Last updated: July 2026
Customer experience metrics fall into three families: perception metrics that ask how customers feel (NPS, CSAT, CES), loyalty and value metrics that show what they do next (retention rate, churn rate, net revenue retention, customer lifetime value, health score), and operational metrics that explain why (first response time, first contact resolution, average resolution time). The rule that keeps a scorecard honest is simple: never track a feeling metric without the operational metric that moves it, because a score tells you there is a problem and the operational number tells you where.
Most companies have too many CX metrics and too little idea what any single one is supposed to change. They report an NPS every quarter, watch it drift, and write a deck about it. The reason that pattern never improves anything is that NPS is a symptom. It moves because something operational moved underneath it, and if you are not watching the operational number too, you are reading a thermometer with no idea where the fever is.
This is the full list, organized by the question each metric answers rather than by acronym. If you want the discipline that sits above the numbers, start with the customer experience strategy framework; this page is the measurement layer underneath it.
The complete list of customer experience metrics
Here is every metric worth tracking, what it answers, and where each one is covered in depth.
| Metric | Family | Question it answers |
|---|---|---|
| Net Promoter Score (NPS) | Perception | Would customers recommend you? |
| Customer Satisfaction (CSAT) | Perception | Were they happy with this specific interaction? |
| Customer Effort Score (CES) | Perception | How hard was it to get what they needed? |
| Retention rate | Loyalty and value | What share of customers stayed? |
| Churn rate | Loyalty and value | What share left, and how fast? |
| Net revenue retention (NRR) | Loyalty and value | Is the existing base growing or shrinking in dollars? |
| Customer lifetime value (CLV) | Loyalty and value | What is a customer worth over the whole relationship? |
| Customer health score | Loyalty and value | Which accounts are about to leave? |
| First response time (FRT) | Operational | How fast do we reply? |
| First contact resolution (FCR) | Operational | Do we fix it on the first try? |
| Average resolution time | Operational | How long until the problem is actually solved? |
| Average handle time (AHT) | Operational | How much agent time does each contact consume? |
Customer experience metrics vs customer service metrics
These two lists get merged, and merging them is why so many scorecards are noise. Customer experience metrics measure the whole relationship: how a customer feels about you and whether they stay and spend. Customer service metrics measure one channel, the support desk: how fast and how well you handle inbound requests. The support numbers are a subset, not the whole picture. A company can have excellent customer service metrics, fast replies and high first contact resolution, and still lose customers because the product is slow to deliver value or the first invoice is always wrong. Service metrics are operational inputs. Experience metrics are the outcome. Keep them on the same scorecard, but never mistake one for the other.
There is a third layer worth naming: the operations that produce the experience in the first place. Onboarding cycle time, billing accuracy, and contract turnaround are not on most CX dashboards, yet they move every perception metric on this page. That layer is covered separately in the guide to the operational metrics that predict churn.
The perception metrics: NPS, CSAT, CES
These three are the ones people mean when they say "CX metrics." Each asks a different question, and using the wrong one is the most common measurement mistake. Ask NPS after a transaction and you will get a confused answer; ask CSAT about the whole relationship and you will get a vague one. The rule of thumb: CSAT for a specific interaction, NPS for the overall relationship, CES for a task you want to be effortless. The full comparison lives in the CSAT vs NPS vs CES breakdown.
Here are the 2026 benchmarks so you know whether your number is good or just familiar.
| Metric | Scale | 2026 average | Good | Excellent |
|---|---|---|---|---|
| NPS | -100 to +100 | 32 (B2B 38, B2C 49) | 30 to 50 | Above 50 |
| CSAT | 0 to 100 percent | 78 percent | 75 to 85 percent | Above 85 percent |
| CES | 1 to 7 | 5.99 | 5 or higher | 6.5 or higher |
One useful and slightly uncomfortable fact from the 2026 data: NPS explains only about a quarter of the variance in CSAT. The two genuinely measure different things, so a strong relationship score can hide a string of poor individual interactions. That is exactly why you track more than one perception metric, and why you pair all of them with the operational numbers below.
