CSAT vs NPS vs CES: Which Metric Should You Use?

Your CEO wants a customer satisfaction number for the board deck. Your VP of Product wants to know which features are failing. Your support lead wants to measure ticket resolution quality. CSAT, NPS, and CES each measure something fundamentally different, and choosing the wrong one means blind spots that cost you customers. This guide compares all three with formulas, benchmarks, and a decision framework.

The Three Metrics at a Glance

Before diving into each one, here's a side-by-side comparison:

Now let's examine each metric in detail.

What Is CSAT (Customer Satisfaction Score)?

CSAT measures how satisfied a customer is with a specific interaction, transaction, or experience. It's the most direct satisfaction metric: you ask people if they're happy, and they tell you.

The CSAT Question

"How satisfied were you with [specific experience]?"

Respondents typically answer on a 1 to 5 scale:

CSAT Formula

CSAT = (Number of satisfied responses / Total responses) x 100

"Satisfied" typically means respondents who selected 4 or 5. Some companies include 3 (neutral), but standard practice counts only the top two scores.

Example Calculation

You survey 150 customers after a support interaction:

• 45 give a 5 (Very Satisfied)
• 60 give a 4 (Satisfied)
• 25 give a 3 (Neutral)
• 12 give a 2 (Unsatisfied)
• 8 give a 1 (Very Unsatisfied)

CSAT = (45 + 60) / 150 x 100 = 70%

When CSAT Shines

Post-support interactions: "How satisfied were you with the help you received?" This tells you whether your support team meets expectations at the individual ticket level.

After onboarding: "How satisfied are you with the setup process?" A low CSAT here reveals onboarding friction before it shows up in churn numbers weeks later.

Feature-specific feedback: "How satisfied are you with our reporting dashboard?" This lets you measure satisfaction with specific product areas rather than the product overall.

When CSAT Falls Short

CSAT is inherently short-term. A customer can give your support team a 5/5 today and still churn next month because your product is missing a critical feature. CSAT captures the moment but not the relationship.

It's also susceptible to recency bias. According to Zendesk's 2024 CX Trends Report, 73% of customers say a single bad interaction can override months of positive experiences. CSAT reflects this: the most recent interaction dominates the score.

CSAT Benchmarks for SaaS

Most SaaS companies aim for 75% to 85% on post-support CSAT. The American Customer Satisfaction Index (ACSI) reports an average of 77% across the software industry.

Key takeaway: CSAT is your best metric for measuring specific touchpoints. If you only measure one thing about your support team, measure CSAT. But never use CSAT alone to judge overall product health.

What Is NPS (Net Promoter Score)?

NPS measures customer loyalty by asking how likely someone is to recommend your product. Unlike CSAT, which measures a moment, NPS attempts to measure the overall relationship.

We cover NPS in depth in our complete NPS guide, but here's the essential summary.

The NPS Question

"On a scale of 0 to 10, how likely are you to recommend [product] to a friend or colleague?"

NPS Groups

NPS Formula

NPS = % Promoters − % Detractors

Example Calculation

300 survey responses:

• 150 Promoters (50%)
• 90 Passives (30%)
• 60 Detractors (20%)

NPS = 50% − 20% = +30

When NPS Shines

Board and investor reporting: NPS is universally understood. Bain & Company reports that approximately two thirds of the Fortune 1000 use NPS, making it the most recognized customer loyalty metric in business.

Trend tracking over time: Measuring NPS quarterly reveals whether your product decisions are improving or hurting customer loyalty. A drop from +40 to +28 over two quarters is a clear signal something is wrong.

Competitive benchmarking: Because so many companies use NPS, you can compare against industry averages. Retently's 2024 benchmark data puts the average B2B SaaS NPS at 36.

When NPS Falls Short

NPS doesn't tell you what to fix. A score of +25 means you have more promoters than detractors, but it gives your product team nothing to act on. That's why NPS surveys should always include an open-ended follow-up question.

The 0 to 6 detractor range is also notoriously blunt. A score of 6 and a score of 1 carry equal weight in the formula, losing important nuance.

NPS Benchmarks for SaaS

A score between 30 and 40 is considered good for B2B SaaS. Above 50 is excellent. Below 0 means more detractors than promoters and signals an urgent need to act.

