40 Customer Satisfaction Survey Questions (CSAT) for B2B Teams

For a B2B SaaS team, good customer satisfaction survey questions ask about one customer moment at a time. Use one overall 1 to 5 rating, one question for each decision you will act on, and one open question for the why.

The 40 customer survey questions are grouped as overall satisfaction, onboarding, support, product use, renewal, cancellation and open-ended questions.

A CSAT score measures satisfaction with one moment, and in a B2B account the person who answers decides what the score means. A buyer, a daily user and an admin can rate the same product very differently.

You get each question with its scale and what it tells you, plus an 8-question survey ready to send. Then comes how to score it per account, and what to do when a question scores low.

What Is a Customer Satisfaction Survey?

A customer satisfaction survey measures how happy customers are with a company’s products, services and capabilities, which is how ASQ’s quality glossary defines customer satisfaction.

It captures satisfaction with one moment, product or service at the time you ask. That makes it different from a relationship metric such as NPS, which the next section compares.

The core question is question 1 below, “How satisfied are you with [product] overall?”, answered from 1 (very dissatisfied) to 5 (very satisfied). In Survicate’s 2025 CSAT benchmark data, the 5-point scale appears in 80.6% of the 1,259 surveys analysed (vendor data).

For a B2B team the value is timing: the survey shows which moment to fix before an account reaches renewal. The CSAT survey questions below each target one of those moments.

CSAT, NPS, and CES: Which One to Ask

Ask CSAT about one moment, NPS about the whole relationship and CES about friction in one task. Each metric answers a different question, so each belongs at a different point in an account’s year.

Metric	The question	Scale	When to send	What it tells you
CSAT	How satisfied are you with [product] overall?	1 to 5, very dissatisfied to very satisfied	Right after a moment you want to judge	Satisfaction with that moment
NPS	How likely is it that you would recommend [product] to a friend or colleague?	0 to 10; promoters 9 to 10, passives 7 to 8, detractors 0 to 6	On a relationship cadence, not after a ticket	Loyalty to the whole account relationship
CES	[Company] made it easy for you to handle your issue.	1 to 7, strongly disagree to strongly agree	Right after a task or support request	How hard it was to get something done

The three share a survey format but not a purpose: only NPS, whose standard question comes from Bain & Company, spans the whole relationship. CES was developed by CEB, now part of Gartner, and asks how easy one interaction was.

A high NPS can sit beside a bad support week, so never blend the answers to these customer experience survey questions into one score.

Questions to Ask in a Customer Satisfaction Survey

These customer satisfaction survey questions are grouped by customer moment.

Replace [product], [feature], [first key task] and [Company] with your own names. The 40 survey questions for customer satisfaction below are numbered so you can pick by number.

We wrote each question for one decision it feeds. We then checked the set against the questions the ranking guides share and removed duplicates.

Overall Satisfaction Baseline

Use when: you want a baseline from every contact on an account. Ask them in every survey so the trend holds.

#	Question	Scale	What it tells you
1	How satisfied are you with [product] overall?	1 to 5, very dissatisfied to very satisfied	Your baseline CSAT, tracked over time
2	How well does [product] meet your team’s needs?	1 to 5, not at all well to extremely well	Fit, apart from one bad week
3	How satisfied are you with the value you get for what you pay?	1 to 5, very dissatisfied to very satisfied	A price-value gap before it shows up at renewal
4	Compared with three months ago, is your experience better, the same or worse?	Better / the same / worse	The direction of change a single score hides
5	How satisfied are you with how our team communicates with you?	1 to 5, very dissatisfied to very satisfied	Account-team and update quality, apart from the product

Onboarding and First Value

Use when: an account has just finished setup or reached its first milestone, and the admin or champion who ran it answers.

