is that we are still seeing thousands of existing sessions tagged in this way every month. So...


1. Why has this not stopped, given that the campaign timeout is set to 6 months? I would have expected that by September 2016 this would have dried up.


2. I mentioned that the vast majority of traffic now with internal referrals is existing traffic, but it's not all. Every month there seems to be maybe 10 new sessions tagged in this way. Why might that happen? 







1 Expert replyverified_user

Referral Exclusion and Campaign Timeout

Rising Star
# 2
Rising Star

How large a % of your sessions are the self referred sessions? Troubleshooting self referrals is time consuming and, when you get down into the weeds, very much a question of diminishing returns. I would say that if the self referrals are only a few % of your sessions then allocating large amounts of time to solving this may not be entirely effective - a glaring exception to this, of course, is if your self referrals are, for example, coming from converting sessions. 


Having said that, it may be worth spending some time on identifying what is happening with these self referrals and where they come from. What do they have as a landing page, for example. Use tag assistant to test common, or business critical, visitor journeys to seek out places where a new session may be started. The latter is described here:



John Wedderburn, Advertiser Community, Rising Star
Was my response helpful? If yes, please mark it as the ‘Best Answer.’ Learn how here.

Re: Referral Exclusion and Campaign Timeout

Visitor ✭ ✭ ✭
# 3
Visitor ✭ ✭ ✭

Thanks for replying, Jon. Unfortunately they make up a sizeable percentage. Last month they made up 14% of all sessions, but 50% of all revenue. Of the 6795 sessions that were self-referred, 21 were new users.


Is my understanding of how this should work correct, and the logic i used to trace back when the self referral was added to the exclusion list sound?


I don't see anything in the landing pages that gives me any clues. No doubt if i spent long enough trying to replicate the right user journey i would see what is going on, but it makes no sense to me that this would happen under any condition.


I've look at Annie's article and posted the same question on that page so perhaps she can offer some insight.

Marked as Best Answer.
Accepted by topic author Analytics R

Referral Exclusion and Campaign Timeout

Visitor ✭ ✭ ✭
# 4
Visitor ✭ ✭ ✭

Just wanted to post that i found out the answer, and as John suggested, using Tag Assistant did help. It turns out that within the cart there was a php page being called in the background. This page and the calling html page both sent data to the same GA property, but the php page was still using classic analytics while the calling page uses universal analytics.


For anyone looking at this post in future, one thing that is rarely mentioned is that referral exclusion list only works with UA. This is the first time i've seen an implementation with both UA and Classic scripts on a page, and it's interesting to see the effects.