Choosing among customer experience metrics can feel like alphabet soup when you just want to know whether customers are happy. Use NPS for loyalty, CSAT for satisfaction with a moment, and CES for how easy something was.
| Item | Value |
|---|---|
| NPS score range | -100 to +100 |
| Promoter ratings | 9 or 10 |
| Detractor ratings | 6 or under |
| Average NPS, insurance (2024 Retently study) | 80 |
| Average NPS, cloud and hosting (2024 Retently study) | 39 |
| Global average CSAT | 86% |
These figures come from IBM’s explainer on net promoter score and Help Scout’s comparison of satisfaction metrics. Averages and definitions can change, so check the source before you rely on them.
What customer experience metrics measure
Customer experience metrics turn how people feel about dealing with a company into numbers you can track over time. CustomerGauge says the three most popular tools are CSAT, NPS and CES. Each one answers a different question, so the choice depends on what you need to know.
The first split to understand is relationship versus transactional measurement. According to CustomerGauge, relationship studies show how a customer views your brand as a whole. Transactional studies are all about touchpoints, such as a service call or a purchase.
Most companies do not stop at one number. CustomerGauge research found that 49% of NPS users also measure an additional metric, most commonly CSAT (33%) or CES (1 in 6). In practice the metrics work as a set of tools rather than rivals.
Scores describe results, while the habits behind them belong to your customer service culture. A metric that nobody acts on is only a dashboard decoration.

Net Promoter Score: loyalty in one question
Net Promoter Score is built on a single idea: ask customers how likely they are to recommend you. IBM describes the method as asking customers to rank that likelihood in a survey. A dedicated explainer covers what is net promoter score in more depth.
In its 2020 profile, Fortune noted that two short questions are the essence of the system. The second question simply asks the customer why they answered as they did. That open comment is often where the useful detail sits.
The metric was introduced in a 2003 Harvard Business Review article by Bain consultant Fred Reichheld, Fortune reports. Reichheld spent two years testing questions, and he said likelihood to recommend won because it best predicted future customer behavior. IBM adds that Satmetrix and Bain & Company were also involved in its development.
Responses are sorted into three bands, as IBM explains:
- Promoters score 9 or 10.
- Passives score 7 or 8.
- Detractors score 6 or under.
To get the result, subtract the percentage of detractors from the percentage of promoters, a step Wikipedia also describes. IBM notes that the outcome runs from -100, when every answer is a detractor, to +100, when every answer is a promoter. Step-by-step arithmetic is in the guide on how to calculate NPS.
Adoption is wide. Fortune reported in May 2020 that at least two-thirds of the Fortune 1000 use the score, including most or all of the financial service companies, airlines, telecom companies and retailers.
A weak result has a practical meaning. IBM says a low score can signal too many detractors, which hinders growth in the customer base.
CSAT and CES: satisfaction and effort
Customer satisfaction score (CSAT)
CSAT measures how satisfied customers are with an experience. IBM defines it as the percentage of respondents who say they are satisfied (4) or very satisfied (5) in surveys. The page on customer satisfaction score CSAT goes deeper on formula and survey design.
Help Scout puts the global average at 86%, while warning that it can vary wildly by industry. It also observes that every company measures CSAT slightly differently, so compare like with like. A score from one help desk may not match a score from another.
Customer effort score (CES)
CES asks how easy or hard it was to get something done. IBM says it determines how easy or difficult customers found accomplishing their goals in an interaction with a company. Most surveys rate the difficulty from 1 (easy) to 5 or 7 (difficult).
Help Scout uses an agreement statement instead: customers react to the line “Company made it easy to handle my issue” on a scale from 1-7. Because the scale direction differs between tools, read the survey wording before you compare results.
The same source cites research that 94% of customers who report “low effort” interactions will repurchase in the future. A separate guide to customer effort score covers question wording and timing.

How do NPS, CSAT and CES compare?
The three metrics overlap but they are not interchangeable. The table sets out the question each one asks and where it fits best.
| Metric | Question it answers | Best suited to |
|---|---|---|
| NPS | Would customers recommend you? | Overall loyalty across the relationship |
| CSAT | Were customers happy with this experience? | A specific interaction or general experience |
| CES | How easy was it to get this done? | A specific action, such as a purchase or an issue resolved |
The comparison draws on IBM and CustomerGauge; vendors word these definitions differently, so check the sources before you rely on them.
IBM points out the limits of each score. CSAT shows satisfaction but cannot tell you whether a person would pass the company on to their network. CES focuses on a specific action and can say less about overall satisfaction. A side-by-side guide to NPS vs CSAT vs CES expands on these trade-offs.
For business-to-business brands, CustomerGauge reports that NPS is the most popular experience metric among B2B companies. It adds that most B2B companies use NPS as their primary metric and layer in CSAT or CES for specific use cases. A consumer help desk may reasonably begin elsewhere.
Benchmarks and limits of the scores
Reading benchmarks
Benchmarks help only if they match your industry. IBM cites a 2024 Retently study that found large gaps between sectors, with insurance averaging 80 and cloud and hosting averaging 39.
Help Scout offers a looser rule of thumb: Amazon and Chewy report NPS results in the mid-50s, so anything above 40 can be considered decent. Treat figures like these as context rather than targets. The good NPS score benchmark guide takes the question further.
Your own history is the fairest comparison. IBM advises companies to deploy surveys frequently so they can establish benchmarks and see whether changes are improving the score.
Where NPS falls short
NPS is widely used, yet Wikipedia records several criticisms:
- Scholarly critique has questioned whether the score is a reliable predictor of company growth.
- It does not offer detailed insight into specific issues or identify specific areas for improvement.
- Reichheld recommended de-linking it from employee compensation, saying linkage made employees care less about pleasing customers and more about getting a high rating.
Because a single number hides detail, the open-ended follow-up matters. IBM says some companies ask why people chose their score and what the company can do to improve it.

Putting customer experience metrics to work
A measurement routine can stay simple. These steps give a workable order:
- Decide whether you need to know about loyalty, satisfaction with a moment, or ease, and pick the metric to match.
- Start from tested wording, such as the examples in customer satisfaction survey questions.
- Choose the timing. IBM says many companies send transactional NPS surveys after a customer made a purchase.
- Add an open question asking why, so the number comes with a reason.
- Feed answers into a wider voice of customer program so scores stay tied to the comments behind them.
- Follow up on unhappy responses, using the approach in how to handle customer complaints.
IBM lists several ways companies can engineer better NPS results, including paying attention to passives and running root-cause analysis. The point of measuring is to find what to fix, not to chase a higher number.
Frequently asked questions
How is NPS different from CSAT?
Show answer
NPS reads as a loyalty signal because it asks whether a customer would vouch for you. CSAT asks whether they were pleased, which suits a recent interaction. A customer can be pleased without being willing to recommend you, which is the gap IBM highlights.
Which metric should a business measure first?
Show answer
Start with the question you most need answered. Help Scout says CSAT is almost always the first metric businesses use, mainly because it is easy to understand and most help desks offer it as standard. Add NPS or CES once you want to track loyalty or friction.
Can a net promoter score be negative?
Show answer
Yes. When detractors outnumber promoters the result drops below zero, and the lowest possible figure is -100, according to IBM.
How often should you send surveys?
Show answer
Often enough to see trends, but not so often that people tire of it. Wikipedia notes that many consumers in 2025 complain about being bombarded with too many requests for ratings. Space out requests and keep each survey short.



