Wrong customer metrics

Why NPS, CES & CSAT aren’t good customer success metrics


Improving CES, NPS and CSAT may help organisations track and improve customer convenience and experiences and ultimately create happy clients. But they may not be appropriate for the goals of customer success leaders, says Sampson Lee. 


23rd Sep 2019

A reader of my previous article A customer success mindset can't save CX alone - but it's a start asked me how a customer experience (CX) manager can transition into a customer success (CS) manager. It prompted me to write a new series of articles.

In my opinion, there are three changes that CX folk must make in order to develop into an effective ‘customer success’ leader.

  1. Goal: drive repeat customers in place of “happy” customers.
  2. Strategy: keep promises instead of enhancing interactions.
  3. Philosophy: customer success in lieu of customer obsession.

This article is about the first change – goal.

“How likely are you to recommend Sampson Lee to a friend?”

Let’s assume Sampson is highly creative but isn’t very easygoing. His ‘personal’ NPS score would likely be low. It’s because people tend to favor those who could make them feel pleasant or “happy”. NPS, however, would do a disservice to someone who is mainly looking for Lee to fulfill an inventive task.

Similar things happen to brands: “How likely are you to recommend Virgin Atlantic / Ryanair to a friend?”

The NPS score of Virgin would probably be high and Ryanair’s would be low. What if you ask your friends, “Which airline would you suggest for cheap tickets?” I trust that they are much more likely to recommend Ryanair than Virgin.

Given such an oversimplified question, NPS is to a large extent biased towards service-focused brands (e.g. Virgin Atlantic and Zappos) and service-related brands (like Southwest Airlines and Amazon), as respondents naturally give high scores to those brands that make them feel pleasant or “happy”. Hence, simply measuring and comparing NPS scores without specifying ‘the task’ misses the point and could be vastly misleading.   

Why are customers paying $5 for a cup of Starbucks coffee?

When you aim to save time and have a quick meal, you’ll probably go to McDonald’s since it has stores at virtually every corner and their transactions are highly efficient; when you desire a third place between work and home to slow down and enjoy your precious moments, Starbucks would likely pop up in your mind.

Apparently, the principle jobs-to-be-done for McDonald’s and Starbucks customers are completely different: McDonald’s’ is saving time and effort while Starbucks’ is enjoying affordable luxury. Delivering a “fast and easy” experience will help McDonald’s customers – but won’t help Starbucks’ – reach their prioritised goals.

Customer effort score (CES), as its name implies, is an ideal metric for brands with “effortless” as their desired outcome, such as McDonald’s. However, if your jobs-to-be-done aren’t about quick and frictionless – like Starbucks – pursuing CES at critical touch-points would dilute resources, damage customer success and reduce brand loyalty (see Why brands must stop trying to eliminate customer effort).

Moment of Buying (MOB)


With reference to 2,318 respondents from a global research [1], ‘exclusive feel from wearing and owning LV products’ is the most important attribute in driving repeat purchases – Moment of Buying (MOB) – out of all 27 sub-processes and attributes during the Louis Vuitton shopping experience.

In figure 1, the horizontal axis represents the importance level of each attribute in driving customer satisfaction (CSAT); the vertical axis represents the importance level in driving repeat purchases. The red circles and the grey dots aggregate the 27 sub-processes and attributes of the Louis Vuitton in-store experience.

The red circles are the top five MOB of the LV in-store experience, including the No.1 repeat purchase driver, ‘exclusive feel.’ Of the Top 5 MOB, only one is an important driver of CSAT (located in the upper-right quadrant); the other four MOB aren’t important drivers of CSAT.

Improve CSAT scores may reduce repeat purchases

Customer metrics

Customer metrics

Customer metrics

Based on over 6,000 valid responses derived from the research projects for Supermarket In-store Experience, Starbucks In-store Experience and B2B Purchase Experience [2], we identified the top 5 MOB for each and denoted them with red circles in figure 2, 3 and 4 correspondingly.

The findings are similar to Louis Vuitton. Of the top 5 MOB, half of them were not important in driving CSAT. In other words, if you focus your resources on enhancing CSAT scores, you will miss 50% of the repeat purchase drivers. Perhaps you need to ask yourself a question: Why spend resources to improve CSAT scores at the expense of repeat purchases?

Branded NPS

Despite NPS, CES and CSAT being the three most important customer experience metrics, they aren’t necessarily fit for measuring and driving customer success. That said, it doesn’t mean that MOB is the only appropriate customer success metric.   

