What are our customers thinking?
Do we know which customers are likely to churn?
How can we stop customers from going to the competition?
All customer support organizations should be asking these questions. They should be thinking about their teams’ influence on customer loyalty and brand perception. Because their impact is huge: When it takes customers too long to reach a representative, when they have to call repeatedly to get a resolution, or when it takes a long time or repeated back-and-forth interactions to get an answer, satisfaction will be low and those customers are at risk.
Most companies measure customer satisfaction, and while it’s a very important metric, it looks backward at events that have already happened (“How would you rate the customer service you received?”). But what if you could predict in the moment how likely a ticket is to receive a good or bad rating from the customer, allowing your agents to take action to ensure a positive outcome?
That’s the intent behind Zendesk’s new Satisfaction Prediction engine, which applies machine learning and predictive analytics to determine whether customers are at risk of churn, prioritize routing based on customer risk, and guide agents to handle interactions more effectively.
In a recent report, Aphrodite Brinsmead of Ovum outlines why customer satisfaction scores are so important, and how Zendesk’s Satisfaction Prediction marks a change in the way analytics will be packaged and sold.