Can the effect of word-of-mouth be measured?

Reading Guy Kawasaki's post on the Art of Customer Service part II and the comments in particular, made me think of the effect that word-of-mouth has in the context of customer service. Some comments highlight the frustrations of bad customer service and the very fact of these experiences being shared, the phenomenon of word-of-mouth.

Word-of-mouth is an elusive ally, but can have a potentially devastating effect on your company. The truth is that people are more likely to tell their friends they've had a bad experience than an OK, or acceptable one. No news it seems are good news. Frederick Reichheld, of the Harvard Business Review has ventured into trying to define or even calculate the effect of word-of-mouth using what he calls the 'Net Promoter Score" in an article published in 2003, titled  “The One Number You Need to Grow”. Reichheld argues that most of the effort survey research companies expend to measure customer satisfaction and loyalty is wasted because they ask the wrong questions. Instead, he boldly asserts that the only question that needs to be asked is: “Would you recommend this company to a friend?”

Reichheld argues that this response should be used to calculate the number of “net promoters”a firm has. Basically, he recommends asking the question on a scale of zero to ten, where ten means “extremely likely” and zero means “not at all likely.” He then calculates the proportion of brand users rating their likelihood of recommending it as a nine or ten (referred to as “promoters”) and subtracts from it the proportion rating their likelihood a zero through six (referred to as “detractors”). And presto, one now has all the information needed from customers to grow a business – the proportion of “net promoters.”

On the surface of it the simplicity of this idea is very appealing. It seems that all the complicated customer surveys are a waste of time, when all you need to know is the answer to this one question.
Ipsos Ideas Canada however, pose some very poignant questions contradicting this theory:

There are numerous problems with Reichheld’s proposed calculation. First, the implication that all customer satisfaction and loyalty measurement systems fail to correlate with profits or growth is ridiculous..For example, research has found that customer satisfaction has a measurable impact on purchase intentions, on customer retention, on financial performance, and on word-of-mouth (also known as the likelihood of recommending). Furthermore, researchers have found that companies that perform relatively better on the University of Michigan’s American Customer Satisfaction Index (ACSI) produce significantly more Market Value Added, a cumulative measure of corporate performance reflecting the difference between what investors have put in and what they can take out. This documented relationship between stock market performance and firm growth contradicts Reichheld’s assertions.

As a designer and a product developer, I'm actually far more interested in knowing what are the reasons behind a customer giving a company a grade 4 as opposed to say, grade 8 on the scale of 1(abysmal) to 10 (brilliant). As Ipsos puts it: "Instead of being a guide to action, it is simply a thermometer. Imagine a doctor taking your child’s temperature, announcing that he has a high fever, and saying, “He’s definitely ill, good luck.” Only by understanding what actually drives customer loyalty can managers effectively prioritize their improvement efforts, and this requires a great deal more diagnostic information from customers than simply their likelihood to recommend a firm to a friend."

What really confuses me in his article is that Reichheld claims that the “net promoter”score is independent of other measures (satisfaction and loyalty). Surely overall satisfaction levels + the likelihood to repurchase + the likelihood to recommend the firm should equal the final net promoter score? Moreover, in terms of improving something it may be vitally important to know that people rated the service they got, but found that the product was substandard and hence they would not recommend the company to others? If you only know the result, i.e they will not recommend the product to their friends, you will see the fact in a balance sheet and there is no need to ask customers about this. If, on the other hand, you did find out that it is the product's fault that they are not willing to recommend you to others, then you at least know that your customer service is in order and you need to sort your product out.

Going back to the original question, can word-of-mouth be measured - the crux is that loyal customers don’t always act as advocates for brands, services, and companies. So if the presence of word-of-mouth is actually more constrained than conventionally thought, how exactly does the “net promoter” concept work? Besides, recommending things to a friend you tend to mention those things in the right context, rather than bang on about them whenever and where ever, that is: if you want to have some friends left. Many companies and products we are entirely happy with, we trust them to deliver the product or service we expect, but we are not evangelical - in fact, I don't think you will ever catch me being evangelical about toilet paper, deodorant, washing powder, electricity and gas providers etc. but you will find me ranting on about computer interface design, product design on various consumer products, certain airlines, restaurants, and so on.

Why is that? I believe we are more likely to be evangelical or recommend something to a friend the better the experience it provided. Therefore I will return with a checklist for whether your company provides a commodity or an experience and if it does provide an experience, is it one people would recommend? More on that later.

Comments

[J]

Hey C,

thanks for your comments on the blog - very much appreciated! (hugs) Can't wait to hear what you come up with next! - your previous mixes are still in rotation in iTunes... :-)

Happy Easter! - promise to cheer up... :-)