#	Question	Scale	What it tells you
6	How satisfied are you with onboarding so far?	1 to 5, very dissatisfied to very satisfied	Whether setup is on track
7	How easy was it to set up [product] for your team?	1 to 5, very difficult to very easy	Setup effort
8	Has your team completed [first key task] yet?	Yes / no	A fact: whether first value happened
9	How clear were the onboarding guides and training?	1 to 5, very unclear to very clear	Whether materials or people are the gap
10	How confident are you that your team can use [product] without our help?	1 to 5, not at all to completely confident	Adoption risk after handoff
11	What almost stopped you from getting started?	Open text	The blocker to remove for the next customer

In our reading, Q8 comes first: an account that never reached first value rates setup differently from one that did.

Support and Service

Use when: a ticket has just closed and the person who raised it gets these customer service survey questions.

#	Question	Scale	What it tells you
12	How satisfied are you with the support you received?	1 to 5, very dissatisfied to very satisfied	Post-ticket CSAT
13	Was your issue fully resolved?	Yes / partly / no	Resolution, the first thing to check on a low score
14	[Company] made it easy for you to handle your issue.	1 to 7, strongly disagree to strongly agree	Customer Effort Score, CEB’s statement in second person
15	How satisfied are you with how quickly we responded?	1 to 5, very dissatisfied to very satisfied	Speed, apart from the outcome
16	How well did our support agent understand your problem?	1 to 5, not at all well to extremely well	Agent skill against process
17	How many times did you contact us about this issue?	Once / twice / three times or more	Repeat contacts, a sign the first fix failed
18	What could we have done better on this request?	Open text	The why behind a low support score

Ask Q13 before Q12 in your customer support survey questions, because an unresolved issue explains a low rating.

Product and Feature Use

Use when: a team has used a feature for a while, and you want to separate users from non-users.

#	Question	Scale	What it tells you
19	How satisfied are you with [feature]?	1 to 5, very dissatisfied to very satisfied	Feature-level satisfaction
20	How often does your team use [feature]?	Daily / weekly / monthly / never	Whether a low score comes from users or non-users
21	How easy is [product] to use day to day?	1 to 5, very difficult to very easy	Everyday usability
22	How reliable has [product] been over the last month?	1 to 5, very unreliable to very reliable	Perceived bugs and downtime
23	Which one improvement would make [product] most useful to your team?	Multiple choice (4 to 6 options) plus “other”	Roadmap input with one forced pick
24	How well does [product] work with the other tools your team uses?	1 to 5, not at all well to extremely well	Integration gaps

Ask Q20 beside Q19, since a low score from a team that never uses the feature says little about the feature.

Value, Renewal, and Loyalty

Use when: an account is approaching renewal and the buyer and champion answer.

#	Question	Scale	What it tells you
25	How likely is it that you would recommend [product] to a friend or colleague?	0 to 10	Net Promoter Score
26	How likely are you to renew when your contract ends?	1 to 5, very unlikely to very likely	Renewal risk ahead of the date
27	How well does [product] help you reach the goal you bought it for?	1 to 5, not at all well to extremely well	Value delivered against the promise
28	How does [product] compare with other tools you have used for this job?	1 to 5, much worse to much better	Your relative position
29	If you could change one thing before your renewal, what would it be?	Open text	The agenda for the renewal call
30	How likely are you to add seats or teams in the next year?	1 to 5, very unlikely to very likely	Expansion signal

In our judgment, Q28 is the one competitor question worth asking, because it shows your relative position without a feature-by-feature debate.

Cancellation and Churn

Use when: an account gives notice or reaches the cancel step, and the person who owns that decision answers.

#	Question	Scale	What it tells you
31	What is the main reason you are cancelling?	Multiple choice plus “other”	The churn reason, countable
32	What did you expect from [product] that it did not deliver?	Open text	The expectation gap sales or onboarding created
33	Which tool, if any, are you moving to?	Open text	Where churned accounts go
34	Overall, how satisfied were you with [product]?	1 to 5, very dissatisfied to very satisfied	Exit CSAT, compared with the baseline
35	What would have made you stay?	Open text	What to offer the next account at risk

Put Q31 first: a countable reason lets you sort churn by cause before you read the open answers.

