The Hair Care sample data table results from a marketing campaign designed to increase purchases of a hair coloring product targeting both genders. For purposes of designing the study and tracking purchases, 126,184 “club card” members of a major beauty supply chain were identified. Approximately half of these members were randomly selected and sent a promotional offer for the product. Purchases of the product over a subsequent three-month period by all club card members were tracked.
The data table shows a Promotion column, indicating whether the member received promotional material. The column Purchase indicates whether the member purchased the product over the test period. For each member, the following information was assembled: Gender, Age, Hair Color (natural), U.S. Region, and Residence (whether the member is located in an urban area). Also shown is a Validation column consisting of about 33% of the subjects.
For a categorical response, the Uplift platform interprets the first level in its value ordering as the response of interest. This is why the column Purchase has the Value Ordering column property. This property ensures that “Yes” responses are first in the ordering.
Select Help > Sample Data Library and open Hair Care
Select Analyze > Consumer Research > Uplift.
Select Promotion and click Treatment.
Select Purchase and click Y, Response.
Select Gender, Age, Hair Color, U.S. Region, and Residence, and click X,  Factor.
Select Validation and click Validation.
Graph after Three Splits
The graph indicates that uplift in purchases occurs for females with black, red, or brown hair and for younger females (Age < 42) with blond hair. For older blond-haired women (Age  42) and males, the promotion has a negative effect.
To launch the Uplift platform, select Analyze > Consumer Research > Uplift. Launch Window for Uplift shows a launch window for the Hair Care sample data table. The columns that you enter for Y, Response, and X, Factor can be continuous or categorical. In typical usage, the Treatment column is categorical, and often has only two levels. If your Treatment column contains more than two levels, the first level is treated as Treatment1 and the remaining levels are combined in Treatment2.
Launch Window for Uplift