Natural Language Processing Virtual Assistant Improves Customer Satisfaction and Saves Money

Reducing call-in rates and improving customer experience for a Fortune 50 Telecommunications & Media Company

Keeping score

Keeping Score

A Fortune 50 Telecommunications & Media Company shifted their strategy to put an emphasis on improving their Net Promoter score. The current state of their customer troubleshooting and support engagement was largely handled by a blend of service calls, technician deployments, and website support; they found that when customers called into their customer support line they had more negative experiences and rated the company more poorly. In order to elevate the customer experience, they planned to develop a self-service platform that would serve as a single point of contact for all initial help and support cases.

Using a vendor NLP product, they set out to create a virtual assistant that could help their customers solve their problems and answer their questions quickly and efficiently on their own accord through a mix of articles, on-screen prompts, and instructional videos. By placing the power to self-address numerous issues in the hands of the customer, the company stood to reduce customer frustration, gain a better understanding of customer needs, reduce costs, and improve their Net Promoter score.

Talk to me

Talk to Me

CapTech was brought in to help manage the development of this NLP-driven virtual assistant and evaluate its performance. Partnering with an external NLP vendor, we worked with our client to develop use cases and treatment options for customers while monitoring and evaluating in-app behavior to continually refine and improve the customer experience.

We defined requirements and the way the chat product would work. As well as completing all of the configuration of the system and designing and writing the business logic. CapTech also developed requirements for metrics and reporting, NLP performance evaluation, data pipeline operations, and executive reporting.

Fewer Calls, Happy Customers

After successfully testing pilot program on different platforms, an iOS solution was launched. Since its release, the system has handled over 2 million queries were handled from 1 million + users on iOS.

Compared to Android users who had a 21% 24-hour call-in rate, users on iOS using the virtual assistant had a call-in rate of only 7% in comparison — a 66% difference. With this reduction, our client stands to realize cost-savings associated with customer deflection from traditional service resolution outlets.

The telecom company intends to use the data collected from the interactions as a baseline for customer service solutions and click-through success rates, with the success serving as a catalyst to develop an in-house NLP solution that covers all their platforms.

66% Reduction in call volume