Customer Satisfaction Survey: Everything You Need to Know in 2026

A customer satisfaction survey is how teams turn a recent experience into a clear signal: Was this good enough? What broke? What should we fix next?

In 2026 that signal matters more than vanity dashboards. The American Customer Satisfaction Index reported a national score of 76.7 out of 100 in Q1 2026, down slightly year over year and still stuck near levels last seen years earlier. Collecting a score is easy. Building a survey program that changes products, support, and retention is the harder job.

This guide covers the full loop: what a customer satisfaction survey is, how CSAT is calculated, when to use it (and when not to), question templates, channels, timing, interpretation, and how to close the loop. Pair it with 100 customer service survey questions when you need a larger question bank.

A useful CSAT program asks a clear question, captures context, and drives action.

What Is a Customer Satisfaction Survey?

A customer satisfaction survey is a short questionnaire that asks people how satisfied they were with a product, service, interaction, or overall relationship.

Most programs center on one rating question (the CSAT question), then optionally add:

• One “why” follow-up (open text or multiple choice)
• One attribute question (speed, clarity, resolution, value)
• Optional contact permission for follow-up

That is different from a long market-research questionnaire. Satisfaction surveys succeed when they are specific, recent, and actionable. If the answer cannot change a queue, a page, a policy, or a coaching note, cut the question.

Survey vs score vs index

Teams often mix three different ideas:

Term	What it is	How you use it
Customer satisfaction survey	The questionnaire you send	Collect feedback at a chosen moment
CSAT score	% of positive ratings on your scale	Track a touchpoint or team over time
ACSI-style index	A modeled 0–100 economic indicator	Broad economy / industry context only

Your internal CSAT percentage and an ACSI industry score are related themes, not identical math. Compare like with like: same scale, same definition of “satisfied,” same channel mix.

What Is CSAT?

CSAT (Customer Satisfaction Score) is the percentage of respondents who gave a positive rating on your satisfaction question.

Typical question: “How satisfied were you with [interaction / product / service]?”

Common scales:

Scale	Labels (example)	Usually counts as satisfied
1–5	Very dissatisfied → Very satisfied	4 and 5 (top-2 box)
1–7	Strongly dissatisfied → Strongly satisfied	5–7 (top-3 box)
Stars	1–5 stars	4–5 stars
Smileys	3 or 5 faces	Positive faces only

Five-point scales are the default for most support and post-purchase programs because they are fast on mobile. Seven-point scales add nuance when you need to detect smaller shifts. Pick one scale and keep it stable so trends stay comparable.

CSAT is a percentage of positive ratings, not the average star value alone.

The CSAT formula

CSAT = (satisfied responses ÷ total responses) × 100

Example: 160 of 200 people rate 4 or 5 on a 1–5 scale → 80% CSAT.

You can also track average rating as a secondary metric, but the headline CSAT percentage is what most CX teams report. If you already have rating tallies, plug them into the CSAT calculator to get the percentage, average rating, and an optional industry reference check.

CSAT vs NPS vs CES

These three metrics answer different questions. Mixing them without a plan creates noise.

CSAT = moment. NPS = relationship. CES = friction.

Metric	Core question	Scale	Best cadence
CSAT	How satisfied were you with this?	Usually 1–5 or 1–7	After a specific interaction
NPS	How likely are you to recommend us?	0–10	Quarterly / lifecycle milestones
CES	How easy was it to get this done?	Often 1–7 agree/disagree	After resolution or self-service

NPS (Net Promoter Score) is defined by the Net Promoter System as the share of promoters minus detractors on a 0–10 recommend question. See how Bain describes measuring NPS for the standard promoter / passive / detractor split. Calculate yours with the NPS calculator.

CES grew from research popularized in Harvard Business Review’s “Stop Trying to Delight Your Customers”, which argued that reducing effort often predicts loyalty better than chasing delight in service interactions.

Key takeaway: Use CSAT when you need a clean read on a moment. Use CES when friction is the hypothesis. Use NPS when you need a relationship loyalty signal.

