Uplift modeling, also known as incremental modeling, true lift modeling, or net modeling is a predictive modeling technique that directly models the incremental impact of a direct marketing action on an individual's behavior.
Uplift modeling uses a randomized scientific control to measure the effectiveness of an action and to build a predictive model based on the incremental response to the action. The response could be a singular variable (i.e., a website visit) or a continuous variable (i.e., customer revenue). Uplift modeling is a data mining technique that’s used predominantly in the financial services, telecommunications, and retail marketing industries for up-sell, cross-sell, and retention activities.
The uplift of a marketing campaign is defined as the difference in response rate between a treated group and a randomized control group. This allows a marketing team to isolate the effect of a marketing action and measure the effectiveness or otherwise of that individual marketing action.
Many marketers define lift (rather than uplift) as the difference in response rate between treatment and control, so uplift modeling can be defined as improving (upping) lift through predictive modeling.
Traditional response modeling typically takes a group of treated customers and attempts to build a predictive model that separates the likely responders from the non-responders. This modeling uses decision trees or regression analysis.
In contrast, uplift modeling uses both the treated and control customers to build a predictive model that focuses on the incremental response. Uplift modeling provides a scoring technique that separates customers into the groups based on how likely they are to make a purchase. The higher their score, the more likely they are to make a purchase (at least once).
Traditional response modeling often targets people they know for sure will buy and, unlike uplift modeling, is unable to distinguish them from the moveable middle.
Read more about how to scale your social ad spend by understanding algorithms and audience segment, like the moveable middle.
Yes! Uplift modeling typically delivers strong ROI for demand generation and retention campaigns. This is because uplift modeling focuses on incremental responses only, so the return per unit is increased.
News, articles, case studies, and other performance marketing resources – straight to your inbox, once a month.