Because you can objectively determine who is in an "actor's" (customer) network and the relative strength of ties between customers. You can see who they call, how often and for how long.

This is exactly what 2 researchers have done in this recently published study.

My interest is in gathering insights that can help refine the application of bottom-up (transactional) NPS to the improvement of customer experiences - to increase the percentage of promoters and therefore returns for our clients.

Generally marketers look for positive case studies; NPS up then sales up. But this is a little unbalanced, as we have known since The Loyalty Effect in 1996 that customer churn reduces company profits as inevitably as loyalty increases them. And nobody I have read suggests that NPS and customer loyalty are independent. I am currently also reading Bob Hayes latest on Measuring Customer Satisfaction and Loyalty and he makes a strong argument that NPS and other customer satisfaction metrics in fact measure the same underlying psychological intermediary in our customers; loyalty. We think NPS makes it easier to measure.

If transactional NPS helps organisations identify process improvements that improve advocacy that improves results - we should also expect the opposite to hold true. Promoters attract more customers, detractors should cost us customers. Nitzan and Libai did not explicitly relate churn effects to NPS, but their findings are directly relevant to this discussion - and it would be very interesting to see if the defectors in their study were detractors & if the customers they influenced were fellow detractors or neutrals.

Nitzan & Libai found, in the database of a cellular phone company;

• Defection of a network neighbour increases the likelihood of a customer's defection. By 150% in raw numbers, by 80% when allowing for tie strength and customer similarity (homophily).
• • Tie strength they measure by looking at the volume of contact between the 2 customers 
  • Homophily they measure by assigning points for each of 4 characteristics that were similar between the 2; gender, age, marketing segment, socio-economic status - it is important to control for homophily as a similar response to the same product by similar people may be mistaken for an indication of social influence when it is not.   
• 'With their feet' detractors who leave impact customers more strongly the closer they are in social network terms; "...a 1% increase in a customer's tie strength with defecting neighbours is associated with a 2% increase in that customer's hazard of defection."
• Similar customers are likely to act the same but at a much lower rate; "...1% increase in similarity between a customer and a defecting neighbour is associated with a 1.1% increase in the customer's hazard of defection..."  
• The influence of a neighbour's defection decreases markedly with the passage of time  
• A loyal customer is less likely to defect following a neighbour's defection - customer loyalty 'immunizes' against the effects of a defecting neighbour. (Loyalty was measured by an index of tenure and call volume, ARPU) 

 
An interesting study.

1. The NPS of network contacts should be considered when trying to understand and manage customer churn. This is likely to be easier in a B2B environment where network structures, the formal ones at least, are easier to discover and track. No customer is an island!
2. If we needed more urging to respond to customer defection quickly here it is, "If a firm aims to deal with this possible influence, it should act fast, as close as possible to the neighbour's defection event". If you use member-get-member programs and a referrer defects, look to his referees asap.
3. Customer loyalty has multiple benefits, making this whole 'Active NPS' endeavour worth it. Loyal customers;

• provide higher profits for longer 
• have higher social value as they influence others to adopt products AND as this study shows
• loyal customers are less likely to be affected by the defection of others   

  
 React to bad customer news quickly!

Good to see 'common sense' backed by hard data - what do you reckon?