We started by identifying target outcomes such as increased revenue growth, improved customer satisfaction, and reduced costs. Next, we selected target metrics, such as call volumes per day, conversion rates, average order size, and outbound revenue per day, to measure progress toward the target outcomes.
The company used insights from the metrics data to create models for NPS improvement, revenue improvement, call and revenue forecasting, and capacity planning. This included:
- ML-driven bots analyzing complex data (~40 parameters relating to calls and chat) to make recommendations to improve performance
- A personalized ML-based prediction model that identifies dissatisfied customers and detractors on 100% of calls to improve the customer experience
- A call forecasting model to map call volumes to factors such as marketing campaigns, new product launches, and seasonality
- A revenue forecast model to estimate conversion rates
- Capacity planning modeling-aligned customer support with demand
Finally, we implemented a centralized tracking system to track callbacks and resolution, enabling the client to monitor this component of call conversions.
Additionally, a smart-scheduling system helped align the right skills within teams and links top-performing teams with the prime shifts or peak loads.
The result is an intelligent operating model in which business processes could now sense the environment, act appropriately, and continuously learn from the effectiveness of those actions at scale.