NPS vs CSAT
How to calculate NPS and CSAT, and when to use each, with a free calculator for both scores.
NPS asks whether customers would recommend you; CSAT asks whether one experience satisfied them. They answer different questions, so the two numbers can disagree.
Key Takeaways
- •NPS asks how likely customers are to recommend you on a 0 to 10 scale (Bain & Company): NPS = % promoters (9 and 10) minus % detractors (0 to 6), so it can run from −100 to +100.
- •CSAT asks how satisfied customers were with one experience: on a 1 to 5 scale it is the share of 4 and 5 answers, from 0% to 100% (Qualtrics).
- •The two scores answer different questions and can't be converted, so a company can have 85% CSAT and an NPS of +10 from the same 200 customers.
- •Our view: use CSAT right after a specific moment such as a support case, and NPS once or twice a year for the whole relationship.
- •From 100 answers, an NPS of +30 could really be anywhere from about +15 to +45. Industry NPS averages in Retently's 2026 benchmark range from 26 to 68.
What is the difference between NPS and CSAT?
NPS and CSAT differ in what they ask. NPS (Net Promoter Score) asks whether a customer would recommend you, so it tracks the whole relationship. CSAT (customer satisfaction score) asks how satisfied a customer was with one experience, such as a support case or a delivery. NPS runs from −100 to +100; CSAT is a percentage.
| NPS | CSAT | |
|---|---|---|
| What it measures | Whether customers would recommend you | How satisfied customers were with one experience |
| The question | How likely are you to recommend us to a friend or colleague? | How satisfied were you with [this experience]? |
| Scale | 0 to 10 | 1 to 5 in Qualtrics; Zendesk also offers 1 to 2 and 1 to 3 |
| Who counts | Promoters 9 and 10, detractors 0 to 6 | Satisfied answers: 4 and 5 on a 1 to 5 scale |
| Formula | % promoters − % detractors | Satisfied answers ÷ all answers × 100 |
| Range | −100 to +100 | 0% to 100% |
| When to ask (our advice) | Once or twice a year, about the whole relationship | Right after a purchase, a support case or onboarding |
The NPS question and groups follow Bain & Company's definition. The CSAT column shows the setup Qualtrics describes; other tools offer other scales (see below). The "when to ask" row is our advice.
How do you calculate NPS and CSAT?
You calculate NPS by subtracting the share of detractors from the share of promoters. Bain defines promoters as people who answer 9 or 10 on the 0 to 10 scale, and detractors as people who answer 0 to 6. Passives (7 and 8) count in the total but in neither group.
NPS = % promoters − % detractors
You calculate CSAT by dividing the satisfied answers by all answers and multiplying by 100. On the usual 1 to 5 scale, satisfied means a 4 or a 5. So 170 satisfied answers out of 200 give a CSAT of 85%.
CSAT = (answers of 4 or 5 ÷ all answers) × 100
Enter your own counts to get both scores and how far each could be off by chance. For a full NPS check with a before and after comparison, use the NPS calculator; for satisfaction alone, the CSAT calculator.
Calculate both scores
Your NPS appears here once you enter at least one count.
Your CSAT appears here once you enter at least one rating count.
The calculator runs in your browser and sends nothing you type.
Can CSAT be high while NPS is low?
CSAT can be high while NPS is low, because satisfied is a lower bar than would recommend. Take a worked example with round numbers. A software company asks 200 customers after their support case how satisfied they were: 170 answer 4 or 5, a CSAT of 85%. In its relationship survey, the same 200 customers split into 70 promoters, 80 passives and 50 detractors, an NPS of +10.
Both numbers can be true at once. The support team solved the case, so the CSAT question gets a 4 or a 5. The NPS question asks about everything else too: price, the product, the last outage. A customer who answers 7 or 8 may well be satisfied and still counts as a passive, and passives lower the share of promoters without counting as detractors. SurveyMonkey puts the same point simply: a satisfied customer is not necessarily a loyal advocate.
Which scale does CSAT use?
CSAT uses whichever scale your survey tool is set to, so check it before you compare scores. Qualtrics says CSAT surveys are always marked on a 1 to 5 scale, but Zendesk supports 1 to 2, 1 to 3 and 1 to 5 scales and maps each answer to good or bad:
| Scale | Counts as satisfied | Source |
|---|---|---|
| 1 to 5 | 4 and 5 ("satisfied", "very satisfied") | Qualtrics |
| 1 to 5 (Zendesk) | 4 and 5 count as good, 1 to 3 as bad | Zendesk |
| 1 to 3 (Zendesk) | 3 counts as good, 1 and 2 as bad | Zendesk |
| 1 to 2 (Zendesk) | 2 counts as good, 1 as bad | Zendesk |
Three mistakes make scores misleading or hard to compare:
- Averaging the answers. An average of 4.1 hides how many people were unhappy, yet FeverBee's guide works CSAT out as an average and says NPS is calculated the same way. Count the share of 4s and 5s instead.
