Customer surveys are short sets of questions that ask people who buy from you what they think, feel, and expect. Done well, they tell you why customers stay, why they leave, and what to fix first.
The hard part is not writing questions. It is picking the right survey, sending it at the right moment, and acting on what comes back.
In this guide, we’ll cover the survey types, how to choose and run one, and how to use the results.
What are customer surveys?
A customer survey is a structured set of questions that collects feedback directly from existing or recent customers. It measures things like satisfaction, loyalty, effort, product fit, and expectations.
Most customer surveys mix two kinds of questions:
- Closed questions give you numbers you can track, such as a 1 to 5 rating.
- Open-ended questions give you the reason behind the number, in the customer’s own words.
Customer surveys can run once, such as a post-purchase check, or repeat on a schedule, such as a quarterly loyalty pulse. The data they produce is often called the voice of the customer, meaning what customers say about your product and service in their own terms.
Customer surveys are often confused with feedback forms, interviews and polls. Each collects opinions, but they differ in depth, scale and purpose.
| Term | What it is | When it fits |
|---|---|---|
| Customer survey | A planned set of questions sent to a defined group of customers | You need comparable data across many customers |
| Feedback form | An always-on form where customers volunteer comments | You want unprompted issues and ideas |
| Customer interview | A one-on-one conversation led by a researcher | You need depth on a small number of customers |
| Poll | One question, usually with instant results | You need a fast read on a single choice |
| Market research survey | A survey of prospects or the wider market | You need to understand people who are not yet customers |
Why customer surveys matter
Customer surveys matter because they put a measurable number and a reason on customer sentiment before it shows up as lost revenue.
Retention is the clearest business case. Harvard Business Review reports that acquiring a new customer can cost anywhere from five to 25 times more than retaining an existing one. The same article cites Bain & Company research showing that a 5% increase in retention can lift profits by 25% to 95%.
Surveys also help you stay in step with the market at large. The American Customer Satisfaction Index (ACSI) is the national cross-industry measure of customer satisfaction in the US economy. It shows that satisfaction moves with prices, service quality, and economic conditions, so your own scores need regular checking.
In practice, customer surveys help teams to:
- Spot problems before customers churn
- Rank fixes by how many customers they affect
- Test ideas before building them
- Show customers that their input changes something
- Give leaders a trend line instead of anecdotes
Types of customer surveys
The best-known customer survey types are CSAT, NPS, and customer effort score. Beyond those, many teams run surveys tied to a specific moment in the customer journey.
Here is how the types group together:
- Metric surveys: CSAT, NPS, and CES, which produce a trackable score
- Moment surveys: Onboarding, post-purchase, support and churn surveys, which are triggered by an event
- Discovery surveys: Product feedback and voice of the customer surveys, which explore needs and ideas
Customer satisfaction (CSAT) surveys
A CSAT survey asks customers how satisfied they are with a specific interaction, product, or service. The usual question is “How satisfied were you with your experience today?” on a 1 to 5 scale.
To calculate the score, divide the number of satisfied responses (4 or 5) by the total number of responses and multiply by 100. A result of 80% means four in five customers were satisfied. You can see the full method in this guide on how to measure customer satisfaction.
CSAT works best right after an event, because memory fades fast.
Net Promoter Score (NPS) surveys
An NPS survey asks one question: “How likely are you to recommend us?” Customers answer on a 0 to 10 scale. The Net Promoter System groups them into three bands:
- Promoters (9 to 10): Loyal customers who keep buying and refer others
- Passives (7 to 8): Satisfied but open to competitors
- Detractors (0 to 6): Unhappy customers who may spread negative word of mouth
Subtract the percentage of detractors from the percentage of promoters. The result runs from -100 to 100. Pair it with an open follow-up such as “What is the main reason for your score?” Read more on the NPS score formula and benchmarks.
NPS measures loyalty over time. It suits relationship surveys sent quarterly or twice a year.
Customer effort score (CES) surveys
A customer effort score survey measures how easy it was to get something done, such as resolving an issue or completing a purchase. The customer rates a statement like “The company made it easy to handle my request” from strongly disagree to strongly agree.
