The Cost of Bad Customer Service

(and the need for back-end service recovery)

Neet Shah

Nathan Johnson

While most companies recognize the importance of good customer service, they are often not aware of the financial benefits of good customer service and the potential costs of a service failure.

The definition of a service failure is simple – any customer service interaction that results in a dissatisfied customer. There could be many reasons for dissatisfaction. Some customer dissatisfaction is due to control issues at the service center such as long hold times, a customer service representative’s lack of knowledge, or accent and language barriers. Other reasons not under the service center’s control may include dissatisfaction with the company’s product, company policies, forces of nature, and more.

Today, many large companies have giant contact center networks that comprise tens of thousands of customer service agents, located locally and offshore, dealing with millions of customer contacts per year. Servicing such a large number of customers will inevitably lead to some service failures.

While most contact center managers are tasked with improving customer satisfaction, they face significant constraints:

Customer service centers are viewed as cost centers.   

Company management often requires costs to be held constant despite growing servicing volumes.

Finding qualified agents is difficult.

In recent years, attrition has been high – as average tenures have declined, the quality of customer service agents has declined.

Legacy technology systems (“tech debt”) are a challenge.

Important improvements often require significant technology investment to replace legacy systems.

In this environment, investing large amounts in preventing every potential failure upfront is not cost- effective. However, our analysis indicates that in many cases, aggressive back-end recovery may be a less expensive and more effective strategy.

Further, Bitner et al. [1] have shown that it is not necessarily the failure itself that leads to customer dissatisfaction, but rather the lack of a quick response to a service malfunction. Companies that have internalized this insight have seen great benefit from the implementation of a service recovery mechanism.

Starbucks is a good example. Baristas are empowered to solve almost any situation leading to a dissatisfied customer by such means as exchanging the product or refunding the payment. By so doing, Starbucks assures that any failure is addressed as soon as possible and that almost no customer leaves a Starbucks branch dissatisfied.


Every so often, a prominent company suffers from a substantial public service failure. United Airlines observed just such a failure in 2017 when security guards forcibly removed a passenger from an overbooked flight. The incident occurred not in response to hostile behavior, but to the passenger’s refusal to surrender his seat for a UA employee. The altercation left him with injuries including a concussion, missing teeth, and a broken nose requiring medical intervention to restore his breathing ability.

An offense of this magnitude would reasonably merit generous efforts to repair public relations, but the airline’s initial reaction to the highly-publicized incident did little to restore their sullied reputation. Early responses from United placed blame on the passenger for being uncompliant. Immediate public outcry and a continued stream of negative press urged the company to go further, including a private settlement with the accosted passenger, refunded tickets for the entire flight, and an amended policy on over-booking flights. Additionally, company representatives stated that use of law enforcement to remove passengers would no longer be allowed unless in matters of safety or security.


In the United Airlines case the service meltdown was a large, one-time event and hence received a lot of attention from the public and the company’s management. However, customers all-too-frequently experience similar frustration and dissatisfaction – albeit on a much smaller scale – on an ongoing basis. These service failures are caused by day-to-day service issues that don’t receive as much publicity. The damage caused to a company from this type of dissatisfaction varies by the service failure type. Some examples include:

Manpower/Infrastructural Related Issues

  • Long hold time
  • Representative accent (due to offshore servicing)
  • Slow fulfillment

Customer Service Policies

  • Dissatisfaction with service policies, e.g., lack of representative empowerment
  • Representatives cross-selling additional products during a service call 

Incompetent Service

  • Conflicting information
  • Multiple calls to resolve issue
  • Multiple transfers between representatives
  • Representative lacking knowledge
  • Lack of follow-up
  • Rude service 

Dissatisfaction from the Company's Product

  • General dissatisfaction with product
  • Damaged product


While the ratio varies by industry and type of service requested, our experience (even with companies that are perceived to have superior customer service) is that 10-20% of all customer service calls result in a dissatisfied customer. Today, the systems that companies have in place typically detect only a small portion of these service failures. Undetected (and thus untreated) failures lead to customer disengagement and often, customer attrition.

Figure 1 – Probability of a customer experiencing at least one undetected service failure during a year

For large companies like cellular service providers, financial institutions, airlines and more, the average customer usually contacts customer service several times a year. The customer is at risk of experiencing a service failure in each of these calls. In most cases, one service failure in one of these calls is enough to cause dissatisfaction and eventually lead to a disengaged customer. As illustrated in the above figure, for customers calling more than 5 times each year, the probability of experiencing at least one undetected service failure during a one year period may be higher than 50% – clearly an unacceptable situation.


While the probability of a customer experiencing service failures is high, it is important to understand that not all service failures are equal as far as impact on satisfaction is concerned. When creating a service recovery methodology, a company needs to look at the overall potential negative impact that a specific service failure can have on customer satisfaction. This is governed by both the frequency of the service failure and severity of disengagement caused. 


Figure 2 – Frequency and impact on engagement level by service failure type. 


Companies tend to focus efforts primarily on prevention. While this is cost-effective for high-frequency failures, it becomes increasingly expensive to fix less frequent failures. Further, as overall failure rates decline, the marginal cost of improvement climbs steadily.

As companies get closer to 100% satisfaction, back-end recovery can often be more cost effective than upfront prevention. The main steps for successful service recovery are:

  1. Acknowledge – Recognize that a failure has occurred and acknowledge it
  2. Apologize – and explain the reason for the failure
  3. Solve – Rectify the situation or escalate to someone who can solve
  4. Make good – Depending on the severity of failure, provide compensation to show the customer how much the company values the relationship

We have observed that existing processes and systems aren’t always aligned with implementing service recovery. Some key changes required include:

  1. Appropriate incentive structures: Call center representatives often have a disincentive to escalate problems. Similarly, they are often rated on how quickly they can complete a call. Neither are aligned with ensuring effective service recovery
  2. Highly qualified recovery team: A group of highly empowered and highly experienced representatives with the ability and authority to solve and make good.
  3. Good detection: Clear instructions to representatives defining situations where escalation is required (voice auto-detection may be an option).
  4. Limiting “make goods”: Clear rules on when reps should offer “make goods” to prevent overuse.
  5. Differentiated treatment: Varying treatment based on customer value and severity of failure.

Most companies tend to under-invest in service recovery. Complementing prevention efforts with appropriate recovery processes can lead to a differentiated customer experience, particularly during critical customer interactions. Further, our experience indicates that implementing a well thought out service recovery process pays for itself very quickly through reduced attrition and increased engagement. 

Unsure of how to translate improved customer experience into top line and bottom line impact? Let’s discuss how we can exchange complexity for clarity.


[1] Bitner, M.J., Booms, B.H. and Tetreault, M.S. (1990), “The service encounter: diagnosing favorable and unfavorable incidents”, Journal of Marketing, Vol. 54 No. 1, pp. 71-84.