Case Study
IVR Optimization & Enhanced Customer Experience
Problem statement
A telecom and data provider faced the need to rationalize on call centre spend by reducing the workload on the agents. How can they proactively use customer data to understand their pain point, optimize IVR flow for better customer experience and automate the solution without the need for agent intervention – were achieved through Predictive Analytics Modeling
Tech Stack
GCP (Google cloud platform) for data handling
Python for data modeling
Solution
First, “Exploratory Data Analysis” was ran on existing customer data to understand the various reasons customers reach out to IVR and the segregate most prominent ones. Among them, “Troubleshooting” was the important category which needed agent intervention. Hence, a model was developed to confidently predict if the customer was calling for “Troubleshooting”. And if that was the case, then the first option in IVR was directed to Troubleshooting instead of the standard flow. Secondly, the granular data from customer’s phone was analyzed to predict the customer problem and thus redirect him to the required self-help section. Thereby, increasing the customer experience by reducing the waiting time – both in terms of navigating to the right flow and conversing with an agent. Also, this reduced the workload on the agents and thus optimized call centre spend