NPS and customer experience
Editor’s note: John A. Goodman is the vice chairman of Customer Care Measurement and Consulting, Alexandria, Va.
In a November 2021 Harvard Business Review article, “Net Promoter 3.0,” Fred Reichheld, Darci Darnell and Maureen Burns acknowledge several of the weaknesses and misuses that the Net Promoter Score (NPS) has suffered from over the last decade, many of which I noted in my writing. For instance, they state it is inappropriate to use NPS to evaluate individual service staff as it leads to pleading, bribery and manipulation.
However, their introduction of the “earned growth metric” does not address some of the most severe barriers to practical application of NPS scores. Briefly, earned growth requires estimating the sum of revenue retained from existing customers and new revenue from new customers using internal accounting data. The article does attempt to measure customers from positive word of mouth but measures new customers, not the events that led to the word of mouth. Again, the aggregate impact but not the specific customer experience.
While this approach will support rough estimation of the overall payoff of CX efforts, it still misses several critical intermediate steps which are important to both gaining buy-in and action.
Net Promotor 3.0: What is missing?
1. Granular issue identification with rigorous analysis of strategic solutions.
My analysis of inaction on CX suggests that it is not a lack of overall belief in the potential impact of CX actions, it is being able to justify the investment of X dollars in a specific CX action vs. advertising or a new distribution center. These decisions are made easier when you are talking about investment in a new customer onboarding process or modifying the invoice adjustment process. In each case, you can estimate the number of customers currently having a bad experience due to the issue.
Further, you can work with quality or lean process professionals to define several possible approaches to fixing the issue and how to pilot test it in a rigorous manner. This analysis must be at the granular level, not at the overall CX level. NPS does not add to granular analysis.
Finally, you can measure whether the needle moved if the number of problems decline after your fix. Problems can often be measured via internal operations data, e.g., delayed shipments, invoice adjustments and complaints.
2. Business case by granular issue.
NPS 3.0 does mention using the lifetime value of the customer, but then uses it to calculate the overall increase in revenue from all CX activities. Investment is usually not made in a blanket fashion but rather by defining critical projects and adding to a total investment. There is almost always a requirement for cost benefit calculation at the project level before an overall strategy is allocated millions of dollars.
CCMC’s Market Damage and Market-At-Risk Models have covered this process. (See details of both models in Goodman, “Strategic Customer Service,” Harper Collins, 2019 as well as “Customer Experience 3.0,” Harper Collins, 2015.) Our calculation estimates the number of customers, and the CFO can attribute as much revenue per customer as feels comfortable. For each type of problem, you can then estimate the revenue damage by seeing how many problem occurrences take place.
3. Word of mouth and margin impact by granular issue.
The key to winning new customers is word-of-mouth referrals. Successful B2C companies like Harley Davidson, Chick-fil-A and USAA get more than 70% of their new customers from word of mouth. Our surveys of B2B put the percentage as high as 90%, especially for small and medium businesses. Resolution of specific issues as well as different kinds of delighters create different magnitudes of word of mouth. This impact can be quantified by issue. Likewise, customer’s willingness to pay more for a product or service is affected by problem and delight experiences – again this varies by CX issue.
When researchers need to address granular issues
A broad estimate of the overall revenue retained from existing customers and added from new customers in aggregate is somewhat helpful. But it begs the question of how much we get by addressing specific problems, which is where most decisions are made. NPS 3.0 does not appear to answer that question. I welcome your comments and complaints.