Net Promoter Score (NPS) – Still useful or ineffective?


Share on LinkedIn

Photo by Lukas from Pexels

The Net promoter score (NPS), introduced by Fred Reichheld from Bain in 2013, is still one of the most commonly used metrics for businesses to measure and evaluate their customer experience.

At its core the Net promoter score is a customer loyalty metric that uses an index score between -100 to 100 and is determined by asking a customer’s likelihood to recommend your organisation, product or service to a friend. The net promoter score is calculated by subtracting the detractors from the promoters.

A 0-10 scale is used to measure which customers are promoters, passives & detractors.

Customers who score:

– 0-6 are considered detractors (i.e., likely to have a negative perception of your company, product or service or less likely to spend money with your company)
– 7-8 are considered passives (i.e., exhibit behaviour somewhere in between a promoter and detractor)
– 9-10 are considered to be promoters (i.e., more likely to have a positive perception of your company and more likely to continue to purchase from your company and recommend them to their friends).

Yet, given the changes and evolution in customer experience evaluation techniques, emerging new technologies and Big Data, is NPS still an effective metric to assess the health of your customer experience and predict customer loyalty?

Many organisations would argue NPS is still relevant to their business for the following reasons:

It is simple to implement. For example, companies just need to ask customers one key question to predict their loyalty. This means they can quickly implement NPS via surveys over the phone, email or via a third party survey provider, without needing to wait for more detailed and complex research to be conducted. The simple methodology and segmentation of promoters, passives and detractors can also make it easier for employees to analyse the results and action improvements.

It’s easier to benchmark their customer experience against competitors and the industry. For example, the NPS measure methodology design is a globally standard methodology, which makes it easier for companies to compare ‘apples’ to ‘apples’, as NPS is measuring the likelihood to recommend the specific company, product or service, rather than specific product and service attributes or components of the customer journey that may be unique to each organisation.

NPS is simple for everybody in the organisation to understand due to the scale between -100-100+ and the simplicity of being able to interpret that if NPS is going up, it infers customer loyalty is increasing and if it’s going down, it is decreasing.

Also if NPS is implemented in a more effective way, companies can often ask follow-up questions to determine why customers are providing the scores. This can help organisations more effectively pinpoint the drivers of loyalty, rather than just understanding the ‘what’ of the aggregated NPS number.

NPS can drive collective accountability in delivering the customer experience. As NPS is often designed as a whole of company measure, this encourages everybody to be accountable for the customer experience as they are all KPI’d on the same goal. Having a consistent KPI for all employees can influence employees to collaborate more effectively to deliver a unified customer experience.

Putting these benefits aside, there are several limitations and constraints of NPS, which are often associated with the methodology design and implementation approach.

The most common NPS challenges we see include:

NPS provides limited insights into the true health of a customer’s experience and cannot be relied upon as a predictor of a customer’s propensity to be loyal and purchase more from your company. For example, the NPS question itself is asking about a customer’s advocacy towards an organisation’s brand, rather than their propensity to purchase from them again. Whereas, understanding a customer’s propensity to purchase would usually require a more specific intention to repurchase question alongside the NPS question.

While organisations implementing NPS more effectively than others often include follow-up questions in their surveys to understand the root causes of the NPS score, the insights are not often translatable or useful. For example, many organisations often ask generic questions which means a customer might not be too specific about whether they had a product related issue, service issue or were generally dissatisfied with one small aspect of the experience rather than an array of things.

What’s more customers are often bombarded with NPS surveys from every company that they purchase a service or product from, which can often result in them getting survey fatigue. This factor combined with the time poor nature of most consumers these days, often results in customers providing little detail about the drivers of their experience or not responding at all. As a result, organisations find it challenging to obtain the breadth and depth of insights they require to obtain a comprehensive evaluation of their CX performance, pinpoint root causes and identify impactful improvements to their customer experience.