The loyalty and value metrics: what customers do, not what they say
Perception metrics ask customers what they think. Loyalty and value metrics watch what they actually do, which is usually the more honest signal. Retention rate and its mirror, churn rate, are the headline pair: the share of customers who stay versus leave over a period. For any subscription business, net revenue retention is more revealing still, because it captures expansion and contraction in dollars, not just logo counts. An NRR above 100 percent means your existing base grows even if you never sign a new customer.
Two more belong here. Customer lifetime value is what a customer is worth across the whole relationship, and it is the number that tells you how much you can afford to spend keeping them. Customer health score is the forward-looking one: a composite of usage, support, and sentiment signals that flags accounts heading for the exit while you can still do something about it. Perception metrics tell you how last quarter felt. Health score tells you which accounts will churn next quarter.
The operational metrics that move everything above
These are the levers. When a perception score drops, the cause is almost always here. First response time is the single strongest predictor of CSAT: 2026 data shows replies within an hour score 12 to 18 points higher than replies that take a day. First contact resolution is the effort metric in disguise, because nothing raises CES faster than making someone explain the same problem three times. Average resolution time measures how long the customer waits for an actual fix, and average handle time measures the cost of each contact to you. Watch AHT carefully: pushing it down while resolution rate falls just means agents are closing tickets faster without solving anything.
How to build a CX measurement stack
You do not track all twelve at once. You assign one primary metric to each stage of the journey, so every number has an owner and a decision attached to it.
- Onboarding: time to value and activation rate. The relationship is decided here more than anywhere else. Track the onboarding metrics that predict retention before you worry about a quarterly relationship survey.
- Support interactions: FRT, FCR, and a post-ticket CSAT. The operational numbers plus one perception check on the interaction itself.
- The ongoing relationship: NPS and health score. One periodic relationship read, one always-on early warning.
- The business outcome: retention, NRR, and CLV. The numbers that turn experience into money and prove the rest of the stack is worth funding.
- Close every loop. A metric that does not change what you build is a vanity number. Route each score to the team that owns the underlying work, which is the whole point of a voice of customer program.
The hard part is not collecting the numbers. It is connecting a flat percentage to the specific accounts behind it, so "activation is 38 percent" becomes a named list of stalled customers you can act on. A product analytics layer that unifies product usage with survey feedback per account is what turns a dashboard of averages into that list. Without it, you know the score dropped but not who dropped it.
Frequently asked questions
What are the main customer experience metrics? The core CX metrics are NPS, CSAT, and CES for perception; retention rate, churn rate, net revenue retention, customer lifetime value, and health score for loyalty and value; and first response time, first contact resolution, average resolution time, and average handle time for the operations underneath. Perception metrics tell you how customers feel, loyalty metrics tell you what they do, and operational metrics tell you why.
What is the difference between CX metrics and customer service metrics? Customer experience metrics measure the whole relationship, including product, billing, and onboarding, and whether customers stay and spend. Customer service metrics measure one channel, the support desk, such as response and resolution speed. Service metrics are a subset that feeds the broader experience; strong service numbers do not guarantee a strong experience if the product or billing fails elsewhere.
What is the single most important customer experience metric? There is no universal answer, but for a subscription business net revenue retention is the closest thing, because it captures whether customers stay and spend more over time. For most operational teams, pair one perception metric (usually NPS or CSAT) with one operational metric (usually first contact resolution), so you always have a feeling and its cause on the same screen.
What is a good NPS, CSAT, and CES score in 2026? A good NPS is 30 to 50, with anything above 50 excellent and above 70 world-class. A good CSAT is 75 to 85 percent against a cross-industry average of 78 percent. A good CES is 5 or higher on the 7-point scale, where the 2026 average is 5.99. Benchmarks vary widely by industry, so compare against your own sector and, more importantly, against your own trend.
Every metric on this page reduces to one test: did it change a decision? If NPS dropped and nobody knows which operational number moved with it, you have a dashboard, not a measurement program. Pick one metric per journey stage, pair every feeling with its cause, and route each number to the team that can move it. That is how measurement stops being a quarterly ritual and starts being the thing that runs customer experience operations.