Low effort predicts loyalty well in service settings, because customers rarely leave over a quick fix but often leave over a hard one. Learn how to run it in this guide to customer effort score.
Lifecycle and touchpoint surveys
Lifecycle surveys fire at a specific point in the customer journey. Each asks about one moment, so the answer points to one fix.
| Survey | Trigger | What it tells you |
|---|---|---|
| Onboarding | First week after signup or purchase | Where setup or first use confuses people |
| Post-purchase | After delivery or checkout | How ordering, shipping, and product match expectations |
| Customer support | Right after a ticket closes | Whether service solved the problem |
| Product feedback | After a feature is used several times | Which features matter and what is missing |
| Website usability | After a task, or on exit | Where navigation fails |
| Churn or exit | On cancellation or long inactivity | The real reason customers leave |
| Voice of the customer | Quarterly or yearly | Needs and expectations in open language |
How to choose the right type of customer survey
Start with the decision you need to make, then pick the survey that answers it. A survey without a decision attached produces numbers nobody uses.
If you are still deciding between the three core metrics, this comparison of NPS vs CSAT vs CES walks through the trade-offs. As a quick rule, match the question you need answered to the survey:
- How loyal are customers overall? Use NPS, quarterly or twice a year.
- How did one interaction go? Use CSAT, within minutes of the interaction.
- Is a process too hard? Use CES, right after the task.
- Why do customers cancel? Use a churn survey, at the moment of cancellation.
- What should we build next? Use product feedback or voice of the customer surveys, after regular use.
Two practical limits apply. Do not survey the same customer twice in a short window. And send each survey to the group that had the experience, not your whole list.
Step-by-step guide to creating a customer survey
A good customer survey takes six steps: set a goal, define the audience, write the questions, choose the channel, test it, then send.
- Set one goal.
Write it as a decision, such as “Decide whether to simplify checkout.”
- Define the audience.
Pick the customers who had the experience, such as buyers from the last 30 days.
- Write the questions.
Use one rating scale per survey, add one or two open-ended questions, and keep each question to one idea.
- Pick the channel.
Email, in-app, SMS, web link, QR code, or in-person all work. Match the channel to where the customer already is.
- Test it.
Have five colleagues or customers take it. Fix anything confusing, and check it on a phone.
- Send and set a close date.
Send at the right time and plan one reminder.
Keep the whole survey under about five minutes. Use skip logic so customers only see questions that apply to them.
Customer survey question examples
Good customer survey questions are specific, neutral, and tied to one idea. The examples below cover the most common goals. For many more, see this list of customer survey questions.
- Satisfaction: “Overall, how satisfied are you with your recent purchase?” (1 to 5)
- Loyalty: “How likely are you to recommend us to a friend or colleague?” (0 to 10)
- Effort: “How easy was it to get your issue resolved?” (1 to 7)
- Product fit: “Which feature do you use most, and what is missing?” (open-ended)
- Service: “How would you rate the helpfulness of the person who assisted you?” (1 to 5)
- Expectations: “How well did our product match what you expected?” (much worse to much better)
- Churn: “What was the main reason you stopped using us?” (multiple choice plus other)
- Improvement: “If you could change one thing about us, what would it be?” (open-ended)
Avoid leading wording such as “How great was our service?” It pushes customers to agree and skews your scores.
Customer survey examples in practice
The scenarios below are illustrative. They show how survey type, trigger, and question fit together in four common settings:
- Food delivery: CSAT after each order
- Ride-hailing: Separate rider and driver surveys
- Video calling: A one-tap rating after each call
- Local clinic: A CES-style text after each visit
Food delivery
A delivery app sends a two-question CSAT after each order: a 1 to 5 rating and one open box asking what went wrong. Late orders and missing items surface within days, and the team can fix routes or restaurant partners.
Ride-hailing and transport
Two-sided businesses need a survey for each side, because each sees a different half of the problem.
- Rider survey: Rating, safety, vehicle condition, and fare clarity
- Driver survey: App reliability, pay clarity and pickup experience
Video calling and messaging software
A calling app shows a one-tap thumbs up or down after a call ends. A thumbs down opens a short list of causes, such as audio, video, or connection. The structured answer feeds the product team’s bug list directly.
