Customer Recovery and its Role in Reducing Customer Churn

According to a recent survey, 56% of customers would churn after a single negative customer service experience. That is a damning statistic and shows just how important it is to keep churn to a minimum, after all, increasing customer retention rates by even 5% can improve profitability by 25-95%.

Which leads to an important question, how do you reduce customer churn and increase customer retention?

One way of dealing with such a dilemma is by improving Customer Recovery.

Customer recovery is the process of transforming unhappy customers who have had a negative or unsatisfactory service experience during customer service into a positive and trustworthy experience. In brief, sometimes customers get dissatisfied and frustrated with the service or assistance they get through agents for their problems during customer service.

In such cases, agents sense the annoyance of the customers and make them happy by using all kinds of help and various service strategies to solve their problems. This whole process of turning an angry customer into a satisfied customer is called 'Customer Recovery'. It is crucial because it allows organizations to match customer expectations and prevent them from potentially churning.

In the long run, customer service recovery can positively influence customer retention, satisfaction, brand reputation, word-of-mouth behavior, and customer loyalty. And while most companies focus more on customer acquisition than customer retention, Harvard Business^[1] says getting a new customer is five to twenty-five times more expensive than retaining an existing customer.

Therefore when existing customers are dissatisfied, it is important to respond to them as soon as possible, identify errors and correct them. Otherwise, it will lead to loss of customers or risk of escalation.

We have two main types of customer recoveries:

(i) Recover already lost customers

(ii) Recover the customers instantly from the poor customer experience during the conversation.

This article will discuss the second type of recovery and why it's essential for every business.

Customer service recovery paradox?

The customer service recovery paradox [2] is when a customer thinks more highly of our business once the issue has been fixed. This paradox means that even though the customer had terrible experiences with customer service, if we recover that situation well enough, customers will be more loyal than they ever were before.

Consider the below graph (Fig 1) showing that customers who have experienced hostile service and successful recovery will be more reliable over time than those who have not experienced service failure. Therefore, customer recovery can have a very positive impact on the business.

‍

Is it difficult to detect customer recovery?

There are two significant challenges in customer recovery. One of them is "solving the problems of a dissatisfied customer through good customer service". For this, we need to train agents with different customer service strategies to correct customer issues and earn customer loyalty. The second is to check for negative experience conversations about whether agents solve customer problems and make them happy. The second one is a bit tough compared to the first challenge as we have to look at the whole conversation to check "Customer Recovery".

Let us discuss this challenge with an example. Suppose a company gets a volume of 10000 tickets per month. Finding customer service failure tickets in those 10000 tickets and locating successfully recovered tickets in those found tickets is a very time-consuming and tedious process. For this purpose, we have developed an AI / ML algorithm that automatically detects the customer's negative experience and verifies whether the ticket has been successfully recovered or not.

How does our model work?

To detect customer recovery, we have developed an AI model with two main components:

(i) Customer Sentiment and Emotion Models

(ii) Recovery Strategy

Below is a described model with a diagram and steps:

‍

• Our AI/ML Recovery model works on all types of tickets regardless of ticket type voice ticket, chat/email ticket, or hybrid ticket (combination of voice and text).
• Our model takes every customer comment in the ticket and then deeply analyzes the different customer emotions such as negativity, frustration, abuse, anger, negative sentiment, and happiness.
• To identify each of the customer emotions mentioned above, we have trained transformers-based, semantic textual similarity-based deep learning (DL) models with large amounts of real-time customer interaction data. We have fused syntactic-semantic-based keyword models along with DL models for better results.
• We have added our own effective recovery strategy on top of the customer emotion models.
• Our model assesses whether customer recovery has taken place with the help of customer negative and positive emotions and the recovery approach.

Conclusion

Detecting and analyzing Customer Recovery is a key metric in CX analytics that should not be ignored. It helps organizations identify the most effective agents in their contact center, draw key learnings from successful ticket resolutions and use that to improve agent performance.

In turn, this can reduce customer churn as more users end up having better interactions with agents.

References:

[1] https://hbr.org/2014/10/the-value-of-keeping-the-right-customers

[2] Matos, Celso & Henrique, Jorge & Rossi, Carlos. (2007). Service Recovery Paradox: A Meta-Analysis. Journal of Service Research - J SERV RES. 10. 60-77. 10.1177/1094670507303012.