Types of Customer Satisfaction Surveys

Transactional CSAT

Sent right after a defined event: ticket closed, order delivered, onboarding step finished, appointment completed.

Strength: High recall, clear ownership (this queue, this flow, this store).
Risk: Over-surveying people who contact you often.

Relationship CSAT

Asks about overall satisfaction with the brand, product, or account, usually on a monthly or quarterly cadence.

Strength: Broader pulse across the journey.
Risk: Vague answers if you do not add theme questions or segment by tenure and plan.

Match the survey type to the decision you need to make.

Common journey moments

Moment	What to ask about	Watch for
Post-support	Resolution quality, agent clarity, wait	Sampling only ticket creators
Post-purchase	Product fit, delivery, returns ease	Shipping delays dominating the score
Onboarding	Clarity, time-to-value, setup help	Early enthusiasm bias
In-product feature	Usefulness after first real use	Interrupting mid-task
Event / in-person	Venue, staff, wait, overall visit	Incomplete scans if the form is long

For in-person moments, a phone-first form behind a QR code often beats a paper card. See how to collect customer feedback with a QR code.

When to Use CSAT (and When Not To)

If you cannot name the decision the score will change, do not send the survey yet.

Use CSAT when

• You need a fast read on a specific touchpoint
• Owners can act within days (support, ops, product)
• The experience is still fresh in memory
• You will pair the score with a “why” and a follow-up path for low scores

Prefer another approach when

• You care more about loyalty and advocacy than a single interaction → NPS
• You suspect friction, not happiness, is the problem → CES
• You need deep discovery (jobs to be done, pricing research) → interviews or longer research studies
• Volume is too low for a stable percentage → wait, widen the window, or use qualitative outreach
• You already survey the same people weekly → consolidate; survey fatigue destroys signal

How to Create a Customer Satisfaction Survey

1. Write one primary question

Keep the object of the question precise:

• Weak: “How satisfied are you with us?”
• Stronger: “How satisfied were you with the support you received on this request?”
• Stronger: “How satisfied were you with your checkout experience today?”

2. Choose a scale and stick to it

Document:

• Scale length (5 vs 7)
• Which boxes count as satisfied
• Whether neutrals sit in the denominator (they should)

3. Add one diagnostic follow-up

Examples:

• “What is the main reason for your score?”
• “What could we improve?”
• Multiple choice themes: wait time, clarity, resolution, product quality, pricing, other

Open text after a low score is often more valuable than three extra rating questions.

4. Keep it short

Aim for under a minute on a phone. A reliable anatomy:

1. CSAT rating
2. One reason / theme
3. Optional comment
4. Optional “may we follow up?”

Fewer questions usually mean more completes and cleaner trends.

5. Design for mobile first

Large tap targets, no tiny radio grids, minimal typing, one question per screen if your tool supports it. If you build from a prompt, turn the prompt into a shareable form and test on a real phone before you launch.

Customer Satisfaction Survey Questions and Templates

Use these as starting points. Adapt nouns to your product and moment.

One precise rating plus one reason beats a long laundry list.

Core CSAT stems

1. How satisfied were you with your overall experience today?
2. How satisfied were you with the support you received on this request?
3. How satisfied were you with [product / order / appointment]?
4. How well did we meet your expectations for this interaction?
5. Overall, how satisfied are you with [company / product] right now?

Support / service follow-ups

6. Was your issue fully resolved?
7. How clear was the guidance you received?
8. How reasonable was the time it took to get help?
9. What is the main reason for your score?
10. What is one thing we should change next?

Product / purchase follow-ups

11. How satisfied were you with product quality?
12. How easy was checkout or setup?
13. How satisfied were you with delivery or access timing?
14. Did the product do what you expected?
15. What almost stopped you from buying or finishing setup?

Relationship pulse extras (use sparingly)

16. How satisfied are you with the value you receive for the price?
17. How well does [product] fit your needs today?
18. How easy is it to get help when you need it?
19. How likely are you to continue using us over the next 6 months?
20. What should we prioritize improving this quarter?