Open-Ended Feedback Questions

Use when: you follow a rating and want customer feedback survey questions that explain it in the customer’s own words.

#	Question	Scale	What it tells you
36	What is the main reason for your score?	Open text	The why behind the rating it follows
37	What is one thing we should keep doing?	Open text	What not to break
38	What can we improve first?	Open text	The first fix, in the customer’s words
39	Is there anything we did not ask that you want us to know?	Open text	Issues the survey missed
40	How would you describe [product] to a colleague?	Open text	The words customers use for your product

Put Q36 straight after the rating it explains. Treat the other customer feedback questions as a pool, and rotate them instead of asking all five at once.

Who to Ask in a B2B Account

One account holds many people with different jobs. In Gartner’s 2024 survey of 632 B2B buyers, buying groups ranged from five to 16 people across as many as four functions.

Your survey reaches the same mix of roles, so ask each role only what it can see:

• Economic buyer: can answer Q3, Q26, Q27 and Q30; cannot judge daily usability or a single ticket.
• Day-to-day user: can answer Q12 to Q22; cannot judge price, value or renewal.
• Admin or champion: can answer Q7, Q9, Q10, Q24 and Q28; sees setup and integrations, and often speaks for the users.

The person who signs is rarely the person who uses the product. In a B2B customer satisfaction survey, record the role with every answer and read each question from the role that can see it.

Question Types and Answer Scales

The question type decides what you can do with the answer: rating questions track change, open-ended questions explain it, and binary questions record a fact.

Type	Example	Use when	Watch out
Rating (Likert) scale	Q1: How satisfied are you with [product] overall?	You will track the number over time	Change the scale and you lose the trend
Binary (yes/no)	Q8: Has your team completed [first key task] yet?	The answer is a fact	A yes/no on satisfaction has no degree
Multiple choice	Q31: What is the main reason you are cancelling?	You need to count reasons	Always add “other”
Open-ended	Q36: What is the main reason for your score?	You need the why	Each open question adds time to the survey

Likert scale questions give you the number to track over time. Use yes/no questions only where the answer is a fact, such as whether the first key task is done.

Write the rating scale as 1 to 5 with both ends labelled, from very dissatisfied to very satisfied. The 5-point scale is the norm in Survicate’s 2025 data.

We keep the midpoint, 3 (neither satisfied nor dissatisfied), and use balanced labels, never an all-positive scale. That is our judgment: an all-positive scale leaves no room to report a problem.

If you will chart the answer, use a rating scale; if you will act on it, add one open question beside it.

A Ready-to-Send Survey for a B2B Account

These eight of the 40 questions make one survey for an account’s relationship check, sent in this order. The overall rating goes first so later questions do not colour it, following Pew’s order-effect principle that earlier questions give context to later ones.

Most customer satisfaction survey examples stop at a list. This one is ordered, and each question feeds a named decision.

We picked one overall score and one open question at the end. In between sits one question for each decision a B2B team acts on: fit, usability, account team, value, renewal and loyalty.

#	Question	Scale	Decision it feeds
1	How satisfied are you with [product] overall? (Q1)	1 to 5, very dissatisfied to very satisfied	Baseline trend
2	How well does [product] meet your team’s needs? (Q2)	1 to 5, not at all well to extremely well	Product fit and roadmap
3	How easy is [product] to use day to day? (Q21)	1 to 5, very difficult to very easy	Usability fixes
4	How satisfied are you with how our team communicates with you? (Q5)	1 to 5, very dissatisfied to very satisfied	Account-team coverage
5	How satisfied are you with the value you get for what you pay? (Q3)	1 to 5, very dissatisfied to very satisfied	Pricing and packaging
6	How likely are you to renew when your contract ends? (Q26)	1 to 5, very unlikely to very likely	Renewal risk list
7	How likely is it that you would recommend [product] to a friend or colleague? (Q25)	0 to 10	NPS for the relationship
8	What can we improve first? (Q38)	Open text	The first fix, in their words

Send these client satisfaction survey questions to every named contact on the account and record each person’s role. Read Q3 and Q26 from the economic buyer, who signs the contract, and Q21 from the people who use the product daily.

