F100 Media Broadcaster – Hyper Personalized Customer Experience Score (CES)
Business problem
Annual customer-initiated churn rate greater than 10% with discount-driven customer retention initiatives leading to a $1B+ financial overhead annually. The client wanted to hyper-personalize metrics for every customer to avoid being pressed into rolling out deep offers and for upstream decisioning through churn detection 2-4 months in the future.
Fractal solution
Principles:
Working closely with multiple client teams, Fractal developed a solution to:
Define value-based relationships and move away from offer-based retention
Provide an in-depth understanding of drivers to design customer-centric strategies
Design an easy way to implement the relationship score metric (0 - 100) to provide a holistic view of customer’s relationship
Fractal harmonized the customer view (Customer 360) from 25+ raw data sources and created an ensemble AI framework with end-to-end automation on the cloud
Development:
AI: Ensembled AI framework with superior performance and ability to leverage:
Static information like demographics, pricing, billing
Dynamic journeys, such as a sequence of past transactions/interactions
Engineering: Harmonized customer view (Customer 360) at weekly level, from 25+ raw data sources
Design: ‘Simple to measure’ metric, model explainability, and business actions
Testing & Operationalization:
Results validated on OOS & OOT samples and against existing baseline models
End-to-end automation on the cloud and incremental execution on a weekly basis
Integration with the marketing platform (through the on-premise database) to ensure regular automated feeds of customer responses, lists, and drivers
Innovation
-
First customer management solution on cloud for the organization resulting in accelerated migration from on-premise data and technology platforms
-
Simultaneously captured dynamic and static behavioral patterns
-
Self-learning models enabled adjustment over time to account for change in customer behavior
-
Process automation, parallelized data processing and modeling, resulting in efficiencies at scale and delivered a time reduction of 20X compared to traditional frameworks
-
CES is a dynamic indicator of the experience customer is having with the business throughout the relationship
Insights
40% of drivers discovered by the model were completely new/ unrelated to offers and provided incremental intelligence about:
DTV and broadband product holdings
Billings/Pricing, marketing outreach
Engineered 800+ granular features to identify early signals of poor experience and churn-risk behavior across 150+ customer segments
Established the hypothesis that experience is defined not only by the current state but by the overall journey the customer has been through with the organization
Business outcomes & impact delivered
40% of drivers discovered were new/unrelated to offers and provided incremental intelligence about - DTV and broadband product holdings and Billings/Pricing, marketing outreach
Engineered 800+ granular features to identify early signals of churn-risk behavior across 150+ customer segments. Established the hypothesis that experience is defined not only by the current state but by the overall journey
Call reduction: 7-10% reduction in the total annual volume of cancellation calls. Churn reduction: 5-6% reduction in the total annual volume of churn
Revenue-loss prevention: Arresting average drop in subscription revenue per customer/quarter, leading to $40M+ in value savings in the first year
Cost reduction: $20M+ savings due to the reduction in cancellation calls and churn prevention activities
Looking for a similar project?
Google Cloud services
Big Query
Compute Engine
Cloud Storage
AI Platform Notebook
Fractal's value proposition
