Time to let your AI out of the box?

Time to let your AI out of the box?

Why predictive models are not enough for decisioning in marketing.

Now that enterprises have consolidated their data in warehouses and CDPs, they are turning to machine learning (ML) to leverage the value of this data. Businesses have typically used a type of machine learning (ML) called supervised learning to build predictive models. These models are used for applications like churn prediction, category affinity, cross-sell propensity, and next best offer. 

But an ML strategy that relies only on predictive models will run into problems. In this white paper, we explain how predictive models need to be supplemented with reinforcement learning (RL) models for decisioning.