Zendesk taps machine learning for its Satisfaction Prediction tool

Machine learning has been making its way into a growing range of enterprise software in recent months, and on Wednesday Zendesk provided a fresh example: Satisfaction Prediction, a new tool designed to help companies catch potential customer-service problems before they happen.

Essentially, Satisfaction Prediction learns from historical customer-interaction data and uses that knowledge to predict which current customer-service interactions are likely to run into trouble.

The tool's knowledge base draws from an analysis of factors in past interactions that can precede customer dissatisfaction -- the amount of effort involved to solve a ticket, for example, and the language used -- along with the customer’s ultimate satisfaction rating at the time.

To read this article in full or to leave a comment, please click here