All free tools

Customer Satisfaction (CSAT) Score Calculator

Free CSAT calculator. Enter counts per score on a 1–3, 1–5, 1–7 or 1–10 scale to get CSAT %, DSAT %, average score and margin of error. No signup.

CSAT % is the share of survey responses in the top boxes of your scale (4–5 on a 1–5 scale), divided by all responses, times 100. Enter how many customers chose each score; the calculator returns CSAT %, DSAT %, average score and a 95% margin of error, and states the top-box rule it used.

Inputs

Survey scale

Count as satisfied

Choices that would reach the neutral midpoint are not offered. Dissatisfied (DSAT) = 1–2.

Every answered score counts, neutral and negative included. Blank or skipped answers are left out of both the top and the bottom of the fraction. Everything runs in your browser and nothing is uploaded.

Results

Updates live

CSAT score

77.0%

± 5.8 points

CSAT = share of responses scoring 4–5 on a 1–5 scale (top 2 boxes) · 95% range 71.2% to 82.8%

Response split

DSAT (1–2)
13.0%(26)
Neutral / other
10.0%(20)
CSAT (4–5)
77.0%(154)
Average score
4.05 / 5
Responses (n)
200
Responses needed for ±5 points
273

CSAT % from different scales or different top-box rules is not comparable.

How this is calculated

Pick your survey scale and how many top boxes count as satisfied (4–5 on a 1–5 scale by default), then enter how many customers chose each score or paste the raw scores. CSAT % is the top-box responses divided by all answered responses, times 100; DSAT % is the bottom-box share; the average is the sum of score × count divided by responses. The 95% margin of error is 1.96 × √(p(1 − p) ÷ n) × 100 points. Everything runs in your browser and nothing is uploaded.

How CSAT is calculated

CSAT % = (responses in the top box(es) ÷ all valid responses) × 100. Valid responses are every answer with a score on the scale, neutral and negative ones included. Skipped questions are left out of both numbers. DSAT % is the same calculation for the bottom boxes, and neutral is whatever is left, so CSAT and DSAT do not add up to 100%.

ScaleSatisfied (default)DSATWhy
1–33 (top 1)1Top 2 would include the neutral midpoint
1–54–5 (top 2)1–2"Satisfied" plus "Very satisfied", the standard CSAT convention
1–76–7 (top 2)1–2Same top-2 rule; Top 3 (5–7) available
1–109–10 (top 2)1–2Same top-2 rule; many teams use 8–10, so Top 3 is available

There is no 0–10 option: that is the Net Promoter scale, so use the NPS calculator for it. A 1–10 result is not converted to a "1–5 equivalent". Sources for the formula and the top-box terms: SurveyMonkey CSAT calculator; Jeff Sauro & Jim Lewis, "Top Box, Top-Two Box, Bottom Box, or Net Box?", MeasuringU, June 4, 2024.

Worked example: post-ticket survey, 1–5 scale, 200 responses

  • Counts: 5 → 96, 4 → 58, 3 → 20, 2 → 12, 1 → 14 (total 200)
  • CSAT % = (96 + 58) ÷ 200 = 154 ÷ 200 = 77.0%
  • DSAT % = (12 + 14) ÷ 200 = 13.0%; neutral = 20 ÷ 200 = 10.0%
  • Average = (96×5 + 58×4 + 20×3 + 12×2 + 14×1) ÷ 200 = 810 ÷ 200 = 4.05 / 5
  • Margin of error = 1.96 × √(0.77 × 0.23 ÷ 200) × 100 = ±5.8 points → 95% range 71.2% to 82.8%
  • The same 77% from only 40 responses: ±13.0 points
  • Responses needed for ±5 points at 77%: 1.96² × 0.77 × 0.23 ÷ 0.05² = 272.1 → 273

Margin of error uses the normal approximation, 1.96 × √(p(1 − p) ÷ n) × 100 points. Below 30 responses, or when every answer is on one side of the rule, the calculator shows no range and labels the score directional.

CSAT % vs average score

CSAT % counts how many customers were satisfied. The average (Σ score × count ÷ responses, shown as "4.05 / 5") moves when answers shift anywhere on the scale, including from 1 to 2 or from 4 to 5, which a top-box share ignores. Report both; never show the average as a percentage.

For AI support, compute CSAT separately for AI-resolved, AI-then-human and human-only tickets. A blended CSAT can rise while handoffs get worse. The Customer Effort Score guide and the customer service survey questions cover the tags to add so each path can be reported on its own.

CSAT benchmarks

The most-cited public satisfaction benchmark is not a CSAT percentage. Read it as context, not a target.

SourceScale and methodFiguresAs of
American Customer Satisfaction Index (ACSI), U.S. Overall Customer Satisfaction0–100 index ("100-point scale"); more than 400 companies in over 40 industries, about 200,000 interviews a yearNational ACSI 76.1 (Q2 2026); 76.7 (Q1 2026); 76.9 (Q2 2025)Latest published quarter as of Nov 17, 2026: Q2 2026

ACSI scores are on a 0–100 index scale. They are not the percentage of customers who chose the top two boxes, so a 76 ACSI score and a 76% CSAT do not mean the same thing. Do not use ACSI as a CSAT % target.

There is no public, method-matched top-2-box CSAT benchmark for customer support. Track your own trend at the same scale, rule and survey trigger, and check the margin of error before calling a change real.

Related RTCI guides

Frequently asked

How do you calculate CSAT percentage?

Count the responses in the top boxes of your scale (4 and 5 on a 1–5 scale), divide by all valid responses including neutral and negative ones, and multiply by 100. 154 satisfied out of 200 responses is 77%.

How do I calculate CSAT on a 1–10 scale?

Pick a top-box rule and keep it. This calculator defaults to 9–10 (top 2); some teams use 8–10. On the same 100 answers the two rules gave 55% and 75%, so always state the rule next to the score and never compare scores built with different rules.

What is DSAT?

DSAT is the share of dissatisfied responses, usually the bottom two boxes (1–2 on a 1–5 scale). CSAT and DSAT do not add up to 100% because neutral answers sit in neither.

Should I report CSAT % or the average score?

Both. CSAT % shows how many customers were satisfied; the average shows shifts across the whole scale that a top-box share can miss. MeasuringU recommends reporting both because they answer different questions.

What is a good CSAT score?

There is no public, method-matched benchmark for top-2-box CSAT in customer support. Compare against your own trend at the same scale, rule and survey trigger, and check the margin of error before calling a change real. The ACSI national score (76.1 in Q2 2026) is on a 0–100 index scale built differently, so it is not a CSAT % target.