For a much larger bank organized by metric, use the customer service survey questions guide.

Mini templates

Post-support (3 questions)

1. CSAT on support received
2. Issue resolved? (Yes / Partially / No)
3. What could we improve?

Post-purchase (4 questions)

1. CSAT on purchase experience
2. Product met expectations?
3. Main friction theme (multiple choice)
4. Optional comment

Quarterly relationship (5 questions max)

1. Overall CSAT with brand/product
2. Value for price
3. Ease of getting help
4. Top improvement theme
5. Optional open comment

Channels and Timing

Send while memory is fresh, on the channel where the experience happened when you can.

Channel	Best for	Watch-outs
In-app / web	Feature and workflow moments	Do not block a critical task
Email	Receipts, resolutions, relationship pulses	Lower urgency; subject line matters
SMS / messaging	Time-sensitive transactional asks	Keep extremely short; respect consent
Chat / ticket widget	Immediately after chat close	Avoid surveying unresolved threads
QR / kiosk / table tent	Stores, events, clinics, venues	Form length decides completion after scan
Phone	High-value accounts	Interviewer effects; use sparingly

Timing rules that hold up in practice:

• Transactional: send as soon as the event is complete (minutes to a few hours, not days later)
• Delivery / visit: same day when possible
• Relationship: fixed cadence (monthly or quarterly), not after every click
• Reminders: one reminder max for most programs

Track response rate alongside the score. A rising CSAT on a collapsing response rate is not a win. Use the survey response rate calculator when you want a clean rate, and the confidence interval calculator when sample sizes are small.

What Is a Good CSAT Score?

There is no universal “good” number. Context beats slogans.

As a rough cross-industry guide many teams still use:

CSAT %	Common reading
85%+	Strong / excellent for many service contexts
75–84%	Solid; dig into themes and segments
65–74%	Mixed; prioritize friction fixes
Below 65%	Urgent investigation

Industry and channel matter. Banking and SaaS programs often run higher than telecom or utilities in published tables. Your own baseline, by channel and segment, is the real target. National indexes like ACSI are useful for macro context; they are not a substitute for your top-box CSAT definition.

Read the trend and the comments, not a single week’s average in isolation.

How to read a score responsibly

1. Trend over time on a stable definition
2. Segment by channel, team, product, tenure, and issue type
3. Coverage (who responded vs who was invited)
4. Comments and themes behind lows and highs
5. Operational couples (first response time, reopen rate, refund rate)

A 78% that is climbing with healthier response coverage can beat an 88% built from ten fans.

From Score to Action: Closing the Loop

Unhappy respondents are a recovery opportunity, not just a red cell in a spreadsheet.

A survey without action teaches customers that feedback disappears.

Minimum closed-loop practice:

1. Route low scores (for example 1–2 on a 5-point scale) to a human owner within one business day
2. Tag themes weekly (wait, clarity, bugs, billing, delivery)
3. Fix the highest-volume root causes first
4. Tell customers when a fix ships when you can
5. Re-measure the same touchpoint after the change

High scores deserve a lighter loop: thank-yous, testimonials (with permission), and notes on what to protect.

Best practice: Separate “alert follow-up” (individual recovery) from “system fix” (process or product change). Doing only alerts burns agents out. Doing only dashboards leaves customers hanging.

How to Improve Customer Satisfaction Survey Results

Improving the score and improving the survey are different jobs.

Improve the experience the survey measures

• Cut first response and resolution time where they drive lows
• Raise first-contact resolution; repeats destroy CSAT
• Clarify policies before customers hit support
• Fix the top three comment themes before launching a redesign
• Make self-service actually resolve the top intents

Improve the survey itself

• Shorten it
• Ask about one moment only
• Send sooner
• Survey on the same channel as the experience when possible
• Stop over-sampling power users
• Publish an internal “definition of satisfied” so teams argue less about math

Most lasting CSAT gains come from removing friction, not from rewriting the thank-you page.