Eight questions sit above the 2 to 3 question band. In Survicate’s 2025 vendor data, CSAT surveys of 2 to 3 questions complete at 82.5% to 86.3%. At 4 to 10 questions, completion drops to 74.2%.

In Survicate’s 2025 completion-time data across 7,677 surveys, one open-text question raises the median from 41 to 72 seconds. This set trades some completion for decisions, and a team that can send only five questions uses the five-question cut in the FAQ.

Takeaway: Send the full set to one account this week and record each respondent’s role before you compare any scores.

Common Mistakes When Writing Customer Satisfaction Survey Questions

Wording mistakes produce answers you cannot act on. Each row below shows the bad question, the fix and why it works better.

Mistake	Bad question	Fixed question	Why
Leading	“How much did you love our new dashboard?”	“How satisfied are you with [feature]?” (Q19, with the dashboard as the feature)	The wording suggests the answer, so the score measures your phrasing
Double-barreled	“How satisfied are you with our product and support?”	Two questions: “How satisfied are you with [product] overall?” (Q1) and “How satisfied are you with the support you received?” (Q12)	One answer cannot score two things, and Pew Research Center advises asking one question at a time
Double negative	“Do you disagree that setup was not difficult?”	“How easy was it to set up [product] for your team?” (Q7)	Pew notes that double negatives confuse respondents
Yes/no with no frame	“Are you satisfied?”	“How satisfied are you with [product] overall?” (Q1)	There is no degree and no moment, so a yes tells you little
Hypothetical	“Would you use a mobile app if we built one?”	“How often does your team use [feature]?” (Q20), asked about a feature that exists	In our judgment, a hypothetical answer is a guess, so ask about what the team does today
All-positive scale	“Good / very good / excellent”	The balanced labels of Q1, from very dissatisfied to very satisfied	An unhappy customer’s only option is the lowest label, which still reads as positive

Two biases sit behind these rows. Acquiescence bias means some respondents agree with any agree-disagree statement, and Pew’s fix is to offer alternative statements. Q14 is the bank’s one such item. Keep it only to compare effort over time; otherwise use Q7’s “How easy was it…” wording.

The order of answer options can also change answers, so rotate the options in Q23 and Q31. Most customer satisfaction survey best practices come down to one question, one idea and neutral words.

When to Send It and How Long to Make It

Send each customer satisfaction survey right after the moment it asks about, and keep it to the questions that feed a decision. The timing and channel below are our judgment, not a researched rule.

Moment	When to send	Channel	Questions to use
Support ticket closed	Right after the ticket is marked solved	Email or the support widget	Q13, Q14, Q12, Q18
Onboarding	When the team completes its first key task, not on a fixed date	In-product or email	Q6 to Q11 (pick 3)
Feature use	In-product, after the feature is used	In-product	Q19, Q20, Q21
Before renewal	Early enough to act on the answers before the renewal conversation	Email from the account owner	The ready-to-send set
Cancellation	At the cancel step	In-product	Q31 to Q35

On channel, Survicate’s 2025 data shows mobile surveys had the highest median CSAT response rate of the formats it could measure, at 20.2%. If your product has a mobile app, run the in-product rows there. Link-shared surveys are left out of that analysis because their views cannot be counted.

On length, Survicate’s data shows surveys of 2 to 3 questions complete best and open-text questions add the most time. Cut any question that feeds no decision.

Run one survey per moment, and never put two moments in one survey.

How to Score a Customer Satisfaction Survey

To score a customer satisfaction survey, divide satisfied responses by all responses and multiply by 100. Count a 4 or 5 on the 5-point scale as satisfied (top-two box scoring).