Even if companies tailored their NPS research to obtain a comprehensive level of insights about the drivers behind their NPS scores, this information would unlikely provide a holistic understanding of customer loyalty and experience performance. For example, NPS wouldn’t accurately help predict customer churn or the impact that employee engagement and culture would have on influencing and improving customer loyalty. Nor would it pinpoint operational and service issues that were impeding the delivery of a unified experience. This is why your NPS methodology would need to be evolved to align with other key interrelated metrics and insights across your key business functions and teams.

NPS insights can often be biased due to the fact that customer responses are not standardised and they are qualitative in nature. This means that organisations are often not able to interpret these effectively and even if these are interpreted, they rely on humans to interpret them manually, resulting in potential inherent bias. This in turn leads to poor quality insights that are used to inform CX improvement investment decisions, resulting in ‘garbage in’, ‘garbage out’.

NPS is an attitudinal metric meaning it is based on what customers say and perceive at the time, rather than it being an actual reflection of how they behave. For example, a customer may demonstrate their dissatisfaction of their experience via NPS but may not act on it for several reasons, such as switching barriers or the transaction may not be considered an important influencer of loyalty. For instance, I have been with the same mortgage provider for the past 3 years and despite the increasing fees and amount of effort required to make changes and updates to the mortgage, it is easier to suffer through the poor service experience, rather than switch providers and go through the same admin headache again. In a similar light, I might provide a high NPS score for a specific transaction that I don’t consider influential to my satisfaction with an organisation and then switch the next day. In this instance just because I provided a high NPS score for a specific transaction, this score doesn’t infer my broader and long-term satisfaction with or loyalty towards that company.

NPS can often tell you the channels or transactions that are working well but it does not effectively evaluate the performance of the broader customer journey. We often see many organisations measure the NPS of a specific customer contact channel, such as a recent interaction with a call centre or a website visit, rather than an ongoing and integrated review of a customer’s interactions with their company and a broader review of their entire customer journey. Measuring these transactions isn’t reflective of their overall experience with your company and relying on transactional or channel specific experiences will likely detract you from focusing on the entire customer journey. It will also make it challenging to obtain the critical insights, such as the drivers of loyalty, which are required to inform and shift customer behaviours. While it is important to understand from a bottom-up view how each transaction, channel and product influences NPS, it’s also important to understand from a top-down perspective the key drivers of loyalty that will influence a step change in the customer experience and business outcomes.

The timing of a NPS survey after the experience is likely to have an impact on its effectiveness to predict customer churn behaviour. This has been reinforced by recent global loyalty research from Gartner[1] indicating that customers often forget the details of their experience over time and therefore their attitudes and perceptions of the company are likely to change. The study inferred that after a 3 month time period, a customer is only likely to remember their salient experience (i.e., the things that are most important to them rather than every trivial detail). Given this, if a company conducts NPS research immediately after the end of the transaction with its customers, this is likely to make it more challenging to identify the most critical drivers of loyalty that are required to inform improvements in the customer experience. Therefore, relying on the NPS insights alone in this instance to make decisions is likely to result in ineffective CX investment decisions.

Finally, we need to remember the purpose of why we are measuring CX, rather than getting bogged down by CX metrics, such as NPS, CSAT or Customer Effort Score (CES). For example, it is common for organisations to start the CX measurement conversation with the types of metrics they would like to use to evaluate the effectiveness of their CX.

The issue is that organisations get so caught up with the concept of needing to measure CX, this can often distract them from the broader purpose of measurement, such as the ‘why’, ‘what’ and ‘so what’, ‘now what’. I.e., How is our CX strategy currently tracking, What is driving CX performance and loyalty? What are the levers we can pull to improve performance? And which parts of the customer journey do we need to improve, rather than the individual channels and transactions?

While standard methodologies like NPS can be useful, they still need to be designed and customised to consider the specific nuances of your organisation, such as your culture, strategy, operations, products and customer segments, to ensure CX has a direct impact on your business performance.

In summary, we could spend all day debating NPS benefits and constraints and there are many more we could add to the list. Yet, this will not help us improve the way we measure and enhance the customer experience to increase customer loyalty and improve business outcomes. It’s therefore time we shift the discussion and actions away from whether to use specific metrics towards how can we effectively design, implement and integrate a customer experience performance management approach for the purpose of improving customer, business and employee outcomes.

