How does uplift modeling work?

Uplift modeling is a way of predicting the difference that an action makes to the behavior of someone. Typically, it is used to predict the change in purchase probability, attrition probability, spend level or risk that results from a marketing action such as sending a piece of mail, making a call to someone, or changing some aspect of the service that the customer receives.

Uplift modeling doesn’t know the change in behavior for any individual, any more than a normal model can know the behavior of an individual in the future. But it can predict it. It does this by looking at two groups of people, one of which was subject to the marketing action in question, and the other of which was not (a control group).

Just as it is standard to measure the incrementality of a campaign by looking at the overall difference in purchase rate between the treated group and an otherwise equivalent control group, uplift modeling models the difference in behavior between these two groups, finding patterns in the variation.

The uplift segmentation

Uplift modeling divides your customer base into four segments based on the predicted action from a marketing campaign:

Persuadables - The people who respond to your campaign in just the way you hope. They buy (or renew), but wouldn’t have had they not received your marketing campaign.

Sure Things and Lost Causes - If you did not target Sure Things, they will buy (or renew) anyway. Lost Causes won’t buy (or renew) regardless of anything you might do. Including either segment in a campaign wastes budget.

Sleeping Dogs (or Do Not Disturbs) - Your campaign causes attrition, actively driving customers to defect. Including them in your campaign not only wastes budget but negatively effects response and revenue.

Incremental impact