Common Mistakes

Most “CSAT is useless” complaints are process failures, not metric failures.

1. Surveying too late so memory is fuzzy
2. Asking five ratings and no “why”
3. Changing the scale every quarter and still charting one line
4. Ignoring nonresponse and celebrating a biased sample
5. Treating CSAT as NPS (or the reverse)
6. No owner for low scores
7. Punishing agents for structural product failures
8. Leading questions (“How amazing was our world-class support today?”)
9. Double-barreled items (“How satisfied were you with speed and accuracy?”)
10. Launching on desktop-only layouts for phone respondents

A Practical 2026 CSAT Program Checklist

1. One primary CSAT question with a documented scale and top-box rule
2. One diagnostic follow-up on every transactional survey
3. Clear trigger events (and a do-not-survey list)
4. Mobile-tested form and invitation copy
5. Response-rate tracking beside the score
6. Low-score routing with an SLA
7. Weekly theme review with product and support owners
8. Quarterly relationship pulse (if needed), not a second transactional blast
9. Stable definitions for at least two consecutive quarters before big target changes

Frequently Asked Questions

What is a customer satisfaction survey?

It is a short questionnaire that asks how satisfied someone was with a product, service, or interaction. Most versions lead with a CSAT rating and add a brief follow-up so you know what to fix.

What does CSAT stand for?

CSAT stands for Customer Satisfaction Score. It is usually reported as the percentage of respondents who chose a positive rating on your scale.

How do you calculate CSAT?

Divide satisfied responses by total responses, then multiply by 100. On a 1–5 scale, satisfied usually means ratings of 4 and 5.

What is a good CSAT score in 2026?

Many teams treat roughly 75–85% as solid and 85%+ as strong, but industry, channel, and your own baseline matter more. Macro indexes such as ACSI provide economy-wide context; they are not a direct substitute for your top-box percentage.

How many questions should a CSAT survey include?

Prefer one to four questions for transactional surveys. If people need more than about a minute after a casual invite or QR scan, completion usually falls.

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

Use 5-point for most day-to-day CSAT. Use 7-point when you need finer sensitivity and can keep the definition of “satisfied” consistent over time.

CSAT vs NPS: which should I use?

Use CSAT for a specific interaction. Use NPS for overall willingness to recommend. Many mature programs run both on different cadences rather than forcing one metric to do both jobs.

When should I send a customer satisfaction survey?

As soon as the experience is complete while memory is fresh: after ticket resolution, delivery, checkout, onboarding milestones, or an in-person visit.

How do I improve low CSAT?

Follow up with unhappy respondents quickly, tag recurring themes, fix the highest-volume causes, and re-measure the same touchpoint. Coaching helps when the issue is agent skill; process or product work is required when themes are structural.

Are customer satisfaction surveys still effective?

Yes, when they are short, well timed, and tied to owners who act. They are weak when they are long, late, disconnected from operations, or treated as a PR scoreboard.

How many responses do I need?

Directional reads can start with a few dozen replies. For steadier tracking, many teams aim for 100+ completes per wave, more when comparing segments. Pair small samples with a confidence interval.

Can I run CSAT offline or in person?

Yes. Use tablets, kiosks, or a QR code that opens a mobile form. Keep the form short enough that a twelve-second scan is not followed by a five-minute questionnaire.

Conclusion

A customer satisfaction survey works when three pieces stay aligned: a precise question about a real moment, a stable CSAT definition you can trend, and a closed loop that turns low scores into fixes.

In 2026 the bar is not another dashboard tile. It is a program that respects customer time, resists survey fatigue, and connects comments to owners. Start with one transactional survey, one follow-up question, and a weekly theme review. Expand only after that loop is reliable.

When you are ready to ship, keep the form phone-first and shareable. Formms can turn a short prompt into a live survey with a link and QR if you want that path; the questions and follow-up discipline still matter more than the generator. For the score math itself, use the CSAT calculator, and for question ideas beyond this guide, see 100 customer service survey questions.