- Changing the scale mid-series. 85% on a 1 to 5 scale and 85% on a 1 to 3 scale are different numbers. Keep one scale, or start a new series.
- Mixing up CES. In this article CES means customer effort score, but the same FeverBee guide uses it for a customer sentiment score and describes "recommend" questions on a 1 to 10 scale. Bain's NPS scale starts at 0.
Can you convert NPS into CSAT?
You can't convert NPS into CSAT or back. The two scores come from different questions on different scales, so there is no formula that turns one into the other. Qualtrics says the same: one can't be extrapolated from the other. Rescaling a 0 to 10 answer to 1 to 5 doesn't help either, because the customer answered a question about recommending.
What is a good NPS and a good CSAT score?
SurveyMonkey calls any NPS above 0 generally positive, because then promoters outnumber detractors, and a score above 50 excellent. Retently calls 0 to 30 a good range. Its 2026 benchmark, built from its own customers' surveys, reports industry averages from 26 to 68, including these:
| Industry | Average NPS, 2026 |
|---|---|
| Financial services | 68 |
| Ecommerce | 61 |
| Insurance | 46 |
| B2B software and SaaS | 41 |
| Healthcare | 37 |
| Internet software and services | 26 |
For CSAT we found no public benchmark with a named data source. SurveyMonkey says a good CSAT typically exceeds 80%, without naming the data behind that number or its NPS bands. Our advice for both scores: compare with your own last period on the same scale, which tells you more than another company's average.
Does NPS predict growth better than satisfaction?
Whether NPS predicts growth better than satisfaction is disputed. Bain states that sustained value creators have Net Promoter Scores two times higher than the average company. An independent test found less: Keiningham and colleagues used data from 21 firms and more than 15,500 interviews in the Norwegian Customer Satisfaction Barometer and, as Ipsos summarises the study, failed to replicate the "clear superiority" of Net Promoter in the industries Reichheld had cited.
Our view: choose each score by the question your team will act on, because the evidence on revenue doesn't settle it.
Where does customer effort score (CES) fit?
We use CES (customer effort score) to mean a measure of how easy it was to get something done, such as solving a problem with support. It sits next to CSAT as a measure of one interaction. The case against chasing delight is made in "Stop Trying to Delight Your Customers" by Matthew Dixon, Karen Freeman and Nick Toman (Harvard Business Review, 2010).
How many answers do you need?
You need a few hundred answers before small changes in either score mean much. The table shows the 95% margin of error for an example NPS (50% promoters, 20% detractors, so +30) and an example CSAT (80% satisfied). A score of +30 from 100 answers could really be anywhere from about +15 to +45.
| Answers | NPS margin (points) | CSAT margin (points) |
|---|---|---|
| 50 | ±22 | ±11 |
| 100 | ±15 | ±8 |
| 200 | ±11 | ±6 |
| 400 | ±8 | ±4 |
| 1,000 | ±5 | ±2 |
NPS swings more than CSAT at the same number of answers, because it subtracts two shares that both carry error and runs on a scale twice as wide. To test whether an NPS change between two surveys is real, use the "Did my NPS really change?" check in our NPS calculator, which follows MeasuringU's method.
Should you use NPS or CSAT?
Use CSAT when you want to fix a specific moment, and NPS when you want to track the relationship. Our view at cx.tools: if you can run only one survey, ask the CSAT question right after the moments that decide whether customers stay (the first support case, onboarding, a delivery), because a low score there points at something a team can fix this week. Add NPS once or twice a year, and read it next to churn.
13 of the 64 tools in the cx.tools directory list NPS, CSAT and CES surveys as a feature, so one of these tools can run all three surveys. Compare them in feedback and survey tools, or read our wider guide to which CX metrics matter.
