The Sawyer Partnership logo  
Quotes & Tips

Consulting, training, and coaching to help you solve messy problems that disrupt your operations or cause customers to take their business elsewhere.

  About The Sawyer Partnership

Solving the Right Problem:   Using Metrics Effectively


Jeanne Sawyer

Choosing and Using Metrics

Building Dynamic Partnerships



The message was unmistakable. A senior executive of a large department store chain calls a senior executive of the computer company that sells the store an average of $8M worth of equipment every year. The message is not what a vendor executive ever wants to hear: the store is experiencing too many failures, the computer company's latest products are junk, and the executive is personally sick of the fruitless and time-consuming efforts to fix the problem. A certain amount of profanity was used.

This true story demonstrates some important concepts in creating effective customer services. Let's analyze what was behind the phone call and what the computer company did to solve the problem.

The store executive knows that he is unable to provide service to his customers, and according to his staff, the reason is computer failure. There have been repeated efforts to fix the problem, but nothing has worked. He also knows the holiday shopping season is coming. Angry and frustrated, he finally makes the angry phone call described above.

Meanwhile, the service engineer is also increasingly frustrated by his inability to fix the problem. He has done everything he or the corporate experts can think of. Morale is sinking. Since the engineer is aware that he is repeatedly replacing the same parts, he feels that his own company is undermining him by providing what appear to be unreliable spares.

This situation exhibits a common tendency of customers and vendors alike: blame the product. Product defects often are the root cause of problems, but it can be dangerous to assume that. Especially in complex environments such as the store's data center, there are so many potential causes that making this kind of assumption is likely to result in solutions to the wrong problem—thus customer problems that keep coming back.

In this example, if the computer company had continued to assume the problem was defective product, more resources from both the company and its customer would have been wasted, the customer would have continued to experience failures, losing revenue and becoming increasingly dissatisfied. The customer would certainly have delayed additional purchases, and ultimately might have moved to another platform.

The solution lies in using metrics effectively and building a relationship with customers based on trust and shared responsibility.

Choosing and Using Metrics

Companies today routinely rely on metrics in two key areas: to track financial results and to measure manufacturing processes. However, using metrics effectively in the field to provide customer service has been more elusive.

The first question was what metric to use. The decision is not trivial, and must indicate success and failure to both the customer and the vendor. In this example, the obvious choice appeared to be system crashes. However, by the time a system crash occurs, it is too late to predict and avoid the problem. The customer and computer company wanted a leading indicator, so decided to use unscheduled service call rate to measure overall system and service performance. The choice was driven by the following characteristics of the metric:

  • Easy to understand as an indicator of trouble. If the customer has to place a service call, that's bad.
  • Easy to collect. Service calls were already being logged. Using them for this purpose required new analytical tools, but the base data existed and was routinely collected.
  • Effective driver of improvement for both customer and vendor. Service calls are expensive for both parties, so reducing the rate is to everyone's advantage.

The top level metric, overall service call rate, acts like a thermometer, indicating the general health of the account but not providing a diagnosis. If the overall rate is too high as compared to other accounts, then additional data is collected to help pinpoint the cause of failure.

The call rate graph for the store's system is shown in the following graph. It shows that for a period of time, the call rate was quite stable. Then something changed causing the call rate to skyrocket and lose stability.

Graph: Service Call Rate Trend

This type of graph enabled the computer company to verify that, indeed the store was experiencing an unwarranted number of hardware failures, but also to highlight several other key pieces of evidence:

  • The failure rate was much higher than experienced by other customers,
  • Only certain systems owned by the store were failing,
  • The failing systems were located in the same building,
  • The same components were replaced multiple times at that location, with new components failing rapidly after installation.

The data told the computer company that something unusual was happening at the site, and suggested a probable cause in the environment or customer/field processes rather than in the product itself. It would be impossible, however, to use this information effectively unless the company's relationship with its customers would allow these possibilities to be explored.

Indeed, at this point the computer company executive presented these results at a progress meeting to senior store executives. After the meeting, one of the store officials commented, "That was the best presentation we've ever seen that says we have no clue as to what the source of the problem is." The reason everyone was pleased with progress, despite not yet having a solution, was because possibilities were being systematically eliminated based on facts rather than guesses—and the logic of the investigation was open to everyone. The process was systematic, clear and repeatable.

Building Dynamic Partnerships

An effective partnership is a dynamic working relationship. It is based on trust and mutual regard, which are earned with demonstrated honesty, integrity, and commitment to shared success. This is a very different model than the traditional approach where the customer presents a list of demands and the vendor responds.

In a successful dynamic partnership, both the customer and vendor accept responsibility for the things in their purview. In a partnership, "the customer is always right" is an ineffective model because it closes communication and prohibits the partnership from fully investigating a problem and learning the truth.

If the computer company in our example had already had this type of relationship with its customer, the angry phone call probably would never have occurred. However, although belated, the company was able to change the relationship—initially through intervention by an executive from the company who personally had a sound relationship with the store, and later by the account team using clearly defined methods to sustain the relationship.


The computer company invited the store to participate in a joint team to analyze the data, identify possible root causes and verify them, and develop and implement a mutually agreed upon action plan to eliminate verified root causes. As the team began working, the relationship started to change from adversarial to one of true partnership. The focus became the problem to be solved rather than assigning blame.

Working from the data, the customer/vendor team was able to isolate the symptoms and to identify possible causes. Key information supplied by different team members led to calling in an independent grounding specialist. Together, the team and the specialist were able to determine that the cause was not the product. Rather, it was current circulating through the systems from the local telephone company's T1 installation. The problem was further exacerbated by the janitor's using polyester dusters—when the janitors dusted, systems failed. Each team member had different bits of information that had to be put together to solve the problem. Neither the customer nor the vendor could have successfully solved the problem alone.

Because the store had participated in the entire analysis process that identified the cause, the store quickly fixed the problem—without wasting additional time arguing, and without embarrassment or anger. The call rate graph demonstrates what happened when the problem was solved: the call rate dropped and stabilized immediately.

The customer is now entirely satisfied, not only with the system performance, but with the entire way the relationship works. The computer company has also seen an improvement in morale, with account team members happy to participate in a productive relationship rather than feeling abused by both the customer and the home office.

From this one customer, the computer company has seen at least $20M in revenue that otherwise would have been delayed and possibly lost altogether. In addition, productivity savings are estimated at several hundred thousand dollars resulting from eliminating the destruction of parts, eliminating service calls and reducing the amount executive attention required.


Since then, the computer company has standardized the process and has demonstrated similar success with over 100 customers worldwide. The standard method includes specified steps for profiling system performance (using unscheduled service calls), analyzing the performance and identifying risks, analyzing root causes, and developing and implementing an action plan that eliminates the root causes.

Every company has a relationship with its customers, either by default or by design. Both a company and its customers can achieve significant financial advantages by:

  • Explicitly designing their relationship,
  • Basing it on mutually beneficial, measurable service objectives,
  • Creating and using defined procedures for working with customers.

A company's service relationship with its customers is too important to leave to chance.



Solving Problems PermanentlySM
tel. 408-929-3622

Privacy Policy

Copyright 2010.  The Sawyer Partnership.  All rights reserved.
Jeanne Sawyer, Ph.D.