For instance, you may deploy Branded NPS – adding your brand promise in the NPS question – e.g. “How likely are you to recommend Ryanair to a friend for inexpensive airfares?” Now, the recommendation is based on the ‘specific task’ that your brand aims to help customers achieve rather than on their ‘general feelings’; Branded NPS is, unambiguously, connected with your jobs-to-be-done and driving customer success. What’s more, with brand promises being stated, benchmarking the ‘likelihood to recommend’ among different brands becomes justified and meaningful.

Or, you may emulate Amazon and create your own customer success score. Around two decades ago, before the existence of NPS and CES, Amazon launched the campaign “The Best Service is No (Need for) Service” and applied ‘contact per order’ (CPO) as the singular metric to measure the effectiveness of fulfilling its jobs-to-be-done – minimising or eliminating customers’ time and need for contacting Amazon.

Goal: driving repeat customers in place of “happy” customers

Regardless of whether you use MOB, Branded NPS, or design your unique customer success score, effective customer success leaders should never forget that their goal is to keep customers coming back (repeat purchases / retentions / continued subscriptions). 

To summarise: without specific tasks being stated, measuring and benchmarking NPS scores doesn’t make much sense; unless your brand promises are about “fast and easy”, chasing after CES would adversely affect jobs-to-be-done; conversely, focusing resources to improve CSAT scores would reduce repeat purchases; and so I conclude that NPS, CES and CSAT have significant deficiencies in measuring customer success and driving repeat customers. 

Granted, customers like convenience, being well-served and having an overall pleasant experience - improving CES, NPS and CSAT may help address these needs and create “happy” customers. But the customers’ decisions for repeat purchases are predominately determined by how capable a brand is at helping them achieve their desired goals, rather than by how good a brand makes them feel.    

Since pursuing NPS, CES and CSAT would create “happy” rather than repeat customers, and with these metrics are widely adopted in the CX world, customer experience is very often disconnected from business results. It’s little wonder why conventional CX has a high failure rate, persistently fails to perform and most CEOs still don’t buy into it.    

Don’t make the same mistake CX did.   

Customer metrics

Customer metrics


  1. Global Louis Vuitton In-store Customer Experience Research, Global CEM and CustomerThink (U.S.), October 2008.
  2. Mainland China Supermarket In-store Customer Experience Research, Global CEM and CustomerSat (U.S.), May-June 2007; Global Starbucks In-store Customer Experience Research, Global CEM and CustomerThink (U.S.), September-October 2007; Mainland China B2B Purchase Experience (IT Solution) Research, Global CEM and CustomerCentric Selling (U.S.), July-August 2007.


Related content

Replies (4)

Please login or register to join the discussion.

By joakiren
25th Sep 2019 20:56

Dear Mr.Sampson, as a result of what has been my personal experience when measuring Customer Satisfaction in a great bank, I must say the inconvenience of using NPS. First, when asking for a recommendation score you need to apply it to a service, not to a brand. Otherwise people will include in the mental score they gived other factors, such as everything they've heard about the company lately. So if you want to improve your NPS, launch a marketing campain... Second, many people [as I do] need a period of time to think all the little moments good and bad that conforms the global experience of flying. And they have to be prioritized to have the right weight in you decision.

Thanks (0)
By LinkedIn Group Member
26th Sep 2019 09:41

This comment was posted in the MyCustomer LinkedIn group by user Alan Chin:

I think CES/NPS/CSAT are a good high level metrics.

However, a down fall is that it doesn’t enable customer strategists to understand the “why”. Especially when it comes to identifying pain points which could ultimately increase the NPS, and frequency of repeat purchases, to drive the overall customer lifetime value.

Thanks (0)
By LinkedIn Group Member
26th Sep 2019 09:43

This comment was posted in the MyCustomer LinkedIn group by user Hunter Ricci:

These are 3 KPIs that weren't previously on my radar. Moving forward I want to makes CES, NPS and CSAT central to my companies

Thanks (0)
By PaulineAshenden
30th Sep 2019 15:49

NPS and the like only take you so far - they don’t allow you to find out Why customers feel they way that they do. That’s one of the key reasons they are seen as disconnected from business metrics. Instead, for CX to be effective you need to be listening to customers and using their insight to drive constant improvements - as we outline in this blog post https://www.eptica.com/blog/why-traditional-voc-metrics-don-t-deliver-in...

Thanks (0)