Customer satisfaction metrics: what to track when your score says everything is fine

Customer satisfaction metrics quantify how customers feel about a product, service, or interaction. Each metric converts subjective experience into a comparable number, usually collected through a short survey sent close to the moment being measured.

There is no fixed list of these metrics, and any article claiming a definitive number has made an editorial choice rather than a factual one. What exists is four categories of signal, each answering a different question.

Fourteen measures cover almost every satisfaction program:

That distinction matters more than the acronyms suggest. A metric that measures sentiment will not tell you why a customer left, and a metric that measures behavior will not tell you what frustrated them. Choosing badly means collecting data that cannot answer the question you are being asked.

Satisfaction measurement carries real commercial weight. SurveyMonkey research found that 96% of consumers are more likely to purchase from a company with a reputation for good customer experience, and 57% say price and customer experience carry equal weight in purchase decisions. A score that misreads how customers feel is therefore misreading a purchase driver.

There is also a documented gap between how satisfied companies believe their customers are and how those customers actually feel. SurveyMonkey research found that 49% of CX professionals believed customer satisfaction had improved over the previous six months, while only 18% of consumers agreed and 53% said it had gotten worse.

A team can hold a healthy-looking score and still be losing customers, because the score is measuring the wrong moment, reaching the wrong people, or asking a question too vague to act on.

Greyhound ran directly into the last of these. Its original post-trip survey ran 57 questions, took an hour to complete, and finished with a completion rate below 18%. The reports were long enough that station managers stopped using them, and the survey was eventually shelved. After the company rebuilt it as a five-question survey, completion reached 94% and its Net Promoter Score rose by nearly 15 points within a few months.

Three scores dominate satisfaction programs, and they measure different things.

CSAT measures satisfaction with one specific interaction. Customers rate satisfaction on a scale, most commonly 1 to 5, and the standard calculation counts only the top two responses:

(Number of responses rating 4 or 5 / Total responses) x 100 = CSAT %

If a store collects 100 responses and 82 rate their experience a 4 or 5, CSAT is 82%. SurveyMonkey guidance treats a score above 90% as an indication that most customers are highly satisfied, and 75% to 89% as satisfied with room to improve. Check your figures with the CSAT calculator, and the CSAT score guide covers the alternative averaging method.

NPS measures loyalty rather than satisfaction, asking how likely someone is to recommend you on a scale of 0 to 10. Promoters score 9 or 10, passives score 7 or 8, and detractors score 6 or below. Passives are excluded entirely:

% Promoters − % Detractors = NPS

If half your respondents are promoters and 10% are detractors, NPS is 40. Scores run from −100 to +100, and above +50 is considered excellent. The NPS calculator runs the arithmetic and NPS benchmarks provide comparison points.

Because NPS measures the relationship rather than a moment, run it on a schedule. Transactional and relational NPS explains why running only one leaves a blind spot.

CES measures how hard something was rather than how it felt, asking customers to rate the ease of completing a task. Scores are reported as an average of those ratings. Effort is often the most directly actionable of the three, because a low score usually points at a specific process rather than a general mood. The CES guide and CES survey template cover the mechanics.

The three scores above are all self-reported, capturing stated attitude rather than action. Behavioral metrics close that gap.

Churn rate is the share of customers who leave during a period: subtract your end-of-period count from your start-of-period count, divide by the starting count, and multiply by 100. Retention rate is its inverse, so 15% churn means 85% retention. Repeat purchase rate is the share who buy more than once, and it often moves before a satisfaction score does. Customer lifetime value is the total value a customer delivers across the whole relationship, expressed as a dollar amount.

These are the metrics that connect satisfaction to revenue, which matters because a satisfaction score alone rarely survives a budget conversation. They also explain the most common confusion in satisfaction measurement, which is a high score sitting alongside rising churn. Satisfaction is a lagging indicator of how an interaction went, not a leading indicator of what someone will do at renewal.

Operational metrics measure your own performance rather than customer opinion. Common examples include first response time, tickets resolved, ticket reopen rate, and cost per resolution.

Reading these alongside survey data is where diagnosis happens. A falling satisfaction score next to a rising reopen rate tells a clear story, while the same score with no operational context tells you nothing you can assign to anyone. Our breakdown of customer service metrics and rating scales covers these in depth.

Not every satisfaction signal is a score, and the ones that are not are often the most specific. Open-ended responses are where a rating becomes diagnosable, reviews capture unfiltered sentiment from people motivated enough to post, and complaints surface urgent problems faster than any survey cycle.

Treating these as anecdote rather than data is a common mistake. Grouped into themes, they are the only source that tells you why a metric moved.

Most programs run two or three survey scores, one behavioral metric, and one operational metric. The question to start from is what decision the number will inform.

Most teams do not need to design a satisfaction measurement program from scratch. Expert-built survey templates cover the standard metrics and question wording, which removes the two decisions most likely to introduce bias into a first round of data.

Three starting points cover the majority of use cases:

Once responses arrive, industry and global benchmarks put your number in context, and text analysis groups open-ended comments into themes so you are reading patterns rather than individual replies.

The teams that measure satisfaction well are not the ones tracking the most metrics. They are the ones who picked metrics tied to a real decision, collected them close to the moment that mattered, and paired every score with the behavioral and operational data that explains it.

SurveyMonkey gives you the templates, scales, benchmarks, and analysis features to do that without a research background. Explore customer satisfaction solutions to see how teams turn satisfaction data into decisions their leadership acts on.

NPS, Net Promoter & Net Promoter Score are registered trademarks of Satmetrix Systems, Inc., Bain & Company and Fred Reichheld.