[1] Cited by Gartner CEB, ‘Using Customer Feedback to Predict Behavioral Loyalty’, 2018 Customer Experience Loyalty Survey, Gartner, 2018


  1. I’m going to comment from a customer perspective, as I am now retired. I find the questions asked by companies to be too simplistic. Such as “how likely are you to recommend X-company based on your call today.” Now whether or not I (or anyone else) would recommend a company is most likely determined by the total experience with the company combined with the perception of its competitors, not just one call.

    If the contact was by phone, a phone call or text is often received afterwards.
    If online, a follow up email.
    Again, both these are either too simplistic or too long, with the perception of wasting my time.

    Has anyone ever successful measured the outcomes from all this activity.

  2. Ellie,
    Great piece! You’ve shared many subtle aspects of the limitations of consumer opinion surveys. Generally, I think we obsess over the NPS score itself and spend too little time on the what, why and what now questions that you cite. Regarding Gartner’s inference on “salient experiences”, we’ve found that most businesses spend all their time “reducing friction” in the routines of their consumer journeys and little if any time delivering “emotionally connective experiences.” We’ve found that it is these emotionally connective experiences (especially ones that dramatize the orgs unique values or promises) that consumers value, remember and predispose the consumer to future purchases.

  3. Hi Karen, you’ve raised a good point and I completely agree with your feedback. Just because you rated one transaction as 10/10 doesn’t mean you are loyal or satisfied and it’s not reflective of your complete experience with that company. I’ve never seen any research on the effectiveness of NPS. Have you seen this anywhere? I definitely think this needs to be done to help influence companies to review their measurement and evaluation approach. In your opinion, what do you think would be a better approach?

  4. Hi Matt, I completely agree with your feedback on the emotionally connective experiences. These are vital to understand when comprehensively mapping out your customer journey (I mean with customers too not just doing an internal workshop :)) Unfortunately we find that many organisations don’t always want to invest the time and energy in validating and deeply understanding these emotionally connective experiences, which would enable them to drive true value from their CX. Do you have any case studies you could share that compare the transactional routine improvements to CX vs. the step change improvements in the truly important components of the customer journey and the value they deliver to business outcomes and on customer loyalty?

  5. Hi Matt,
    Further to what you’ve said about customer values. One common issue is that many organisations don’t start with their customer value proposition when developing their vision and strategy. This means that this may often be an afterthought, which can mean that work is done in a disjointed way. The issue is then that the key decisions, processes and strategic levers don’t focus on delivering the right experiences due to the underlying processes, strategy, culture etc, which means they can’t deliver on the attributes that customer care about most in the value proposition. This is one of the things we are seeing which then has a flow on effect on how you effectively measure and improve the delivery of your CVP through the customer experience. Have you also found this and do you know of any good case studies where the customer is truly integrated into the strategy and operations, rather than just lip service or retrofitted?

  6. Hi Ellie

    My recommendation on identifying customer experience/satisfaction would be to revert to something that seems to have been lost, and that is the indepth statistically valid research survey. In my history as a marketing director back in the 1980s, we did these every 3 years, a combination of mail and phone, selected specifically from subgroups of our customers (age, geo, income, etc.). This is not just a self-selected survey, but a very strong representation of the demographic make up of your customers. This type of research, combined with focus groups, helped us design new services to stay ahead of the competition. I’m sure some companies still do this, but there’s an attitude quite often in business that the company knows best for the customer, or that customer ideas may be too expensive.

  7. Show me the money. It was a line made famous by actor Cuba Gooding in the hit movie Jerry McGuire. It is a plea for results, not promises; performance, not intentions. And it reminds me of some of my client companies that rely on the popular Net Promoter Score as their single proof of service excellence.

    Now, don’t get me wrong. I have great admiration for the way the NPS led the C suite to focus on customers and service instead of only drilling down on this quarter’s financial balance sheet. But, unfortunately it enticed some leaders, seeking to simplify the customers complex relationship with their organization, to rely on a single question to gauge customer affinity. It became the search for an answer to an ultimate question, somewhat like, “I promise that someday I will marry you.”

