NPS (Net Promoter Score), CSAT (Customer Satisfaction Score), and CES (Customer Effort Score) each measure a different dimension of customer experience: long-term loyalty, post-interaction satisfaction, and journey ease. None of them is sufficient on its own. Used together, they give a complete picture of how customers feel, where friction occurs, and which actions to prioritize.
This guide covers the definition and formula of each metric, how to read the scores, industry benchmarks, and how to combine them to drive real CX improvements.
Created in 2003 by Fred Reichheld at Bain & Company, the Net Promoter Score measures the likelihood that a customer will recommend your company. It is a relationship metric, not a transactional one: it reflects the overall bond between a customer and a brand, not a single interaction.
The question: “On a scale of 0 to 10, how likely are you to recommend [company/product/service] to a friend or colleague?”
| Category | Score | Profile |
|---|---|---|
| Promoters | 9 to 10 | Loyal, enthusiastic customers who actively recommend |
| Passives | 7 to 8 | Satisfied but not engaged, vulnerable to competitors |
| Detractors | 0 to 6 | Dissatisfied customers who may damage your reputation |
Formula: NPS = % Promoters – % Detractors
NPS can be negative. A score above 0 is positive. A score above 30 is generally considered good; above 50 is excellent. What matters most is the trend over time and comparison with your industry.
CSAT measures satisfaction after a specific interaction: a support call, a purchase, an onboarding. It is a transactional metric that captures customer sentiment right at the moment of contact.
The question: “How satisfied are you with [interaction/product/service]?”
Responses are collected on a 1 to 5 (or 1 to 10) scale. Positive responses are scores of 4 and 5 on a 5-point scale.
Formula: CSAT (%) = (Positive responses / Total responses) × 100
A CSAT of 80% means 8 out of 10 customers declared themselves satisfied with that specific interaction. For a deeper dive into calculation, benchmarks, and improvement levers, see our customer experience resources.
Introduced in 2010 by the Corporate Executive Board following a study published in the Harvard Business Review, the Customer Effort Score measures how much effort a customer must exert to accomplish a task or resolve a problem. Its core hypothesis: reducing customer effort is more predictive of loyalty than creating delight.
The question: “On a scale of 1 to 7, how strongly do you agree: [Company] made it easy for me to resolve my issue.”
Formula: CES = average score across all responses
A high CES (close to 7) signals a smooth experience. A low CES identifies friction points that need attention.
| Criterion | NPS | CSAT | CES |
|---|---|---|---|
| What it measures | Loyalty and recommendation | Satisfaction after an interaction | Effort required by the customer |
| Core question | “Would you recommend us?” | “Were you satisfied with…?” | “Was it easy to…?” |
| Scale | 0 to 10 | 1 to 5 or 1 to 10 | 1 to 7 |
| Recommended frequency | Quarterly / semi-annual | After each key interaction | After a process or task |
| Horizon | Long-term, relational | Short-term, transactional | Operational, functional |
| Predicts | Growth and word of mouth | Immediate satisfaction level | Behavioral loyalty through ease |
| Main limitation | Does not explain the reasons | Does not predict future loyalty | Does not capture the emotional dimension |
| NPS score | Level | What it means |
|---|---|---|
| > 70 | Excellent | Very loyal customers, active brand advocates |
| 50 to 70 | Strong | High satisfaction and solid loyalty |
| 30 to 50 | Acceptable | Average satisfaction, significant room for improvement |
| 0 to 30 | To improve | Customer churn risk to watch closely |
| < 0 | Critical | More detractors than promoters, urgent action needed |
| CSAT score | Level | Recommended action |
|---|---|---|
| > 85% | Excellence | Build on loyalty and recommendations |
| 75 to 85% | Good | Maintain and refine residual friction points |
| 60 to 75% | Satisfactory | Identify irritants, targeted action plans |
| < 60% | Critical | Immediate corrective actions, full diagnostic |
| CES score (1 to 7 scale) | Effort level | Impact on experience |
|---|---|---|
| 6 to 7 | Very low effort | Smooth experience, strong correlation with loyalty |
| 4 to 6 | Moderate effort | Standard experience, some friction points exist |
| < 4 | High effort | Major friction points, urgent optimization required |
Interpreting a score without an industry reference is like reading a map without a scale. A NPS of 35 is outstanding in telecom, average in retail, and below par in B2B insurance. Context defines performance.
