Measure satisfaction. Fix what's broken.

The Complete Guide to Customer Satisfaction Score (CSAT)

What it is, how to calculate it, what a good score looks like, and when to use it.

CSAT is the most direct way to measure satisfaction after a specific customer interaction. Here's everything you need to know to run it correctly.

Definition

What is Customer Satisfaction Score?

CSAT is a transactional survey metric that measures how satisfied a customer was with a specific experience - a support interaction, an onboarding session, a purchase, or a feature release.

Unlike NPS, which tracks overall loyalty over time, CSAT is immediate and contextual. It asks one question: “How satisfied were you with [this specific experience]?”

Responses are collected on a 1–5 scale (Very Dissatisfied → Very Satisfied). The score is the percentage of respondents who rated 4 or 5 - not an average of all scores.

Type

Transactional - tied to a specific interaction

?

Survey question

"How satisfied were you with [experience]?"

Scale

1–5 (Very Dissatisfied to Very Satisfied)

%

Score output

% of respondents who rated 4 or 5

When to send

Immediately after the interaction

Good benchmark

80%+ is good, 90%+ is excellent

The formula

How to calculate CSAT

CSAT Formula

CSAT = (Satisfied responses ÷ Total responses) × 100

Satisfied = rated 4 or 5 on a 1–5 scale. Only top-box responses count.

Example calculation

Surveys sent
200
Responses received
120
Rated 4 or 5 (satisfied)
96
Rated 1, 2, or 3 (not satisfied)
24
CSAT score
96 ÷ 120 × 100 = 80%

CSAT score benchmarks

90–100%

Excellent

World-class customer experience. Very few companies sustain this - investigate and systematize what's working.

80–89%

Good

Above industry average for most SaaS categories. Focus on converting neutrals (3s) rather than fixing the basics.

70–79%

Average

Industry median. Room for improvement - identify which touchpoints are pulling the score down.

60–69%

Below Average

Meaningful dissatisfaction. Customers are noticing friction. Review your lowest-scoring touchpoints immediately.

Below 60%

Critical

Significant customer experience failure. This level of dissatisfaction typically precedes churn spikes.

Timing and context

When to send a CSAT survey

CSAT is most accurate when sent immediately after the specific experience. Wait longer than 48 hours and recall degrades significantly.

Support ticket resolved

Immediately on resolution

Support CSAT is the most common use case. Captures the most actionable signal - fast follow-up means customers remember the interaction.

Measures: Agent quality, resolution speed, first-contact resolution rate

Onboarding completed

After the activation milestone

Low onboarding CSAT is the strongest predictor of early churn. Identify friction before the 30-day mark when intervention is still possible.

Measures: Setup complexity, time to first value, documentation quality

Purchase or upgrade

Within 24 hours of transaction

Reveals whether the sales process set correct expectations. Misalignment between sales promise and product reality shows up here first.

Measures: Expectation gap, pricing clarity, purchase friction

Feature release

On first use of the new feature

In-product CSAT triggered by first interaction captures immediate reaction before habituation sets in. Identifies polish gaps pre-launch.

Measures: Feature usability, discoverability, value delivery vs. expectation

Renewal or contract review

30 days before renewal date

Renewal-period CSAT gives account managers a data point for the renewal conversation and early warning on accounts at risk.

Measures: Overall value perception, ROI clarity, relationship health

Professional services / implementation

At each project milestone

Long-duration projects need milestone check-ins, not just end-of-project CSAT. Issues surfaced mid-project are recoverable; issues surfaced at the end are not.

Measures: Project management quality, communication, deliverable quality

What to ask

CSAT survey questions

One primary question. One follow-up. Keep it short - CSAT response rates drop sharply after the second question.

Primary question

“How satisfied were you with [the support you received today]?”

Replace the bracketed text with the specific interaction. Use a 1–5 scale labeled Very Dissatisfied → Very Satisfied. Score = % who chose 4 or 5.

Follow-up questions by score

All respondents

After any score

"What is the main reason for your score?"

Open text. The single most useful follow-up - surfaces the actual friction or delight driver without guessing.

Low scores (1–3)

Dissatisfied

"What could we have done better?"

Forward-facing phrasing gets more actionable responses than "What went wrong?" - respondents describe the fix, not just the complaint.

High scores (4–5)

Satisfied

"What did we do particularly well?"

Identifies what to systematize. Promoter patterns - fast resolution, specific agent, clear communication - are the playbook for everyone else.

Improvement tactics

How to improve your CSAT score

CSAT improvement comes from fixing specific touchpoints, not from optimizing the metric itself.

01

Close the loop on low scores within 24 hours

Respond personally to every 1–2 score. A direct follow-up from the support team after a bad experience recovers more relationships than any product fix. Speed matters - the window closes fast.

02

Find which touchpoint is dragging the average down

Track CSAT per touchpoint separately. A 75% overall score could mean onboarding is at 60% while support is at 90%. Fix the lowest touchpoint first - aggregate scores hide where the real problems are.

03

Use CES to find friction in low-score patterns

When CSAT dips after support or onboarding, run a CES survey on the same interaction. High effort predicts dissatisfaction better than almost anything else. Reduce steps, reduce wait time, reduce cognitive load.

04

Convert 3s - neutrals have the most upside

Customers who rate 3 (neutral) are one good experience away from becoming 4s. They are not actively dissatisfied but they are not loyal. Identify what neutrals have in common - use case, tier, onboarding path - and target that group with a better experience.

