Total Revenue Goal
I set up a test experiment to do some due diligence on data matching.
I set up a test to only include desktop USA visitors and just split the traffic 50/50 with no variation on the site. The revenue from Original and Variation #1 is not even close to what Google Analytics is reporting as Desktop USA. What's interesting is that it matches the total revenue (including desktop, mow, and tablet - all countries).
Is that expected?
Thanks for reaching out I have a few clarifying questions I'd like to ask.
Have you turned on your Google Analytics Integration within the Project, Experiment, and by adding additional code to your page's source code? This will ensure the scope of how Optimizely and GA defines Revenue is identical. Here is more information about integrating with Universal Analytics:
Also, it sounds like this discrepancy might be related to how segments work. Segments are not set up to be mutually exclusive. For example, a user might be bucketed into the "Mobile" segment and then re-enter the experiment on a desktop. In this case, he would be included in both the "Mobile" and the "Desktop" segments. Does that make sense? Here is more information about Segments.
Looking forward to hearing back from you!
Yes - we did integrate UA with Optimizely and the revenue numbers align nicely (both very close to site total rather than the targeted group).
So - are you saying even though my test was only supposed to show for Desktop USA, the analytics will always show EVERYONE that visited the site for Revenue Goal, and I have to segment from reporting to see my pre-defined audience? Sorry I am still a big confused.
No that's not the case. I'm sorry I think my answer must have been a bit confusing.
Could you perhaps send me a personal message with your experiment ID? I'd like to have a closer look at this particular case and make sure I have a good understanding of your setup before discussing further.