    We all know customers demonstrate their loyalty in a variety of ways. Loyal customers buy again, buy more, trust more, and provide a wider berth for mistakes. Some identify so much with a brand they proudly wear it. Their Bass Pro Shops caps or Harley-Davidson jackets attest to their allegiance. The NPS does remind us that the highest level of customer fidelity is their eagerness to be advocates. The NPS does not seek to learn about performance, only intention. It is a good start. But, the real test of customer advocacy is their actual words of recommendation or their amorous stories that start with phrases like “You’re not going to believe what happened to me.”

    I like Cuba’s line. Promises don’t build trust; performance does. Intentions are not the stuff that drive the bottom line, actions are. The politician asks, “Will you vote for me?” and the citizen, excited by the campaign, says “yes.” But, promises made at a political rally are not the ultimate test of political support, only votes cast behind the curtain. Basing evaluations and scorekeeping on how many customers promise to come back to your store is foolhardy as a gauge when the metric that matters is how many cross your threshold again.

    We were working with a major bank in Central America. They had for years used as their key metric in the customer service arena, a survey question giving customers a chance to register their “overall customer satisfaction.” They believed that if customers were satisfied—meaning their needs and expectations were met—they were very likely to be loyal. They were happy 80% of their customers indicated they were satisfied with their experience and the bank worked to get that percentage higher. Yet, the stark reality is that about 70-75% of customers who leave an organization to go with a competitor report they were satisfied with the one they abandoned. It seems satisfaction is not worth much.

    They then switched to the Net Promoter Score as their new metric–a survey question that sought to learn how many of their customers indicated they were promoters—meaning they would recommend the bank to a friend or colleague. And, they were thrilled that about 80% of their customers were classified as “promoters” and the bank worked to get that percentage even higher.

    That is the point when we came on the scene. We started a conversation about the demonstration of loyalty—a “show me the money” discussion. Do you want to improve satisfaction or do you want to improve the bottom line? Do you want intent to recommend or do you want customers to actually recommend you? When we encouraged them to add a question: Have you recommended our bank to a friend or colleague, 80% dropped to less than 55%. It was an eye-opener on what really matters—performance, not intention.

    Seeking a cause and effect metric is tough in the qualitative world of customer experience. There is so much subjective, eye-of-the-beholder built into the assessment. There are so many variables that make up the anatomy of affection it is challenging to sort out what actually caused what. So, we reach for correlation measures coupled with a lot of hope. We hope that if they say they would recommend us, they actually do. We hope that if they say they found their experience “wonderful” (or the “researchy” phrase—like completely satisfied), they will demonstrate that evaluation by positively impacting our bottom line.

    In the end, no single metric can adequately capture what we need to know to have a high level of confidence that customers will “show you the money.” Lifetime value, the share of wallet, the degree of effort, recency, retention, spend per transaction are all useful in ascertaining how customers evaluate the experiences and translate it into the performance that matters to the success of the enterprise. Like, love, customer devotion will fortunately retain a bit of mystery hopefully leading us to never take it for granted. (Adapted from a post that appeared on April 22, 2019 on

  8. Hi Ellie and Karen,

    First, to Karen’s point: I typically find NPS based surveys too restrictive. The questions almost always ask that I limit my comments to the last interaction with the organization. For example, a call center contact or the agent on my flight from Chicago to Denver! Without exception, my feelings for the organization are based on other aspects of that recent experience. As the client, it’s frustrating to know that the surveyor is not even asking for the relevant information!
    Ellie: You are absolutely right about the value propositions. I work primarily with the financial services industry. Our first step is transforming the typical value proposition from…”committed to provide world class products and services with honesty and integrity” …to something much more human and meaningful. One recent example is an FI in Atlanta that is now “Driven to be the best part of your day.” With this promise, they have set the bar beyond their commodity-based competition and are actively curating emotionally connective experiences for their clients. It still surprises me that so many in financial services think they can “win” with a sexier checking account or a clever new marketing campaign. It no longer works this way! We can’t expect consumers to feel passionately about our organization if all we care about are the nuts and bolts of operations. I’d be happy to share my mini-case studies on this topic.

