Blog Home

Analyze Customer Churn with Speech Analytics


The Team at CallMiner

May 29, 2018

Happy young woman working from home and talking to a customer
Happy young woman working from home and talking to a customer

In today’s business environment, high customer expectations leave many contact centers struggling to understand why customers leave their service or churn. More often than not, these businesses focus their time and energy on acquiring new customers rather than trying to retain existing ones. A decision that costs them significantly since attracting new customers costs seven to ten times more than it does to keep an existing one. Pair that with the fact that 68% of customers stop working with a company because of a bad experience, costing business $62 billion each year, and it is important companies find ways to prevent customer churn.

Solving the Problem

According to Aberdeen Research, best-in-class companies have a retention rate of 85%. Two out of three of these companies use performance measurement to better understand their customers. An area for customer contact measurement is in the contact center.

Speech analytics is a popular solution used by contact centers to track calls, uncover trends, and make changes where needed to improve operations. The analytics collects raw data from every conversation type including phone calls, text, emails, and live chat and transcribes it into a searchable format by tagging it with metadata. Contact center managers can use this information to uncover various trends including reasons for customer dissatisfaction, compliance risks, manufacturer defects, and problem agents. Speech analytics can also identify customers that are most likely to stop doing business with you or churn.

Predicting Customer Churn

Without speech analytics solutions, businesses usually don’t know a customer is at risk for churning until it is too late. By the time they offer a new solution, discounted rates, or an apology, the customer has already begun their transition phase to a new company. Speech analytics can listen for various churn signals on phone calls including keywords or phrases or voice pitch in addition to calculating a churn risk score automatically.

One of the key features of speech analytics is the automated call scoring that listens to 100% of conversations and grades them based on criteria established by managers. Calls are typically reviewed for agent quality, compliance, risk, customer satisfaction, and churn potential. To track customer churn potential, contact centers listen for key phrases including cancel, unsubscribe, remove, stop my service, too expensive, cheaper option, and very unhappy. When these words are mentioned, the call can be graded appropriately.

These keywords can also be set to trigger notices be sent to managers in real-time letting them know an at-risk customer is on the phone. Managers can see who the customer is as well as who the agent is, and decide if they need to intervene or not.

Contact center agents can access these scores in real-time, giving them the opportunity to self-improve on an ongoing basis. Managers can also access these scores to better understand which agents need more assistance and which customers need more attention. If customers are at risk for churn, managers can assign agents to reach out to them using churn prevention procedure or offerings to try and save their business.

Who can benefit from speech analytics?

A healthcare insurance provider implemented customer churn predictions with the help of professional services experts MainTrax and followed up with those at risk with a personal phone call. The results? The provider was able to save 96% of customers that were considered a churn risk by being able to act quickly.

Insurance providers aren’t the only companies that can benefit from using speech analytics to understand customer churn.

  • Telecommunication companies
  • Energy services provider

Without speech analytics, the telco and energy service providers won’t know before a customer calls to cancel that they were a churn risk. But, with analytics, they can listen for certain keywords or phrases including “my bill is too high” or “do you have any plans that cost less?” When these types of questions are asked, it triggers a notice be sent to the supervisor, who can send the agent a new plan, or notify the agent instantly with options to offer on their screen.

Every contact center is expected to deliver a great customer experience, or else they risk customer churn. To uncover the root cause for ongoing customer churn, contact centers can utilize speech analytics. By establishing the right keywords, phrases, and triggers, businesses can reveal reasons for churn, make changes to operations when necessary, and personally reach out to at-risk customers before it is too late.

Learn more about reducing churn and the complete MainTrax case study in our whitepaper, Reduce Churn and Increase Customer Satisfaction with Speech Analytics.

Contact Center Operations Customer Experience Speech & Conversation Analytics North America Utilities Industry