Try it yourself: Set up a feedback board alongside your NPS surveys to capture the "why" behind every score. No credit card required.

What Is CES (Customer Effort Score)?

CES measures how easy it was for a customer to accomplish a specific task. The insight behind CES is counterintuitive. Research from the Corporate Executive Board (now part of Gartner) found that reducing customer effort is a stronger predictor of loyalty than delighting customers.

Their landmark 2010 study, published in Harvard Business Review as "Stop Trying to Delight Your Customers," backs this up. They found that 96% of high-effort interactions lead to disloyalty, compared to only 9% of low-effort ones.

The CES Question

"How easy was it to [complete specific task]?"

Common variations:

• "The company made it easy for me to resolve my issue." (Agree/Disagree scale)
• "How easy was it to get started with [product]?"
• "How much effort did you put into [action]?"

CES Scale

CES Formula

CES = Sum of all scores / Total number of responses

CES is reported as an average rather than a percentage.

Example Calculation

100 customers rate the ease of resolving a support ticket:

• 20 give a 5
• 35 give a 4
• 25 give a 3
• 12 give a 2
• 8 give a 1

CES = (20x5 + 35x4 + 25x3 + 12x2 + 8x1) / 100 = 3.47

When CES Shines

Post-support ticket resolution: CES is purpose-built for this. A low CES on support interactions tells you that even when you solve the problem, the process of getting it solved is frustrating.

Onboarding flows: "How easy was it to set up your account?" Low CES during onboarding predicts higher churn within the first 90 days. A Gainsight study found that customers with high-effort onboarding are 62% more likely to churn in the first year.

Self-service interactions: "How easy was it to find the answer in our help center?" This is invaluable for measuring knowledge base effectiveness.

Checkout and upgrade flows: "How easy was it to upgrade your plan?" High-effort upgrade processes directly cost you revenue.

When CES Falls Short

CES is narrow by design. It tells you whether a specific process was easy but says nothing about overall satisfaction or loyalty. A customer can find your support process effortless (CES 4.8) while being deeply unhappy with your product's core functionality.

CES also doesn't capture emotional satisfaction. A frictionless but cold, impersonal support experience can score well on CES but poorly on CSAT.

CES Benchmarks for SaaS

Head-to-Head: The Full Comparison Table

Here's a detailed breakdown across every dimension that matters for choosing the right metric:

That last row is critical. None of these metrics inherently tell you why customers feel the way they do. They all require follow-up questions for qualitative context, and response rates on those follow-ups are always lower than on the rating itself.

Key takeaway: No single metric covers everything. CSAT measures the moment, NPS measures the relationship, and CES measures the process. The best teams use all three at the right touchpoints.

Decision Framework: Which Metric Should You Use?

The answer for most SaaS companies is: more than one, at different touchpoints.

Use CSAT When...

• You want to measure satisfaction at specific touchpoints (support, onboarding, feature usage)
• You need a quick pulse check after a particular interaction
• You're evaluating the quality of a team or process (support CSAT, onboarding CSAT)
• You want to compare satisfaction before and after a change (redesign, new workflow)

Use NPS When...

• You need a company-wide loyalty metric for leadership and investors
• You want to benchmark against competitors or industry averages
• You're tracking long-term trends in customer sentiment
• You want to identify promoters for referral and case study programs

Use CES When...

• You want to identify high-friction experiences in your product
• You're optimizing support, onboarding, or self-service flows
• You suspect customers are churning because your product is hard to use, not because it lacks features
• You want to measure knowledge base or help center effectiveness

A Practical Setup for SaaS

Here's a realistic implementation that balances coverage with survey fatigue:

This gives you loyalty tracking (NPS), touchpoint satisfaction (CSAT), and friction identification (CES) without overwhelming your users. No customer should receive more than one survey per week.

What All Three Metrics Miss: The Case for Continuous Feedback

Here's the part that doesn't get discussed enough. CSAT, NPS, and CES are all point-in-time survey metrics. They capture how a customer feels at the moment of response. They don't capture what customers think between surveys.

