knowledge base

We're here to help.


Control Groups

  • A control group is a portion of unique targeted users who do not see the actual ads so it’s possible to measure and compare the “performance lift” provided by that campaign
  • Control groups can be set on every type of campaign except Branding & Acquisition
    • Control groups cannot be set for individual Ad Groups and are run at the User level, not the impression level
  • The proper way to measure campaign vs control is by Revenue Per Visitor (RPV)
    • RPV is a composite of Average Order Value and Conversion Rate
  • Measuring the lift of a campaign over control group using either AOV or conversion rate may not tell the whole story. You can have a very high AOV, but just several orders; and simultaneously have an abundance of conversions with a very low AOV
  • RPV is calculated by 'Revenue/Unique Users' and clearly indicates if the campaign or the control performed better as you can measure how much revenue each campaign earned at the User level
    • Important to note that no major marketing decisions should be made until Statistical Significance has been reached in your test
  • Control group tests are put into place to test the validity of view-through conversions (VTCs), when running a test
    • There should be a control group on all campaigns; especially highly segmented campaigns such as "Engaged Shoppers" or "Cart Abandonment" campaigns
    • Conversely, you can run control group tests to measure click conversions
      • Include a control group on one campaign only
Was this article helpful?
0 out of 0 found this helpful
Have more questions? Submit a request


Powered by Zendesk