Challenge

Our customer's data science team had constructed numerous churn prediction models but were unable to identify the root causes.

Increasing the call center team size did not solve the problem. Further, the call center managers were responsible for grading their team's performance, leading to misrepresentative analysis of call center performance. With churn impacting the business revenue and with no solution in sight, our client needed a different approach to stop losing customers.

Our Approach

We trained our Al agent listened to every customer call, transcribing them in real time and extracting all relevant information such as the reason for the call, the customer churn risk, and the call resolution status. Our Al agent then analyzed the customer sentiment on the call, scored the call quality relative to the call center script, and graded the agent quality based on their performance scorecard to provide specific feedback to the agent. All the data was entered into a centralized database that our customer could use to query with Al for tangible insights, allowing for all churn drivers to be identified in real time and providing call center agents with clear instructions on how to most effectively resolve issues when dealing with clients. The Al agent helped to reduce our customer's time to reach a resolution from 11 minutes per client to 2, increasing client satisfaction and reducing churn