How We Used Predictive Modelling to Optimise Towards ROI at VGW
November 18, 2019
Optimizing towards long term metrics, such as LTV (life-time value), vs. short term metrics, such as CPA (cost per acquisition), is imperative to efficiently manage marketing spend, to improve ROI (return on investment) and for overall growth. Executing towards long term metrics in the digital world is challenging but possible. Most advertising platforms are ready with short term metrics, but don’t let that stop you. We had this challenge at VGW: we needed to optimize marketing campaigns towards LTV and we were able to do so by implementing a predictive modelling in partnership with Facebook. We built a supervised ensemble predictive model using click & view attributed features, captured during the on-boarding funnel. The model was back tested and used to analyze via A/B tests various campaign optimization, which led to ultimately improve the ROI of our digital marketing spend. You will walk out with an example on how to implement a predictive model to improve the ROI of digital marketing spend.