Local clinic or service business
A clinic sends a text link 24 hours after a visit with a CES-style question about booking and check-in. It adds one open comment at the end.
- What staff learn: Scheduling friction that satisfaction scores alone would hide
- What they change: Reminder timing and check-in steps
How to measure customer survey results
Two kinds of numbers measure a customer survey. Participation metrics show whether the data is trustworthy. Score metrics show how customers feel.
| Metric | What it shows | How to calculate |
|---|---|---|
| Response rate | How many invited customers replied | Responses divided by surveys delivered, times 100 |
| Completion rate | How many who started finished | Completed divided by started, times 100 |
| CSAT | Share of satisfied customers | Responses of 4 or 5 divided by total, times 100 |
| NPS | Net loyalty | % promoters minus % detractors |
| CES | Average ease of effort | Mean of the effort ratings |
For ways to lift participation, see this guide to survey response rates.
How to analyze and act on customer survey results
Analysis turns raw answers into a short list of actions. The goal is a decision, not a report.
Start by splitting results by segment, such as new vs. long-time customers, plan type or region. An overall NPS of 40 can hide a score of 10 in one segment.
Then read the open-ended answers. Group them into themes, count how often each appears, and match each theme to a score. Themes that appear often among detractors are your first fixes. Text analysis tools can speed up the grouping on large volumes.
Finally, close the loop. Reply to unhappy customers quickly, then tell the wider group what you changed. This practice is called closed-loop feedback, and it raises future response rates because customers see their input matters.
Common customer survey mistakes
Most failed customer surveys fail for the same reasons. Each one is easy to avoid once you know it.
- Too long.
Long surveys cause drop-off and rushed answers. Cut anything that will not change a decision.
- Leading or double-barreled questions.
“How fast and friendly was our staff?” asks two things at once.
- Surveying everyone with the same form.
Send the form that matches the experience. - Asking too often.
Survey fatigue lowers response and goodwill.
- No follow-up.
Customers who see no action stop replying.
- Reading the score without the reason.
A number alone does not tell you what to fix.
- Ignoring bias.
If only happy or angry customers reply, your score is skewed. Compare responders with your full customer base.
Customer surveys with QuestionPro
QuestionPro’s customer experience platform keeps survey collection, analysis, and follow-up in one workflow, so a low score leads to an action instead of a spreadsheet.
With customer experience software, teams can:
- Send NPS, CSAT, and CES surveys, with AI help drafting the questions
- Run AI sentiment scoring and theme extraction on open-ended answers
- Turn a detractor response into a ticket routed to an owner with a suggested fix
That last step is what most teams lack. Many can collect scores. Fewer can act on every unhappy reply within a day.
Customer surveys work when someone owns the answers
A customer survey is only as useful as the action that follows it. Pick one decision, send one well-timed survey to the right customers, and name one person to own the result.
Start small. A short CSAT after support and an NPS twice a year give you a baseline. Add lifecycle surveys once you can act on what you already collect.
Frequently Asked Questions (FAQs)
Limit relationship surveys, such as NPS, to once or twice a year per customer. Transactional surveys can follow each key event, but cap total requestsfor example, one survey per customer per 30 days, to prevent fatigue.
Anonymous surveys often draw more candid answers, but you cannot follow up with unhappy customers. Many teams use identified surveys for service issues and anonymous ones for sensitive topics like staff behavior. Say which type it is up front.
Small incentives, such as a prize draw, can raise response rates. They can also attract people who only want the reward and bias answers. Keep rewards modest, offer them to everyone who finishes, and never tie them to a positive rating.
For a large customer base, about 385 completed responses (see the sample size math) gives a 95% confidence level with a 5% margin of error. Smaller customer lists need fewer. If you can survey everyone, aim for the highest response rate rather than a sample.
Yes. A short form, an email, or a QR code on a receipt can work for a few dozen customers. As volume grows, software helps with scheduling, analysis, and follow-up, which are hard to manage by hand.
