Traffic Allocation
Traffic allocation sets the percentage of your visitors who become part of the campaign, from 1 to 100. It defaults to 100. Lower it to introduce a personalized experience gradually. If your site gets 100 visitors a day and you want 20 of them in the experiment, set it to 20%.This is one figure for the whole experiment. Personalization has no control group and does not divide traffic between variants. Which variant a visitor receives depends on the audience they match, not on a percentage split.
Define Target Type and Its Variants
Target Type is chosen once and governs the whole experiment. Each variant then holds its own values for that one type — you cannot target one variant by location and another by device.
Below the dropdown, each variant appears as its own block. Use the pencil to rename a variant, Remove to delete one, and Add Variant to add another.
The fields inside a block depend on the type. Locations offers Countries, Region/States, and Cities.
Worked Example
A two-variant experiment with Target Type set to Locations:
A visitor browsing from Delhi receives Variant A. A visitor browsing from Anaparthy receives Variant B.
Key Considerations
- Every variant needs values before the experiment can start.
- Target Type applies to the whole experiment. Changing it changes it for every variant.
- Check that each set of values matches meaningful traffic before launching. A variant targeted at a few dozen visitors a week will not produce readable results.