"Experiments are a great feature of the Ads platform. For example, you can run a bid strategy experiment wherein the "control" bids toward a cost-per-lead target (CPL) and the "treatment" toward return-on-ad-sales (ROAS). The ability to run Performance Max experiments is new and very helpful. There are three types. Advertisers can test a control setting against: Another campaign type (Shopping, Search, or Display),"
"The final experiment type, Uplift, is the most interesting as it shows the incremental gains of using new or existing Performance Max campaigns alongside other types. The control and the treatment will each receive 50% of the traffic. The treatment includes the Performance Max and comparable campaigns. Google defines "comparable campaigns" (which are editable) as having the same domain, one or more overlapping conversion goals, or overlapping locations."
Performance Max now supports three experiment types: testing against another campaign type, final URL expansion, and uplift tests that measure incremental gains when Performance Max runs alongside comparable campaigns. Experiments split traffic evenly between control and treatment, enabling direct comparisons such as a 50/50 Shopping versus Performance Max split. Final URL expansion exposes half of the traffic to Google's URL optimization versus advertiser-selected URLs. Uplift tests compare control campaigns to treatments that include Performance Max plus comparable campaigns; comparable campaigns share domain, overlapping conversion goals, or overlapping locations. A new data-exclusions feature enables advertisers to exclude specific audiences from seeing ads.
Read at Practical Ecommerce
Unable to calculate read time
Collection
[
|
...
]