How this was checked
- Bain & Company, "Measuring Your Net Promoter Score"
- Qualtrics, "CSAT vs NPS" (27 April 2022)
- Zendesk help, "Viewing your CSAT (customer satisfaction) score and ratings"
- SurveyMonkey, "CSAT vs NPS" (updated 9 November 2026)
- Retently, "What is a Good Net Promoter Score? (2026 NPS Benchmark)" (15 April 2026)
- Lewis and Sauro, "How to Statistically Compare Two Net Promoter Scores" (MeasuringU, 2 February 2021)
- Ipsos summary of Keiningham, Cooil, Andreassen and Aksoy, "A Longitudinal Examination of Net Promoter and Firm Revenue Growth" (Journal of Marketing, July 2007)
- FeverBee, "How to Calculate CSAT, NPS & CES Scores (With Formulas)"
- Dixon, Freeman and Toman, "Stop Trying to Delight Your Customers" (Harvard Business Review, July to August 2010)
- All sources were opened on 22 November 2026. Margins of error use the unadjusted variance formula from MeasuringU for NPS and the normal approximation for a share for CSAT, both at 95% confidence.
- Left out because the original could not be opened: the "96% of high-effort customers become disloyal" figure (Gartner's page refused automated reading) and the ACSI national index.
- The tool count comes from the live cx.tools catalogue at the time this page was built.
- The HBR article is behind a paywall, so we cite it only for its title, authors and argument.
cx.tools drafted this page with AI, and we checked every sourced number against the sources named above.
Tools that run NPS, CSAT and CES surveys
These tools in the directory list NPS, CSAT and CES surveys as a feature:
Alchemer
Customer feedback platform for surveys, reviews, social and listings data, with AI analysis and integrations.
Buyer fit
Mid-Market, Enterprise teams that need Sentiment Analysis and Email Distribution coverage.
Key Features
- Sentiment Analysis
- Email Distribution
- In-App / In-Product Surveys
- Multi-Channel Distribution
Qualtrics XM
Enterprise experience management platform with advanced survey capabilities, real-time analytics, and AI-driven insights.
Buyer fit
Enterprise, Mid-Market teams that need AI-Powered Text Analysis and Customer Segmentation coverage.
Key Features
- AI-Powered Text Analysis
- Customer Segmentation
- Robust Statistical Analysis
- Email Distribution
SurveyMonkey
Survey platform with a drag-and-drop builder and templates, used by businesses of all sizes.
Buyer fit
Small Business, Mid-Market, Startup teams that need Email Distribution and Multi-Channel Distribution coverage.
Key Features
- Email Distribution
- Multi-Channel Distribution
- Multilingual Support
- API Integration
Survicate
A customer feedback platform for collecting and analysing feedback across the customer journey. Built for product, UX, and CX teams capturing NPS, CSAT, and in-product surveys at scale, and turning multi-source feedback into stakeholder-ready research reports.
Buyer fit
Mid-Market, Small Business, Startup teams that need AI-Assisted Research Synthesis and AI-Powered Text Analysis coverage.
Key Features
- AI-Assisted Research Synthesis
- AI-Powered Text Analysis
- Behavioural Targeting
- Customer Segmentation
Zonka Feedback
Multichannel feedback platform with real-time dashboards, AI-driven insights, and automated ticketing for low scores.
Buyer fit
Mid-Market, Enterprise teams that need AI-Powered Text Analysis and Sentiment Analysis coverage.
Key Features
- AI-Powered Text Analysis
- Sentiment Analysis
- Automated Alerts
- Closed-Loop Feedback Management
InMoment XI Platform
VoC and experience improvement platform providing data collection across channels with text analytics and journey analytics.
Buyer fit
Enterprise, Mid-Market teams that need AI-Powered Text Analysis and Churn Risk Detection coverage.
Key Features
- AI-Powered Text Analysis
- Churn Risk Detection
- Emotion Detection
- Predictive Analytics
Frequently Asked Questions
How do you calculate CSAT and NPS?
Calculate NPS as the percentage of promoters (9 and 10 on a 0 to 10 scale) minus the percentage of detractors (0 to 6). Calculate CSAT as the satisfied answers (4 or 5 on a 1 to 5 scale) divided by all answers, times 100.
Is NPS the same as customer satisfaction?
NPS is not the same as customer satisfaction. NPS asks whether customers would recommend you; CSAT asks how satisfied they were with one experience. A satisfied customer may still answer 7 or 8 on the NPS question, which makes them a passive.
What is a good NPS score?
SurveyMonkey calls any NPS above 0 generally positive, because then promoters outnumber detractors. Retently's 2026 benchmark, built from its own customers' surveys, reports industry averages from 26 to 68.
Should I use NPS or CSAT?
Use CSAT to find and fix a specific moment, such as a support case, and NPS to track the relationship as a whole. If you can run only one survey, our view at cx.tools is to start with CSAT.
Can you convert NPS to CSAT?
You can't convert NPS to CSAT or back. The scores come from different questions on different scales, so no formula turns one into the other.