  9. Chip–The concept of “satisfaction” is about as useful as an old calendar! In a world of interchangeable commodities, we’ve found that retailers need to communicate and appeal on much more fundamental, human levels. Consumers have learned to expect so little from banks, as one example, Meeting their low expectations does nothing to differentiate one institution from another. Chasing the productivity frontier of operations, systems and technologies is just the starting point. Curating experiences that dramatize meaningful values and commitments is where one FI can separate itself from its rivals.

  10. Hi Karen,
    I completely agree with you point in relation to conducting in-depth qualitative research to dig deeper into the customer research rather than relying on internal perspectives. We often see a preference for companies to conduct workshops with internal employees to find out what the root causes of dissatisfaction or poor NPS results are and unfortunately they do not validate this with customers. While customer surveys are often useful to obtain a validate some things, organisations still need to speak to customers to dig deeper to obtain these insights.
    Why do you think organisations are reluctant to do this? Is it time and resources or budget?

  11. Hi Chip,
    You’ve made a great point about not relying on one single metric to effectively understand the customer experience. As you said, intention does not translate into customer behaviours and therefore improving business outcomes. A great example is the Australian Federal elections where the Labour Party scored really well and were rated as the favourite in the pre-poll (intent data) and then the Liberal party ended up winning. In the same way, organisations often make their decisions based on customer intent data whether the insights be from CSAT, NPS, VoC or other general brand research. However, organisations globally often are unable to connect the improvements they make with business outcomes, such as improve customer acquisition or retention. This is where it’s necessary to build an effective customer scorecard which is integrated into the business strategy via cause and effect linkages, as you say. While the suite of metrics that are interrelated are good to keep everybody honest and focused on achieving the same goal, the organisation needs to focus on what the performance analysis is telling them, rather than a suite of KPIs. If this is developed properly it can be a game changer.
    Where have you seen this work well?

  12. Hi Matt,
    Thanks for your comments. I would love to see the mini-case studies on the customer value propositions if you would be able to share them via Linked In. In relation to your comment regarding customer satisfaction, it can be useful if you develop the research and link other insights to it. It all depends on the design and execution. I agree with you that most CSAT design approaches can often be flawed and do not provide a comprehensive level of insights for decision-making but if you effectively translate the attributes of the value proposition and measure these across the most important points in the customer journey, this can be quite impactful. I agree that you can’t rely on this alone as it’s intent data but if you link it with service level, operations, behavioural insights and deeper research, this can be very powerful.
    Where have you seen this done well before? Is this something the Financial Services segment are doing well in the US?

  13. I am curious as to why someone has not advanced the thought process of NPS from “would” you recommend to “DID” you recommend. In business, I don’t really care what you ‘would’ do – I am more interested in what you DID. As the saying goes — “The road to hell is paved with good intentions.” If you intend to recommend a business/service/product how does that actually help the business? Thoughts?

  14. As an undergrad at the University of Illinois, I took a research class from Professor George Beam. His first law of social science research was, “If you want to know, don’t ask!” Of course, he was pushing our buttons, but his point was clear. What consumers say and what they do are often very different things. Beam would argue that if we “ask” we must also “observe” in order to truly understand the subtle motivations and triggers of consumer behavior. The key insight for our client’s successes have come from that deeper observation. The realization that transforms strategy and the daily practice of the business is that human beings are emotional creatures. We make emotional decisions and then rationalize them! Most businesses–particularly complex, highly regulated businesses like banks often miss this reality. Its starts from the beginning. They create value propositions (mission, vision, brand promises, etc.) that are grounded in the mechanical operations of their industry, e.g. “To provide world class financial products and services with integrity and honesty.” Few even set their sights on the emotions that motivate the need for these commodity products and services. Once a bank dives deep enough to truly understand their markets at an emotional level and enroll all staff in delivering experiences that make durable, emotional impressions, amazing things happen–inside the organization and with the public.


Please enter your comment!
Please enter your name here