Explainer

NPS vs CSAT vs CES

These three metrics answer three different questions: loyalty (NPS), satisfaction (CSAT), and effort (CES). The biggest mistake teams make is treating them as interchangeable.

Compare metrics Copy questions Build surveys
By Jordan Keane • CX Ops Writer • Published Jan 15, 2026

What each metric really measures

The easiest way to choose is to start with scope: NPS is relational (overall relationship), while CSAT and CES are usually transactional (specific touchpoints).

Metric Measures Best used when Common output
NPS Overall loyalty / likelihood to recommend. Tracking brand health over time; benchmarking; understanding promoters vs detractors. Score from –100 to +100.
CSAT Satisfaction with a specific interaction or outcome. After support, delivery, onboarding, feature usage, or purchase. Average rating or % satisfied.
CES Effort required to complete a task or resolve an issue. When friction reduction is the goal: support resolution, password reset, checkout, returns. Average ease/difficulty score.

A quick translation

  • NPS: “Do people want to recommend us?”
  • CSAT: “Were they happy with what just happened?”
  • CES: “Was it easy, or did we make them work for it?”

Copy/paste questions (good defaults)

Keep the format consistent: 1 metric question + 1 optional open-text “why?” follow-up. That’s usually enough to drive action without creating survey fatigue.

NPS question (relationship)

“On a scale of 0–10, how likely are you to recommend us to a friend or colleague?”
Follow-up: “What is the primary reason for your score?”

Use it quarterly, after major milestones, or when your product/experience changes meaningfully.

CSAT question (moment)

“How satisfied were you with [support/delivery/onboarding]?” (Very dissatisfied → Very satisfied)
Follow-up: “What could we improve?”

CSAT is best when you control a touchpoint and can make quick fixes (scripts, training, process).

CES question (friction)

“How easy was it to [resolve your issue/complete checkout/return an item]?” (Very difficult → Very easy)
Follow-up: “What made it difficult?”

CES is ideal for identifying friction you can remove with UX changes, automation, or clearer instructions.

When to use each metric (decision guide)

A practical way to choose is to start from the decision you’re making this month: product roadmap, support performance, checkout improvements, or retention strategy.

Use NPS when

  • You need a high-level loyalty trend (up/down) across months or quarters.
  • You’re comparing segments (plan tiers, regions, onboarding cohorts).
  • You want to identify promoters, passives, and detractors for different follow-up actions.

Use CSAT when

  • You need to evaluate quality at a specific touchpoint (support chat, delivery, install).
  • You need fast feedback loops for teams (support, ops, onboarding).
  • You suspect an outcome is failing (late deliveries, unresolved tickets, buggy release).

Use CES when

  • You want to reduce churn caused by “this is too hard.”
  • You want to find friction in self-serve flows (reset password, cancel plan, update billing).
  • You’re optimizing conversions (checkout, onboarding steps, activation).

How to use them together (without over-surveying)

Combining the three metrics works when each has a purpose and a trigger — not when you blast them to everyone all the time.

A simple “balanced” program

  • CES: after support resolution (every time, or sampled).
  • CSAT: after delivery/fulfillment or onboarding milestone (sampled).
  • NPS: quarterly or after key milestones (sampled, not transactional).

How to interpret contradictions

  • High CSAT + low NPS: happy with the moment, not convinced long-term (value, differentiation, trust).
  • High NPS + low CSAT: people like the brand, but a touchpoint is failing (support, delivery, billing).
  • Low CES + okay CSAT: they got the outcome, but it felt like hard work — a churn risk.

Common pitfalls (and fixes)

Most measurement programs fail because the metric isn’t connected to an owner, a workflow, or a decision.

Pitfall: Asking NPS right after a single event

If you ask NPS after every delivery or ticket, you’ll measure mood, not relationship.

Fix: use CSAT or CES at touchpoints; reserve NPS for periodic relationship measurement.

Pitfall: Collecting scores without the “why”

A number alone doesn’t tell you what to change.

Fix: add one follow-up question and tag themes (pricing, delivery, UX, support, reliability).

Pitfall: Too many surveys (fatigue)

Over-surveying reduces response quality and can annoy your best customers.

Fix: sample intelligently, set cooldown windows, and prioritize the touchpoints you can improve.

Build a CX program Reduce drop‑off Close the loop