The Optimization tab of a dialogue displays a table with statistics on the original dialogue and its variants. Open a dialogue and select the Optimization button to see the distribution and metrics for the original dialogue and its variants.
To interpret these numbers, it is important to be aware of the details. Keep these items in mind when you are working on dialogue optimization:
- Distribution determines the chance to view a specific variant upon first viewing the dialogue. So in the example above, when someone enters the page there is a 10% chance they will be in the control group and see nothing; a 45% chance they will see Variant A; and a 45% chance they will see Variant B.
- Once someone is assigned a variant, they will keep seeing the same variant.
- Distribution has nothing to do with the actual number of views registered per variant. BlueConic will not correct the distribution based on the registered number of views when the variants strategy is set to "A/B testing" or "Rotating."
- The settings on the When tab apply to all variants, including the control group variant.
- For regular variants that have a position configured, a VIEW will only be tracked if the position is actually available on the page the dialogue is executed on.
- The control group variant will always get a VIEW tracked when the dialogue is eligible to be executed on a page for visitors in the control group.
See Dialogue Optimization for information on creating and controlling variants and A/B testing.