Net Promoter Score (NPS), a customer loyalty metric developed in 2003 by Fred Reichheld, Bain & Company and Satmetrix, is going to save the world.
At least that’s how some view it. Ever since its inception, NPS has become the darling of executive management teams all over, and, as such, many organizations now use NPS as their primary Customer Satisfaction (C-Sat) measurement tool. The trouble is, by relying so much on a single question (see “NPS at a Glance” in box below), organizations end up oversimplifying the C-Sat measurement process, which can be just as dangerous as having no C-Sat measurement program in place at all.
Our aim isn’t to condemn NPS or to beat up those who adore it, as NPS can be a valuable tool. Rather, our intent here is to point out its shortcomings and to discuss how it can be most useful as a component of a more comprehensive C-Sat measurement process as opposed to a stand-alone miracle metric.
Whoa…slow your NPS roll
It’s always fun to criticize, so let’s start with NPS’ shortcomings:
The “Ultimate Question” is ambiguous. Not every customer interprets the NPS question in the same way, thus leading to skewed responses. Consider the structure of the question – “How likely are you to recommend Company X to a friend or colleague?” A customer who’s asked to answer it after an interaction with a contact center agent may be unsure whether to base their response on their overall interactions with Company X or whether they should base their response on the recently completed transaction.
To eliminate such ambiguity, it’s wise to phrase the NPS question more carefully for post-contact, transactional surveys – something like, “Based on this last experience, how likely are you to recommend Company X to a friend or colleague?”
The question is sometimes inappropriate or irrelevant. Even when phrased more succinctly, the NPS question is awkward or simply doesn’t apply in many situations and settings. For example, posing a referral question to a customer who calls to cancel service or close an account doesn’t make sense. Nor does it make much sense to pose such a question to a customer who just started service or opened an account, as that customer hasn’t had time to develop much of a relationship with or form a strong enough opinion of the organization. It’s like asking somebody to marry you after the first date.
It’s also a tad absurd for monopolies – such as regional utilities – to ask customers the NPS question. It puts the customer in an awkward position. Even if the customer hates the company, they likely don’t want their friends and family to live without electricity or gas.
The question is inherently biased. The phrasing of the traditional NPS question steers customers slightly toward a more positive response. Asking ‘How likely would you be to refer a friend?’ assumes that the customer is at least somewhat likely to do so. It’s the power of suggestion. It’s like when, after riding a roller coaster with someone, you ask them, “How awesome was THAT?” They may be about to throw up, but they give you a thumbs up instead so as to not disappoint you.
The subtle inherent bias of the NPS question is the reason why many companies that report high NPS scores fail to see business growth and/or clear proof of improved customer loyalty.
NPS is strictly an “outcome” metric. Even when companies carefully rephrase and add clarification to the NPS question, NPS is still by nature only an outcome metric. And while such a metric may be useful – as it draws management’s attention to trends in customer sentiment and satisfaction – NPS alone provides little to no insight into what is causing customers to feel the way they do, and shines little light on next steps. Thus, asking it in isolation – as many organizations do – isn’t wise.
Approaching NPS the Right Way
While many companies have blindly hopped on the NPS bandwagon due to the simplicity of the “Ultimate Question” survey approach, smart companies – those committed to accurately gauging the customer experience and making continuous improvements – have taken the NPS question, customized it for their specific business, and worked it into a transactional C-Sat measurement process that features additional key questions.
For example, the post-contact survey may ask the customer to rate how good or horrible the agent he or she spoke to was, as well as whether or not the customer felt his or her issue was resolved during the interaction. (Numerous studies have shown a strong link between call resolution – particularly first-call resolution (See previous post, “First Call Resolution – Difficult to Measure, Dangerous to Ignore”) – and customer satisfaction. Thus, most C-Sat experts recommend including an FCR-related question in any post-contact customer survey.) Good C-Sat/NPS surveys also include open-ended questions that enable customers to express profanity and provide specific reasons for their ratings, which can prove invaluable for things like service recoveries, agent coaching and process improvements.
Recovering. Coaching. Improving. These are all ACTIONS, and that’s really what any C-Sat/NPS initiative should be about – acting on the data to continuously elevate the customer experience and the bottom line.