CAP Auto-Routing Improvements
A/B Experimentation for Contact Routing Optimization
The Problem
CAP auto-routing was at just 3% in Dec 2021 with 27% concession leakages. The system was WW-owned by CAP Data-service but IN-owned by CTAP's Smart Routing Recommender, creating a fragmented ownership model. Manual routing was causing unnecessary transfers and concession costs.
My Approach
I deep-dived manual routing contacts and identified opportunity areas — for example, 40% of 'Wrong Item' contacts had no routing rule at all. I created a business case for CTAP, then designed and enabled a staggered weblab-based launch with Control vs Treatment analysis for each new rule. This allowed us to measure impact incrementally and roll back if needed.
The Outcome
Launched 6 new auto-routing rules. Auto-routing improved from 28% to 45% (1200bps). 1.64M annualized transfers reduced. INR 250M in concession savings. Received the Just Do It award at AIM '24.
Why It Mattered
This project demonstrated that experimentation-driven operations improvement could deliver massive financial impact. The staggered approach became a template for how the team launches routing changes — test small, measure rigorously, scale what works.
