Net Promoter Score is BS due to sampling bias. When a subset of a population is used to extrapolate generalizations about the entire population it is wise to make sure that your sample population properly represents the entire population. If your sample does not represent the population that you wish to understand more about than the data is useless. I don’t have concrete evidence to place before you which indicates sampling bias in NPS, but my intuition tells me this is a truth.
What is Net Promoter Score?
Net Promoter Score, NPS, is a KPI that can be used to determine whether your customers are brand promoters or brand detractors with a simple calculation. The score is created by taking the total number of respondents minus the number of detractors and neutrals divided by the total number of respondents then multiplying by 100. The way in which promoters are identified is via survey. Have you ever participated in a post call, website visit, or app visit survey?
On a scale of 1-10, 10 being most likely, how likely are you to recommend XYZ Inc. to your friend or family?
Now do you recall? Questions like these are what give rise to NPS. The scale of promoters, neutrals, and detractors can be different at each company, but in general 10-9 (Promoter), 8-7 (Neutral), and 6-0 (Detractor). Some companies have found that NPS gives their company a better understanding of brand sentiment. I think companies can sometimes use metrics and KPIs as a false sense of security. Cough..Cough
Who Takes Surveys?
Let’s walk through two scenarios which result in the opportunity to participate in an NPS survey. Joe gets his internet bill and remembers that his service stopped working for a few days in the middle of the month. Joe calls the internet company and complains about loss of service. The customer service representative empathizes with Joe and refunds the cost of 2 days of service. Joe is about the hang up the phone when he hears the operator say, “Please stay on the line for a brief survey.”
Joe is happy with the result of his call, but he has two options. He can take the survey or hang up the phone without taking the survey. Joe may be included in NPS or maybe not.
Jane calls the internet company and complains about loss of service. The customer service representative empathizes with Jane but refuses to refund the cost of 2 days of service. Jane is about the hang up the phone when she hears the operator say, “Please stay on the line for a brief survey.” Jane is now fed up. She ponders why she will waste any more time with the lousy company skips the NPS survey.
Jane is unhappy with the result of her call, but she has two options. She can take the survey or hang up the phone without taking the survey. Jane may be included in NPS or maybe not.
Let’s be real, who has the time to take these surveys? Perhaps the unemployed, retired, or people who have a fear of missing out. These are your net promoters!
What’s the Alternative?
Mining data to identify brand promoters would be my alternative action. For example, as an e-commerce store, I would track average repeat purchase per customer over different time periods. Also, I would track average ticket per customer. If a customer returns something that would mean they are dissatisfied. Customer return rate could identify bad experiences with the brand. Actions speak louder than words or surveys in this case. Use data to find your good customers. Use data to find your bad customers. Make your good customers better. Turn your bad customers into better customers or get rid of them. Creating a metric along the lines of Customer Present Value, CPV, is a step in the right direction. NPS is BS!
Visit DJR Jeeves!