Loyalty expert Frederick F. Reichheld’s article in the Harvard Business Review, The One Number You Need to Grow (December 2003), explores issues around measuring customer satisfaction. Reichheld’s research, linking a customer’s willingness to recommend a firm with its revenue growth over three years, enumerates what many executives have intuitively known for many years: the power of positive word of mouth. The research seems compelling, and as an advocate of customer-focused organizations and measurement of customer experience, I am supportive of anything that reinforces my beliefs.
In a world where the reach and effectiveness of advertising is diminishing, recommendations take on greater importance. Research by the Future Foundation shows the shifting influence of recommendations vis-à-vis advertising.
If the trend continues and experience grows in importance as a vehicle for generating business, organizations will have to pay more attention to building experiences that inspire recommendations. There is little doubt, therefore, that the subject of the article is of real importance to many organizations, but it should come with a health warning—it is a double-edged sword. The problem with Reichheld’s article lies in the literal interpretation that some companies put on it, thinking that they need only one question in their survey. This is just downright wrong and is not what it says.
Why gather feedback?
There are two primary reasons for investing the time and money to measure the customer experience: to track performance and to improve performance.
The article centers on the relationship between the number of people willing to recommend an organization and the organization’s revenue growth. It is a useful way for management to track the performance of the organization and may correlate to some aspects of financial performance. Management information and tracking is, however, only one purpose of feedback and, I would argue, a secondary reason.
I believe that there is a causal relationship between advocacy and financial performance, but one number does not prove that. Correlation and causal relationships are very different. To know something is a causal relationship, I have to isolate other variables that might affect the outcome. A telecoms supplier once reported research that showed that cat owners made the most phone calls. Correlation, yes; causal relationship, doubtful.
The success of an organization is a complex issue and rarely the result of one factor. Growth might be the result of higher advertising spend, an under-supplied market, a patented product or innovation in distribution. This is, of course, reflected in the customer’s perception of the company, but tracking one number does not say why the customer would recommend the company. Knowing how you are doing—tracking performance—is less important than knowing why you are doing—managing performance.
How would you feel if an airline pilot announced that on your next flight, she was going to track only altitude? Well I’d feel good that she knew we were not too close to the ground but a little concerned that she wasn’t watching how much fuel was available or the speed and direction we were traveling. Complex systems require multiple measures to track them. The key is having an effective summary and access to the detail when appropriate.
It is also important to understand that insight usually comes from putting different data together and looking at any relationships. I would argue for having more numbers (within reason), not one.
I believe that Reichheld is right in focusing on advocacy rather than satisfaction. Satisfaction is much measured but a poor indication of performance. To test the value of satisfaction, the next time your spouse or friend cooks you a meal and asks, "How was it?" say you were "satisfied," or even "very satisfied." Be sure to duck when the frying pan flies at you. People are not impressed with satisfaction, although it is still the most common question in customer surveys.
Remember also that, as the article points out, willingness to recommend is not always a good correlate of growth. In some industries—typically business-to-business environments where choice of supplier is constrained—it can be a weak correlate.
The primary reason for gathering feedback is to identify and prioritize the improvements that will increase advocacy and repurchase intention. Most recommendations are the result of positive experiences, so improving the experience is key to improving performance. The real correlate of interest, therefore, is that between advocacy/repurchase intention and attributes of the experience. Read my ideas about the proper way to gather feedback.
Good feedback stimulates the improvements that drive up a customer’s willingness to recommend. Any sports coach will tell you that the only way to change the score is to improve the way the team performs and plays the game. You don’t change the score by looking at the scoreboard. In a complex, multi-faceted customer relationship, understanding how each part of the team is performing is vital. Companies need to understand their strengths and weaknesses from the customer’s perspective.
Enterprise Rent-a-Car (the inspiration for the work described in Reichheld’s article) suggests that this information is best collected by a text question asking about the quality of the rental experience. While this will provide information, capturing this by an open text question alone (as the article suggests) is problematic. Analysis is time-consuming and provides no view of the scale of problems across the organization. This can be achieved by categorizing and coding the responses, but this adds time and cost to the feedback process. A second problem with open questions is that they produce the lowest response rates of any question type. Even in phone surveys, the likelihood of getting a meaningful comment is low.
The people who will readily provide a comment are those whose experience has been either very good or very bad and in this respect, Enterprise’s approach is consistent with its focus on those with a high propensity to recommend. It does not, however, provide the means to readily identify the priorities for improvement.
I want to build an organization where each and every one of our customers sings our praises. I know I can do that only by knowing what matters to all of them and how well we deliver. With this information spread across the organization, I can keep staff focused on continuously driving up the quality of the experience we deliver and, thereby, impress customers enough to get them to continue to recommend and repurchase. For me, one number is not enough.