This creates three significant blind spots:

1. No Continuous Signal

Between quarterly NPS surveys, a lot happens. Features ship, competitors launch new products, pricing changes hit, and customer needs evolve. McKinsey's 2023 State of Customer Care report found that customer expectations change 2 to 3x faster than most companies update their measurement approaches. Survey-based metrics miss all of this.

2. Limited Qualitative Depth

Even with follow-up questions, survey qualitative data is shallow. A customer who writes "needs better reporting" in an NPS follow-up isn't telling you which reports or for which use case. There's no way for them to prioritize that against other improvements. No conversation, no upvoting, no way for other customers to say "yes, me too."

3. No Prioritization Signal

Surveys tell you how many people are unhappy. They don't tell you which improvements would satisfy the most customers. Ten detractors can cite ten different reasons. Which one should you fix first?

Key takeaway: Surveys tell you the score. Feedback boards tell you what to build next. The two are complementary, not competing.

This is where continuous feedback fills the gap. A feedback board with voting gives you always-on signal. Instead of waiting for the next NPS cycle to learn that customers want better integrations, you can see that request accumulate votes in real time.

The combination works like this:

• NPS tells you the overall temperature is dropping
• Your feedback board tells you the top three requests driving dissatisfaction
• CSAT on support tells you whether your team is handling the frustration well
• CES on onboarding tells you whether new users are struggling before they even reach the features they care about

Across 6,035 product teams, ProductLift has collected over 157,624 feedback items and helped ship 39,406 features based on that continuous signal. That's the kind of volume and specificity no quarterly survey can match.

Try it yourself: Launch a feedback board alongside your survey metrics and see how much richer the signal becomes. No credit card required.

How Surveys and Feedback Boards Work Together

Rather than replacing surveys, continuous feedback makes them more powerful. Here's how they complement each other in practice.

NPS Drop? Your Feedback Board Shows Why

When NPS drops from +38 to +29, the follow-up responses give you fragments: "missing features," "too expensive," "competitors are better." Your feedback board gives you the full picture. The top three feature requests all relate to a workflow you changed last quarter, and they have 200+ combined votes. That's actionable.

CSAT Reveals the Moment; Feedback Reveals the Pattern

Post-support CSAT of 65% tells you customers are unhappy with support interactions. Your feedback board shows that 40% of support tickets stem from a confusing settings page. Fix the settings page and support CSAT improves without changing anything about the support team itself.

CES Points to Friction; Feedback Points to the Fix

An onboarding CES of 2.8 tells you setup is painful. Your feedback board shows that the top-voted request is "let me import data from a CSV." Now you know exactly which friction to remove.

Revenue Context Changes Everything

ProductLift's Stripe integration adds a layer none of these metrics can: revenue context. When you see that a feature request has 50 votes, you can also see that those 50 voters represent $45,000 in MRR. Compare that to another request with 80 votes but only $8,000 in MRR. The prioritization decision becomes much clearer.

With user segments, you can filter feedback by MRR range, plan type, and customer status (active, trial, churned). A detractor paying $5,000/month who wants a specific integration is a very different signal than a free trial user who wants the same thing.

Practical Scenarios: Choosing the Right Metric

Scenario 1: Post-Launch Feature Evaluation

You just shipped a redesigned dashboard. Which metric do you use?

Best choice: CSAT. Survey users who have interacted with the new dashboard: "How satisfied are you with the new dashboard?" This gives a direct measure of whether the redesign landed well.

NPS would fail here because it measures overall loyalty, not reaction to a specific change.

Scenario 2: Understanding Why Churn Is Increasing

Your churn rate jumped from 3% to 5% over two months.

Best choice: NPS + continuous feedback. NPS shows whether detractors are increasing (confirming the problem). Your feedback board and open-ended responses tell you why. Maybe a competitor launched a feature you lack. Maybe your pricing change frustrated mid-tier customers.

CES would fail here because churn driven by missing features or pricing has nothing to do with effort.

Scenario 3: Optimizing the Support Experience

Customers complain that getting help takes too long.

Best choice: CES + CSAT together. CES tells you whether the process is easy. CSAT tells you whether the outcome was satisfying. A customer can find the process easy (high CES) but the answer unhelpful (low CSAT), or vice versa.

