Analytics
4K members online now
Understand information in your reports and troubleshoot reporting issues such as self-referrals, (not set) data, and inaccurate information
 
Guide Me
star_border
Reply

Revenue figures not matching in 'All Campaigns' Report between All Users segment and custom segment

Visitor ✭ ✭ ✭
# 1
Visitor ✭ ✭ ✭

Hi everyone,

Thank you for reading this post.

 

I am trying to track the ROI on certain Email campaigns my team is running to target international user-base. The problem I am facing is that I have 3 different figures and I am not sure which is the correct one. 

The campaign I am tracking here is named 'AsktoAstrologer'

 

1st figure (4339.08) - It shows as Revenue in all Campaigns Report when the segment selected is 'All Users'.

 

2nd figure (7928.06) - It shows as Revenue in all Campaigns Report when the segment selected is 'International Email Campaigns' filtered by 'Sessions'.

 

3rd figure (13,825.12) - It shows as Revenue in all Campaigns Report when the segment selected is 'International Email Campaigns' filtered by 'Users'.

 

Also, the number of transactions increases when I view the report in segmented view.

 

I understand that why there is a difference between the segmented views by sessions and users. But I am failing to understand the 1st figure of 4339.08. How is it calculated and should I consider it?

 

Please help me understand if I should figure 1 as my ROI or any figure between 2nd and 3rd based on my need to know the ROI by sessions or users. 

 

Please find below screenshots for reference:

 

All Campaigns - All Users SegmentAll Campaigns - All Users SegmentAll Campaigns - International Email Campaign Segment by SessionsAll Campaigns - International Email Campaign Segment by Sessions

 

1 Expert replyverified_user

Revenue figures not matching in 'All Campaigns' Report between All Users segment and custom segment

Rising Star
# 2
Rising Star

Quick reflection - I suspect part, if not all, of your issue is to do with your sampling. In your second report, you are only looking at ca.50% of the data (shown beside the date selector). This will have an effect on all the reported metrics, including revenue. I suggest you invest some time in removing the impact of sampling, and then see if the figures match. Methods to achieve this can be found here: http://www.lunametrics.com/blog/2013/06/24/solutions-google-analytics-sampling-problems-8-ways-data/

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

Revenue figures not matching in 'All Campaigns' Report between All Users segment and custom segment

Visitor ✭ ✭ ✭
# 3
Visitor ✭ ✭ ✭

Hi John,

 

Thanks for the quick response. That's truly a new insight for me as I am new to this reporting. The figures matched when I reduced the date range to a point where sampling was 100% (May 1 - May 8). 

 

But, is this the figure (4339.08) I should trust and report to my manager? 

 

Thanks!

Revenue figures not matching in 'All Campaigns' Report between All Users segment and custom segment

Rising Star
# 4
Rising Star

It's sampled - so it's not accurate, in the sense that it will drift further away from your back end revenue figures and the reality of the unsampled data. You can 'trust' it in so much as, relative to the other campaigns during the same time period, you can rely on the trends to be reasonably robust. But I would be cautious about using it for any detailed ROI analysis. If the aim was to say "how much money did we make?" then I would definitely be trying to reduce the amount of sampling as much as possible. As you probably know, your GA revenue will almost certainly differ from your back-end anyway so sampling will only further compound this problem.

 

As you are using session based metrics then you can chop up the data (for example, method 8 in the linked article) and get around the sampling issue. 

 

You could go as far as calculating the margin of error (http://www.lunametrics.com/blog/2016/03/03/accurate-sampling-google-analytics/) but, depending on your audience you may confuse them - or worse - undermine their belief in the data. The last thing most managers want to see is a revenue number with a +/- 10% caveat on it!

 

I'd recommend that you get a handle on the way sampling effects this kind of data - particularly if you are going to be reporting it regularly. I've used google sheets and the GA add-on to query session data from the API in sufficiently small chunks to avoid sampling - I could then stitch the results together to encompass my desired data range. I could then compare this with what I got from GA to get a handle on how sampling affecting KPIs such as revenue.

 

You could also use this as a case to buy in Premium, where data is available unsampled!

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