SessionM has a wide variety of machine learning methods to provide further insight into customer behaviors such as product pairings and the detection of opportunities and disruptions in loyalty optimization. These models include neural networks based on convolutional and long/short-term architectures for the natural language analysis of product descriptions, as well as time series models for detecting fraud, anomalies and customer churn from purchase and event history. All these models work in tandem to help understand and predict behavior and achieve personalization.
Calculate affinity scores, probability to purchase, and offer redemption from historical purchase data as well as confidence estimates for every customer/product and customer/offer combination.
SessionM has a proprietary method for automatic testing & pulsing of multi-arm campaigns to allow campaign managers to have insight into which variants perform best towards defined goals/KPIs, maximizing the effectiveness of your overall campaign success.
Calculate probability of churn at an individual customer level as well as an estimate on future transaction frequency and spend. This allows us to calculate customer lifetime value which can be used in personalized segmentation and campaign design.