According to Retently’s 2025 NPS Benchmark Report, based on a sample of over 10,000 surveys per industry, the global average NPS across all sectors is 32. Industry averages vary significantly:
| Industry | Average NPS (2025) | Key driver |
|---|---|---|
| Technology and Services | 66 | Product integration, innovation, UX |
| Retail and E-commerce | 59 | Delivery speed, returns, service consistency |
| Hospitality | 44 | Personalization, repeat experience |
| Banking and Financial Services | 41 | Trust, issue resolution, digital ease |
| B2B Software / SaaS | 40 | Onboarding quality, support responsiveness |
| Telecom | 19 | Billing friction, churn, contract complexity |
B2C companies average an NPS of 49 versus 38 for B2B, a structural 11-point gap driven by transaction simplicity and relationship complexity. According to Bain & Company, companies that lead their industry on NPS grow revenues roughly twice as fast as their competitors.
According to the American Customer Satisfaction Index (ACSI) and Zendesk’s industry benchmarking data, a CSAT score between 75% and 85% is generally considered good across most sectors. In competitive industries such as SaaS and e-commerce, top-performing teams target 90% or above. Survey response rates typically range from 20% to 30%; above 40% is considered strong, and below 15% signals the need to revisit timing, channel, or survey length.
CES lacks universal cross-industry benchmarks because many platforms use different scales. According to Gartner, scores below 70% (on a percentage-based version) indicate areas for improvement, while scores above 90% signal a strong position. On the standard 1 to 7 scale, a CES above 5.5 is considered good. A CEB study found that improving CES from a low score to a mid-range score increases customer loyalty by 22%, while further gains at the top of the scale produce marginal returns.
Tracking scores without acting on them produces no ROI. Each metric responds to specific operational levers.
NPS is a lagging indicator: it reflects the cumulative result of many interactions. The three most reliable drivers are first-contact resolution (resolving issues without callbacks or transfers), reducing customer effort at every touchpoint, and closing the feedback loop with detractors directly. According to research from Lorikeet CX, teams that focus on resolution quality rather than survey optimization see NPS gains of 15 to 25 points. Segment your NPS by customer type, tenure, and channel to identify where detractors concentrate before building any action plan.
CSAT improves fastest when friction is removed at the specific touchpoints where it is measured. Start by identifying the interactions with the lowest scores, analyze the open-ended feedback attached to those ratings, and prioritize fixes that affect the highest volume of customers. In customer service contexts, FCR (First Contact Resolution) is the single strongest predictor of CSAT: according to multiple Zendesk reports, customers who have their issue resolved on the first contact score satisfaction significantly higher than those who need to contact again. Survey timing matters too: sending CSAT within minutes of the interaction produces more reliable data than end-of-week or monthly surveys.
CES improvement is primarily about removing steps, reducing wait times, and making self-service genuinely functional. Map the specific journey step where the CES is lowest and ask one operational question: what stops a customer from completing this task without contacting support? Reducing channel switches (forcing customers to repeat information across touchpoints) and improving IVR or chatbot deflection quality are the two highest-impact levers. A low CES on digital self-service, for example, often points to knowledge base gaps rather than interface issues.
Each metric has its moment. Using the wrong metric at the wrong touchpoint generates unreliable data and misdirected action plans.
| Journey stage | Recommended metrics | Objective | Example question |
|---|---|---|---|
| Discovery | CES, CSAT | Ease of finding information | “Did you easily find what you were looking for?” |
| Evaluation | CES, CSAT | Clarity of offer comparison | “Was it easy to compare our offers?” |
| Purchase | CSAT, CES | Purchase process satisfaction | “Are you satisfied with your purchase experience?” |
| Onboarding | CES, CSAT | Ease of product setup | “Was getting started straightforward?” |
| Daily use | CSAT, CES | Product or service performance | “Does the product meet your expectations?” |
| Support contact | CSAT, CES, NPS | Resolution quality and effort | “Was your issue resolved easily?” |
| Renewal | NPS, CSAT | Loyalty and overall satisfaction | “Would you recommend our company?” |
| Advocacy | NPS | Likelihood to recommend | “On 0 to 10, how likely are you to recommend us?” |
Taken separately, each metric only explains part of the picture. CSAT tells you whether the interaction went well. CES reveals the effort it took to get there. NPS steps back to assess the strength of the relationship over time, and therefore the probability that the customer will stay and recommend.