Scenario 4: Deciding What to Build Next Quarter

You have limited engineering resources and six competing feature requests.

Best choice: None of these metrics. This is where surveys fall short entirely. You need a feature voting board where customers can upvote what matters most. Combine that with revenue data from your Stripe integration to weight votes by customer value. ProductLift's Journey Model connects this signal directly to your roadmap, then communicates what shipped through your changelog.

How to Act on Your Metrics

Collecting data without acting on it's worse than not collecting at all. It wastes customers' time and erodes trust. Here's how to close the loop on each metric.

Acting on CSAT

• Set thresholds: If post-support CSAT drops below 75%, trigger a review of recent tickets to identify patterns.
• Share with the team: Make CSAT visible to support in real time. A shared dashboard creates accountability.
• Correlate with other data: If CSAT is high but churn is also high, the problem isn't support quality. Look at product fit instead.

Acting on NPS

• Follow up with detractors within 48 hours. Personal outreach to understand their frustration.
• Analyze open-ended responses. Categorize by theme (pricing, features, reliability, support) and track which themes grow or shrink over time.
• Feed insights into your roadmap. If "better integrations" is the top theme among detractors, that should influence product priorities. ProductLift's status change notifications keep customers in the loop automatically. When you move a request from "Under Review" to "Planned" or "Shipped," every voter gets an email via the close the loop workflow.

Acting on CES

• Map the customer journey. Measure CES at every key touchpoint and identify the highest-effort steps.
• Redesign high-effort flows. If onboarding CES is 2.5, that's your most urgent UX project.
• Compare self-service vs. assisted. If your knowledge base CES is low, customers can't help themselves. That drives support volume and frustration.

Ready to close the gap between scores and action? Start a free trial and see how continuous feedback turns survey insights into shipped features. No credit card required.

FAQ

Can I use all three metrics at the same time?

Yes, and most mature SaaS companies do. The key is using each metric at the right touchpoint. Use NPS for overall loyalty (quarterly), CSAT for specific interactions (post-support, post-onboarding), and CES for process efficiency (post-support, post-self-service). Be careful about survey fatigue: no customer should receive more than one survey per week.

Which metric best predicts churn?

Gartner's research (originally from the Corporate Executive Board) found that CES is the strongest single predictor of customer loyalty and repeat purchase behavior. In their study, 96% of high-effort interactions led to disloyalty. However, this research focused primarily on service interactions. For SaaS, a combination of declining NPS plus high-effort CES scores is the strongest churn signal. Continuous feedback data adds early warning because customers often voice frustration on feedback boards long before it shows up in any survey metric.

How do I calculate a combined customer health score?

Many SaaS companies create a composite score that weights NPS, CSAT, CES, product usage data, and support ticket frequency. A simple starting approach: 40% NPS + 30% product usage + 20% CES + 10% CSAT. You will need at least 6 months of data to calibrate the weights for your specific context. The exact formula depends on which signals best predict churn and expansion in your business.

Should I use a 5-point or 7-point scale for CSAT?

A 5-point scale is standard and sufficient for most use cases. It's easier for respondents and produces comparable results. A 7-point scale offers slightly more granularity but typically doesn't change your strategic conclusions. The most important thing is consistency: once you pick a scale, don't change it, or you lose the ability to compare results over time.

What's a good response rate for these surveys?

For in-app surveys (CSAT, CES), aim for 30% to 50%. For email-based surveys (NPS), 20% to 30% is solid. SurveyMonkey's benchmark data suggests that in-app prompts triggered at contextually relevant moments consistently outperform batch emails by 2x to 3x. If your response rates fall below these thresholds, revisit timing and survey length before assuming customers don't care.

How do continuous feedback tools differ from these survey metrics?

Survey metrics (CSAT, NPS, CES) are push-based: you send a survey and ask for a response at a specific moment. Continuous feedback is pull-based: customers share thoughts whenever they want, on their own terms. Surveys give you structured, quantifiable data. Feature voting boards give you unstructured, qualitative depth plus a built-in prioritization mechanism through votes. The two approaches complement each other. Surveys tell you the score; continuous feedback tells you the story behind it and what to build next.