CSAT score = satisfied responses ÷ all responses × 100

Per-account CSAT = the average of each account’s own CSAT

ARR-weighted CSAT = sum of (account ARR × account CSAT) ÷ total ARR

Take one survey sent to eight accounts that gets 55 responses, 49 of them satisfied. The per-response score is 89.1% (our calculation). The table shows how those responses split by account, with ARR in US dollars.

The example is our own: eight accounts with made-up ratings and ARR, with every score computed in a script we rerun.

Account	ARR (USD)	Responses	Satisfied	Account CSAT
Account A	$240,000	3	0	0.0%
Account B	$60,000	12	11	91.7%
Account C	$45,000	10	9	90.0%
Account D	$30,000	9	9	100.0%
Account E	$25,000	8	8	100.0%
Account F	$20,000	6	5	83.3%
Account G	$15,000	4	4	100.0%
Account H	$10,000	3	3	100.0%

Seven of eight accounts score above 80% and the largest scores 0%

8 accounts, 55 responses, 5-point scale, top-two box; hypothetical

Account D	100.0%
Account E	100.0%
Account G	100.0%
Account H	100.0%
Account B	91.7%
Account C	90.0%
Account F	83.3%
Account A	0.0%

A CSAT score counted per response can hide a churn risk. In our example, 49 of 55 responses are satisfied (89.1%), but weighted by annual recurring revenue (ARR) the score falls to 43.2%.

Counted once per account, the score is 83.1%. Account A holds 53.9% of ARR but gave 3 of the 55 responses (5.5%), and none was satisfied. Without account A, the per-response score would read 94.2%.

Report the per-response score to the product team, and the per-account and ARR-weighted scores for retention risk. Reading the answers per account is where customer satisfaction survey questions and answers turn into a renewal list.

There is no universal good score, so compare each result with your own earlier scores for the same question. As an outside reference only, Survicate’s 2025 median CSAT is 80.0% (vendor data). Half of its surveys score between 64.8% and 90.0%, and B2B and B2C sit 2 points apart.

The US national ACSI was 76.9 out of 100 in Q4 2025 ACSI, Feb 2026. The ACSI’s quarterly table (Aug 2026) puts Q2 2026 at 76.1. It is an index on a different scale, so do not compare it with a top-two-box percentage.

Takeaway: Report the ARR-weighted score next to the per-response one before any renewal review, so a quiet large account cannot hide behind a high average.

What to Do About a Low Score

Closing the loop on a low score takes three steps. Ask one question tied to a decision, tell the customer what happened with the answer, and change the process it points to.

Bain’s Net Promoter System calls the employee-level version the inner loop. Feedback goes to the people who shaped that customer’s experience, and it has to be granular, timely and lead to follow-up.

Owners and actions are our recommended split, built for the ready-to-send survey in the same order.

Question	Owner	First reply	What to change
Q1 Overall satisfaction	Account manager	Same business day when the score is 1 or 2	Call the respondent, find the moment behind the score and log the cause
Q2 Fit with team needs	Product manager	Next business day	Ask which need is unmet and add it to the next roadmap review
Q21 Day-to-day ease of use	CS lead	Next business day	Book a usage review and log the friction for product
Q5 Team communication	Account manager	Next business day	Name one owner for updates and agree a contact rhythm
Q3 Value for price	CS lead	Next business day	Review the plan and seat mix with the buyer before renewal
Q26 Likelihood to renew	CS lead	Same business day	Add the account to the renewal risk list and book a call with the buyer
Q25 Likelihood to recommend	Account manager	Next business day	Ask scores of 0 to 6 what one change would raise it, and note 9 to 10 as reference candidates
Q38 First improvement	Product manager	Next business day	Tag the theme and report it back to the customer when it ships

A low score from one small account is one data point. The same low score across several accounts is a pattern for the product team.

Which Moment Has No Question Yet?

Pick one moment, send the short set after it, and score the answers by account rather than by response. The CSAT calculator turns your rating counts into a satisfied share per account. Between onboarding and renewal, which moment in your customers’ year has no question today?