The three reinforce each other. Excessive effort (low CES) gradually weighs on satisfaction (CSAT), which in turn erodes loyalty (NPS). Conversely, a smooth and positive experience builds immediate satisfaction and future recommendation. That virtuous cycle is what cross-metric management is designed to create.
| NPS | CSAT | CES | Diagnosis | Priority action |
|---|---|---|---|---|
| High | High | Low effort | Optimal performance | Maintain and capitalize on recommendations |
| High | High | High effort | Satisfied customers but laborious journey | Simplify processes without degrading quality |
| High | Low | Low effort | Specific quality issue, relationship intact | Treat the identified irritant quickly |
| Low | High | Low effort | Adequate satisfaction but no engagement | Work on differentiation and brand experience |
| Low | Low | Low effort | Major quality problem | Full service quality audit |
| Low | Low | High effort | Customer experience crisis | Global emergency plan across all dimensions |
NPS, CSAT and CES give a satisfaction and loyalty view. To connect those data points to operational performance and return on investment, additional indicators are essential.
| KPI | What it measures | Why it matters |
|---|---|---|
| Retention rate | Share of customers retained over a period | Outcome indicator of satisfaction over time |
| Churn rate | Share of customers lost | Early warning signal to cross with CSAT and NPS |
| CLV (Customer Lifetime Value) | Total revenue generated by a customer | Orients investment toward high-value customers |
| FCR (First Contact Resolution) | Share of requests resolved on the first contact | Strongly correlated with CSAT and CES improvement |
| AHT (Average Handling Time) | Average time to handle a contact | Must be combined with FCR to avoid perverse effects |
| Direct recommendation rate | Share of new customers acquired through referral | Concrete measurement of NPS impact on growth |
A low FCR, for example, generates repeat contacts that degrade CES and CSAT before NPS even registers the impact. Anticipating these cascading effects is the hallmark of structured CX management. Our guide on outsourced contact center KPIs covers how to integrate these metrics into SLAs and performance dashboards.
These three metrics sit at the heart of any serious outsourcing arrangement. They are among the most common contractual commitments between a company and its BPO provider: without defined targets for CSAT, FCR and NPS, there is no shared basis for evaluating performance or triggering action plans.
At Armatis, tracking these metrics is embedded in the operational dashboards shared with each client, with real-time monitoring and differentiated action plans by interaction type and team. The objective is to make customer satisfaction an operational data point, not a monthly reporting exercise. To understand how to structure this measurement framework, see our guide on the essential KPIs in an outsourced contact centre.
NPS, CSAT and CES are not interchangeable: each answers a precise question about a different dimension of customer experience. CSAT tells you whether the interaction went well. CES tells you whether it was easy. NPS tells you whether the customer will come back and talk about you.
Smart CX management rests on three principles: measure each metric at the right moment in the journey, cross-reference them to identify the real causes of gaps, and act before a friction point becomes a customer loss. For support on structuring this measurement framework in your organization, our teams are available to discuss your challenges.
What is the difference between NPS and CSAT?
NPS measures the likelihood that a customer will recommend your company: it is a long-term loyalty and relationship indicator. CSAT measures satisfaction after a specific interaction: it is a transactional, immediate metric. A customer can have a high CSAT after a successful support call and a low NPS if their overall relationship with the brand does not satisfy them.
Should you choose between NPS, CSAT and CES, or use them together?
Using them together is recommended. Each illuminates a dimension the others do not cover: immediate satisfaction (CSAT), journey smoothness (CES), and long-term loyalty (NPS). Separately, they only tell part of the story. Combined, they allow precise diagnosis and prioritized action. According to CustomerGauge research, 49% of NPS users also measure an additional metric, most commonly CSAT.
What is a good NPS score by industry?
It depends on the sector. According to Retently’s 2025 benchmarks, the global average NPS is 32. Technology and services companies average 66; retail and e-commerce reach 59; hospitality and banking sit around 41 to 44; telecom averages 19. Always benchmark within your industry before drawing conclusions from an absolute score.
How often should you measure NPS?
NPS is typically measured quarterly or semi-annually. Too frequently, it generates respondent fatigue and loses reliability. It can also be triggered at key relationship milestones: after several months of the relationship, before renewal, or following an offer change.
Is CES really more predictive of loyalty than CSAT?
According to the Corporate Executive Board’s research, low customer effort is indeed more predictive of behavioral loyalty than high satisfaction. A satisfied customer may leave if they find an easier alternative. A customer who resolved their problem without effort has far fewer reasons to look elsewhere. The two metrics are complementary rather than competing.
How do you integrate these metrics into an outsourcing contract?
CSAT and NPS targets are the most common inclusions in BPO SLAs. It is recommended to define a minimum monthly threshold per metric, a measurement method agreed by both parties, and a formal review cadence. Adding FCR alongside CSAT prevents perverse effects where a provider improves handling time at the expense of resolution quality.
Armatis is one of Europe’s leading providers of business process outsourcing (BPO) services in the field of customer experience. For over 35 years, it has supported large enterprises and SMEs in managing and transforming their customer service operations. With a presence in France, Tunisia, Portugal, Poland, Madagascar, and Germany, the group combines industry expertise, a multi-site European presence, and cutting-edge technology integration to meet the demands of European and international markets.
Contact our teams to discuss your challenges.
Join the leaders who trust our multilingual and technological expertise.