CSAT vs NPS

When to use CSAT vs NPS

CSAT and NPS measure different things. CSAT is transactional - it tells you whether a specific interaction went well. NPS is relational - it measures overall loyalty and whether customers would stake their reputation on recommending you.

Run CSAT after every meaningful touchpoint (support, onboarding, purchase). Run NPS quarterly to track the relationship over time. Together they give you both the diagnostic (where is friction happening?) and the trend (is the relationship getting stronger?).

Compare CSAT and NPS in detail

Implementation

How to run a CSAT survey

Five steps from setup to action.

01

Define the specific touchpoint

CSAT works because it is tied to a specific moment. Choose one: support resolution, onboarding completion, feature first-use, or post-purchase. Do not send a generic "how are we doing?" survey - the context is what makes responses useful.

02

Set the trigger timing

For support: send within minutes of ticket resolution. For onboarding: send when the user hits the activation milestone (first value moment), not at an arbitrary day 7 or day 14. For purchases: within 24 hours. For features: on first interaction with the new capability.

03

Write the question correctly

Keep it singular and specific: "How satisfied were you with [the support you received today]?" Not "How do you feel about our product overall?" Add one optional open text field: "What could we have done better?" Open text on low scores (1–3) is your most valuable qualitative signal.

04

Collect and segment results

Track CSAT per touchpoint separately - support CSAT and onboarding CSAT are different metrics. Segment by customer tier, use case, or cohort to find patterns. A 78% overall CSAT is meaningless if onboarding is at 60% and support is at 92%.

05

Close the loop on low scores

Follow up on all 1–2 scores within 24 hours. A personal response to a dissatisfied customer after a bad experience can recover the relationship. Systematic analysis of 3 scores identifies the improvements that move your overall score.

Industry benchmarks

What is a good CSAT score by industry?

78–82%

SaaS / Technology

Wide variance by segment

75–80%

E-commerce / Retail

Post-purchase focus

76–80%

Financial Services

Driven by support quality

77–83%

Healthcare

High stakes interactions

Benchmarks vary by survey channel, customer segment, and methodology. Use them as orientation, not targets - what matters more is your trend over time.

Go deeper

Every angle of Customer Satisfaction Score

Each guide covers one part of CSAT measurement in full detail.

Survey tool

CSAT survey tool

Sending CSAT manually - a form link in every support reply, a spreadsheet to calculate percentages - works for a handful of tickets but breaks at scale. A dedicated CSAT survey tool triggers automatically after each interaction, links every satisfaction rating to the specific user and their attributes, and surfaces which agents, features, or plan tiers are dragging the score without you having to dig through raw exports.

See the CSAT survey tool →

Free tool

CSAT calculator

CSAT is the percentage of 4s and 5s out of total responses - straightforward in theory but time-consuming to calculate correctly across multiple batches or survey periods. The free CSAT calculator takes your raw response counts and returns your score, satisfaction breakdown, and benchmark comparison instantly. No spreadsheet, no formula errors, no signup required.

Open the free CSAT calculator →

Benchmarks

What is a good CSAT score?

Above 80% is generally considered good for SaaS and above 90% is excellent - but those thresholds only tell you whether you're above or below the industry average. The more useful comparison is within your interaction type: support CSAT benchmarks higher than product CSAT, and post-onboarding CSAT benchmarks lower than post-resolution CSAT. Mixing interaction types into one score hides where the real problem is.

CSAT benchmarks by industry and stage →

Ready to use

Free CSAT survey template

The standard CSAT survey is one question - “How satisfied were you with [interaction]?” on a 1–5 scale - plus an optional open-text follow-up. The template is pre-built with the standard question, response scale, and conditional follow-ups for low and high scores, ready to send immediately after support interactions, onboarding, or any product touchpoint without any configuration.

Get the free CSAT survey template →

Survey questions

25 CSAT survey questions

The core CSAT question covers overall satisfaction, but follow-up questions reveal which specific aspect of the interaction - speed, clarity, outcome, or agent quality - is driving the score down. These 25 questions give you the core question plus targeted follow-ups for support, onboarding, feature, and product update interactions, so you can pinpoint exactly what to fix.

See all 25 CSAT survey questions →

Playbook

How to improve customer satisfaction

A low CSAT score is a symptom. The levers that actually move it are following up with dissatisfied customers within 24 hours, fixing the specific interaction type dragging the score, reducing resolution time, and closing the loop so customers know their feedback was acted on. The playbook covers each lever with specific tactics and before/after benchmarks.

How to improve CSAT: full playbook →

Metric comparison

NPS vs CSAT: different time horizons

CSAT measures satisfaction with a single touchpoint right after it happens. NPS measures overall relationship loyalty - the cumulative effect of every interaction across months. They're complementary: use CSAT to measure and diagnose individual interactions, use NPS quarterly to track whether those interactions are adding up to loyalty or eroding it. A high NPS with low support CSAT is a warning sign.

NPS vs CSAT: full comparison →

Metric comparison

CSAT vs CES: satisfaction vs effort

A customer can rate a support interaction 5/5 on CSAT - satisfied with the outcome - while rating the effort 2/7 on CES - it took three contacts to get there. CSAT captures how happy they were with the result; CES captures how hard they had to work. Research shows CES is a stronger churn predictor at the interaction level, while CSAT is better for measuring quality of outcomes.

CSAT vs CES: full comparison →

Frequently asked questions

Everything you need to know about CSAT.

Get started today

Start measuring customer satisfaction today

Run CSAT surveys after support, onboarding, and key product moments. Find where satisfaction drops before it shows up in churn.

Run Your Free CSAT Survey

No credit card required