Numbers Questioned In September Article
I would like to comment on Natalia Scriabina and Sergiy Fomichov’s article “6 Ways To Benefit From Customer Complaints” (September 2005, p. 49).
The numbers in the statement “Just 4% of dissatisfied customers complain, while 96% go to competitors” are very appealing, but I am not sure this is true for all industries. For example, in the auto industry, do only 4% of dissatisfied customers complain that they need their car fixed? I don’t clearly see how the authors came up with these figures.
PSI Qualité & Sûreté De Fonctionnement
Misquoted Data Cause Questions, Confusion
I’m sorry to say there were mistakes in the article “6 Ways To Benefit From Consumer Complaints.” Under the second subheading, “Encourage Customers To Express Dissatisfaction,” the statement that only 4% of consumers complain is very wrong—the real number is about 50%.
I believe the error emanated from a misquote of Technical Assistance Research Programs’ (TARP’s) data. In the original data, we said 96% of consumers with a small ticket problem don’t complain to a company’s headquarters (HQ) but 50% do complain to the local retailer or return the product.
The information was misquoted when an author dropped the two qualifiers, “small ticket” and “to HQ,” in one of his books. This misquote has been picked up by at least one other ASQ publication and then by the British Standards Organization.
I present below a section of the article I wrote for the newsletter of ASQ’s Service Quality Division, Competitive Intelligence, in June 1999 to try and set the record straight:
- On average, across all industries, 50% of consumers will complain about a problem to a frontline person. In business to business environments, 75% of customers will complain to a frontline person. If this frontline person is an employee of a distributor or retailer, the chances are high that the problem will never be reported to the manufacturer or corporate office. This leads to the next finding.
- Only 1 to 5% of customers will escalate their complaint to a local manager or corporate HQ. For packaged goods and other small ticket items, TARP has found 96% of consumers either do not complain or complain to the retailer where they bought it. For large ticket items, the complaint rate is higher, rising to 50% to frontline and 5 to 10% of complainers escalating to local management or corporate. The existence of a toll free number at corporate HQ will, on average, double the number of complaints getting to corporate. However, only one out of 100 to 500 complaints will actually be addressed to a senior executive.
Complaint rates vary by type of problem. Problems that result in out of pocket monetary loss have high complaint rates (50 to 75%) while mistreatment, quality and incompetence problems evoke only 5 to 30% complaint rates to the frontline. For example, TARP recently found only 3% of consumers unhappy about their airline meal complained to anyone and they all complained to the flight attendant. No one complained to HQ or consumer affairs.
The next statement in Scriabina’s and Fomichov’s article is also wrong. It says 90% of customers whose complaints are resolved do not defect. That number is an average of two numbers published in TARP’s original White House Office of Consumer Affairs Study (survey report released in 1976 and final report in 1980) and applied to problems resolved quickly on first contact. Many complaints ultimately get resolved, but in many companies, the percentage resolved quickly on first contact is much lower.
The original TARP White House sponsored
study (see Figure 1, p. 8) recounted the impact of problem solving on loyalty.
I am concerned the article will further muddy the waters. Please highlight the correct data.
JOHN A. GOODMAN
John Scott, author of “ISO 9000 in Service: The Good, the Bad and the Ugly” in the September 2005 issue (p. 42), is director of total quality management for Palmetto GBA LLC, Columbia, SC.
In “The Future of Quality in Indianapolis” in the October 2005 issue (p. 39), the steps of the strategic planning process should be numbered one through four. A corrected version of the article is online at www.asq.org.