Our client, a global telecom, had a legacy recommendations engine (RENG), that was often suggesting repeat recommendations and was not learning from customer choices. Thus, our client was failing to increase customer engagement.
Case Study
Recommendations engine
The Business Challenge
The Solution
We helped our client establish optimal pricing and monitor competitors’ actions We built a robust measurement framework using large volumes of real-time data, to help evaluate the impact of the RENG on different customer interactions. We also improved the functionality of RENG: eliminated repeat recommendations, wrong recommendations, wrong system flags. We also developed and executed a three-step improvement initiative.
Deliverables
Collaborative suggestions
live algorithm providing customer-level propensity scores to watch a show that someone similar to them has seen
Machine learning
live algorithm which retrains the itself based on actual customer behavior
Holistic 360 RENG
three modules (metadata, collaborative and machine learning) providing one-set of recommendations unique to each customer
Business impact
We helped our client increase by 17 %the number of shows watched through recommendations through the refurbished RENG. The groups with the highest increase in recommendations watched also exhibited 11% lower churn rates than their counterparts, thus driving additional value for the telecom.
Related content
Want to know more?
Retention Analytics
Data-driven retention analytics reduces churn and retains customers cost-effectively.
