4.1Satisfaction metrics (NPS, CSAT, CES)
The essentials in 30 seconds
- NPS (“would you recommend us?” 0-10) = loyalty; CSAT (“satisfied with this interaction?”) = the transaction; CES (“was it easy?”) = effort.
- NPS = % promoters (9-10) − % detractors (0-6): an NPS of +30 is good, +70 exceptional — always with the response rate and verbatims.
- Reflex: start with post-interaction CSAT + CES (quickly actionable), add quarterly NPS (trend) — each with an owner + target.
Threaded mini-case — Rania finally measures her satisfaction
Rania runs an online shop: sales hold, but customers do not return and she does not know why. She decides to MEASURE: CSAT after each delivery, CES on journey ease, NPS each quarter. First shock: decent CSAT, but high effort and harsh verbatims on returns. We will measure with her, step by step.
This same case returns in the lesson, the diagram, the script and the 48h challenge — and continues in 4.2, 4.3, 4.4 and 4.5.
1. Three metrics, three questions
The NPS (“would you recommend us?”, scored 0 to 10) measures loyalty and recommendation; the CSAT (“satisfied with this interaction?”) measures one-off transactional satisfaction; the CES (“was it easy?”) measures the effort the customer expended. V2 image: the NPS is the compass, the CSAT the thermometer, the CES the churn predictor — because effort drives away more than mistakes.
Analogy: CSAT says whether yesterday's meal was good, NPS whether you will recommend the restaurant, CES whether you had to wait an hour to be served. Rania discovers her “satisfied” (CSAT) customers struggle with returns (CES) — and do not recommend (NPS).
2. Read honestly, then steer with one well-tracked metric
NPS is computed: % promoters (9-10) − % detractors (0-6); an NPS of +30 is good, +70 exceptional. But a score is only worth reading with its response rate (who replies? often the extremes!) and its verbatims: numbers alert, words explain. An NPS without verbatims is a thermometer with no diagnosis.
Better one well-tracked metric than three decorative ones: Rania starts with post-interaction CSAT + CES (quickly actionable), adds quarterly NPS (trend), and gives each an owner and a target — without both, it is a poster.
🔎 Concrete example — The CES that reveals the bleed (Rania's case)
CSAT 85% (good) but high effort + verbatims “3 calls for…”
→ problem: effort, not quality → simplified journey
→ 6 months: CES 2.1, churn -30%.Result: CSAT masked, CES revealed — Rania knows where to act.
Lesson: measuring effort predicts churn better than satisfaction alone.
Section vocabulary
- CSAT
- Customer Satisfaction Score: satisfaction measured “hot” right after an interaction or purchase.
Simple example: Rania asks “satisfied with your delivery?” right after receipt.
What it is not: a measure of loyalty or recommendation (that is NPS).
Reflex: always pair the score with a “why?” open field.
- NPS
- Net Promoter Score: likelihood to recommend (0 to 10); splits promoters / passives / detractors, close to loyalty.
Simple example: Rania tracks her NPS each quarter: +30 good, +70 exceptional.
What it is not: a verdict on a single interaction (that is CSAT).
Reflex: read it with response rate + verbatims, never alone.
- CES
- Customer Effort Score: effort the customer had to expend to get satisfaction; reducing effort builds strong loyalty.
Simple example: “Was it easy to return the item?” reveals Rania's true irritant.
What it is not: “delighting” the customer (CES measures ease).
Reflex: hunt every imposed effort: waiting, repetition, complexity.
- Detractor / Promoter
- In NPS: detractor (low score, dissatisfied, risk of harm) vs promoter (high score, ambassador).
Simple example: Rania calls back each detractor within 48 hours; she thanks her promoters.
What it is not: two marketing segments to target alike (one alerts, one carries).
Reflex: every detractor deserves a callback, every promoter a thank-you.
- Verbatim
- Open-ended comment left by a customer in a survey, often richer than the score itself.
Simple example: “3 messages for one return” is worth ten scores to steer action.
What it is not: noise to ignore next to numbers (it is the diagnosis).
Reflex: read the verbatims before deciding any action.
In practice — Hands-on - Use satisfaction metrics
- Measure CSAT (hot satisfaction) right after an interaction or purchase.
- Measure NPS (likelihood to recommend, 0-10: promoters / passives / detractors) as a proxy for loyalty.
- Measure CES (customer effort): reducing effort often retains more than delighting.
- Interpret scores (trend, target), mine verbatims, and turn them into improvement actions.
Ready-to-say script — announcing measurement to the team (1 min)
“We will finally measure instead of guessing. After each delivery: one satisfaction question, one ease question. Each quarter: would you recommend us? Every number will have an owner and a target — and every comment will be read. The goal is not the score, it is what we will fix.”
Say in a team huddle, then display the three questions and their owners on the wall.
Copy-paste template — metrics card
METRICS CARD — WEEK OF [date]
CSAT (satisfied?): question [__] — owner [name] — target [__]
CES (easy?): question [__] — owner [name] — target [__]
NPS (recommend? 0-10): rhythm [quarterly] — owner [name] — target [__]
RESPONSE RATE: [__] % — VERBATIMS READ: [yes / no]
DECIDED ACTION: [what] — WHO: [name] — BY WHEN: [date]Small business (2-3 people)
Two questions after each sale (satisfied? easy?) in a notebook are enough: reread them every Friday.
Remote / online
Post-delivery questionnaire by email or SMS, mobile-first: 2 questions, 30 seconds, open “why?” field.
Physical shop
Mini-terminal or QR code at checkout: hot CSAT on the way out, verbatims read every evening by the manager.
Beginner at measuring
Start with post-interaction CSAT + verbatims alone for one month before adding CES then NPS.
✓ Best practices
- Separate the 3 roles: NPS loyalty, CSAT transaction, CES effort.
- Attach verbatims + response rate to every number.
- Give each metric an owner and a target.
✗ Common pitfalls
- Steering by NPS alone, without verbatims.
- Ignoring the response rate (who really replies?).
- Multiplying metrics with no owner or action.
Key takeaways
- CSAT: hot satisfaction; NPS: recommendation / loyalty; CES: customer effort.
- NPS = % promoters (9-10) − % detractors (0-6); +30 good, +70 exceptional.
- Reducing effort often retains more than delighting.
- No metric alone is enough: numbers alert, words explain, action counts.
48h challenge — apply tomorrow
Send this week 2 questions after each interaction (“satisfied?” + “easy?” + “why?”), read every verbatim Friday, and decide 1 action. 20 minutes is enough.
- I separate NPS, CSAT and CES and state each role.
- I can compute an NPS and support it (rate + verbatims).
- I did the 48h challenge or scheduled its date.
Frequently asked questions
What is the difference between CSAT, NPS and CES, and which one should I choose?
They measure different, complementary things: CSAT one-off “hot” satisfaction after an interaction, NPS likelihood to recommend (close to loyalty), CES effort expended by the customer. The choice depends on what you steer: CSAT for interactions, NPS for overall loyalty, CES for friction. Many combine CSAT after each interaction + periodic NPS. What matters is acting on what they reveal.
Is CES really more important than satisfaction?
Not in absolute terms, but reducing effort is a particularly powerful loyalty lever: customers mostly want things to be easy. Waiting, transfers, re-explaining, complexity — each effort erodes the relationship even if the problem is solved. Before trying to delight, do not complicate the customer's life: the best experience is often the simplest.
Rania shows a decent CSAT but her customers do not return; verbatims mention complicated returns. Which metric is she missing to understand?