The client is among the world’s largest restaurant companies and ,wanted to engage with customers through 1:1 personalized marketing and in turn maximize revenues, profits and CLTV. Apart from having a manual testing process with nascent ML capabilities, the existing system offered limited personalization and was unable to correctly attribute sales to specific offers.
Sigmoid defined and implemented core data science modeling logic for email 1:1 personlized marketing to achieve scalable models with tested, well-documented code. We also built personalized customer and offer affinity model using diverse datasets and performed Multi Armed Bandit testing to explore and exploit routing offers, customer content and select attribute values to optimize value over time.
Implemented multi-armed bandit (MAB) approach for personalized marketing effort resulted in 23% increase in CTR and 12% improvement in average conversion rate.