How is it possible that customers are more loyal after failures of products or services than they have been before? Excellent service recovery is the key and with the right activities, companies can fully utilize the service recovery paradox.

The “service recovery paradox” states that with a highly effective service recovery, a service or product failure offers a chance to achieve higher satisfaction ratings from customers than if the failure had never happened. A little bit less academically, this means that a good recovery can turn angry and frustrated customers into loyal customers. In fact it can create even more goodwill than if things had gone smoothly in the first place.

Nevertheless not all service recovery efforts will lead to increased satisfaction ratings as several studies have already shown. The key is to understand that there are certain situations when it is highly likely that a service recovery will lead to increased customer satisfaction. Services recoveries that are likely to be efficient are obviously those where the service failure is perceived to be not systematic or that the company had little control over it. But even in cases when it was a systematic failure and the company had control over the failure there is benefit for when service recovery activities are put into action to ensure that one can win-back customer’s and the source of failure is eliminated.

The key question is this: Are you aware when your customers encounter service failures? Have you thought about an “emergency plan” that can be put in action whenever your customers encounter a service failure? Or do you plan to take ad-hoc action when customers end their business relationship with you?

Read more about the service recovery paradox in these publications:

VP Magnini et. al., “The service recovery paradox: justifiable theory or smoldering myth?,” Journal of Services Marketing 21, no. 3 (2007): 213-225.

CA de Matos, JL Henrique, und C Alberto Vargas Rossi, “Service Recovery Paradox: A Meta-Analysis,” Journal of Service Research 10, no. 1 (2007): 60.