AdWords is now Google Ads. Our new name reflects the full range of advertising options we offer across Search, Display, YouTube, and more. Learn more

Analytics
3.4K members online now
3.4K members online now
Learn how to set up goals and Ecommerce tracking, verify accuracy of attribution traffic sources, and discuss funnel visualization, Multi-Channel Funnels, the Goal Flow Report, and Enhanced Ecommerce
 
Guide Me
star_border
Reply

Email Campaign Metrics Questions

[ Edited ]
Visitor ✭ ✭ ✭
# 1
Visitor ✭ ✭ ✭

I've been using Analytics to track Email Campaign performance and I have a couple of questions; two things that bug me, actually,

1) Date range function failure. I always begin by setting a range of dates, then choosing the metrics I want to compare. Invariably, Analytics displays campaigns far outside the target date range which means I have to sort through a bunch of unrelated data to find the right campaigns. 

Is there a fix for this?

 

2) Data fluctuations. This one's a real nuisance. I pulled data yesterday on some email campaigns we ran in April. I gathered some data, but didn't finish, so I had to reload the data today to continue the project. Several of the reports have changed significantly from one day to the next; number of sessions revised downward. 

How can the number of sessions generated by an email campaign -- that ran two months ago -- drop by 5% overnight? Can anyone explain this?

1 Expert replyverified_user

Re: Email Campaign Metrics Questions

Top Contributor
# 2
Top Contributor
1 - GA uses last-non-direct attribution, meaning if a user visits from your email, then returns a month later direct, GA will attribute the return visit to the email. Unfortunately you can't tell if it was actually direct or not...maybe the landing page is a clue if the email didn't link there. The Multi Channel Funnel reports do not use that attribution model, but have a lot less functionality available.

2 - biggest possibility is sampling, especially when looking at longer time ranges and many sessions in total (>500,000). When you start drilling into specific dimensions with a low count, the sampling causes the numbers to fluctuate greatly from day to day. Shorten your timescale to improve the accuracy.
Mike Sullivan, Google Analytics Top Contributor
Was my response helpful? If yes, please mark it as the ‘Best Answer.’ Learn how here.

Re: Email Campaign Metrics Questions

Visitor ✭ ✭ ✭
# 3
Visitor ✭ ✭ ✭
Thank you Mike.
Now I see what you're saying about the effect the date range has on the campaign results. I guess I was looking at the function of the range of dates backwards, assuming it would only show me data for campaigns *launched* during that period, instead of data for all activity *measured* during that period regardless of when the campaigns were launched.
But if I narrow the date range, as you suggest, it actually skews the data because of the last-non-direct attribution factor you mentioned in point one. For example:
I looked at an April 12th campaign yesterday using a date range of April 1 - May 30, during which time we sent 30 campaigns (some to different segments of our mailing list). When I narrow the date range to just the month of April, the reported sessions and revenue for that campaign drop by 15.2% and 31.9% respectively.
And a lot of the campaigns showing up in this April only report were initiated in March, so again it affirms what you're saying. But since we can't tell if that business was actually direct or not, it's hard to judge the effectiveness of the campaigns.

Is there a tutorial somewhere about the Multi Channel Funnel report? I'd like to do some cross-referencing to see if we can nail this down.
Thanks again for your help!