A customer feedback survey is only worth sending if people reach the end of it. Here is how to choose the right survey, what to ask, and where to send the answers. Then build it free, with no cap on responses.
Unlimited responses on the free plan. All 21 question types included.
5
Questions, our recommended ceiling
∞
Responses on the free plan
21
Question types, all included
$0
To run every survey on this page
The basics
A customer feedback survey is a short set of questions sent to people who have already used your product or service, asking them to rate an experience and then say why. The rating gives you a number you can track from one week to the next. The why is the part that tells you what to change.
Most feedback surveys fail the same way. They get written for the person collecting the data rather than the person answering it: twenty questions, a matrix grid, a mandatory email field before anyone has been asked a thing. People start, get tired, and close the tab. What is left skews toward whoever had the patience to finish, which is rarely your average customer.
So before you write a single question, settle four things. If you can answer these, the survey almost writes itself, and our survey questions guide covers the wording in more depth.
Pick one
Five jobs, five different moments. Run one properly before you run a second.
Send it within 24 hours of a finished interaction: an order delivered, a support ticket closed, an onboarding call wrapped up. Any later and people are rating a memory of a memory.
The one question: “How satisfied were you with the help you got today?” on a 1–5 rating, followed by “What would have made that a 5?”
Send it 30 to 90 days after someone starts, then on a fixed cycle. NPS is about the relationship, not a moment, so it needs time to be worth anything.
The one question: “How likely are you to recommend us to a friend or colleague?” on a 0–10 slider scale.
NPS surveysSend it a few days after delivery, once the thing has been opened and used. This is the only survey where you can still reach the buying decision while it is fresh.
The one question: “What almost stopped you from ordering?” as a long text answer. The objections you read here belong on your product page.
Send it inside the cancellation flow itself, not by email the next morning. At that moment people are honest, briefly, and then they are gone.
The one question: “What made you decide to leave today?” as single choice with five reasons and an “Other” box. Keep it to one screen.
Send it when you have a decision to make and no evidence. This one is not on a schedule: you pick a segment on purpose, ask, and then stop asking.
The one question: “What is the one thing you wish this did that it doesn’t?” Open text, no options, no prompting.
Start with CSAT. It is the cheapest to run, the fastest to read, and it attaches to something you already do every day. Once you have two months of it, add NPS on a quarterly cycle.
Open the CSAT templateCopy these
Eighteen questions we would actually send, grouped by survey. Take two or three, not the whole list.
Length
Every question you add is another chance for someone to quit. A conversational form hides that cost by showing one question at a time, so the survey never looks like work. But the honest fix is still a shorter survey.
1 →
1 is very dissatisfied, 5 is very satisfied.
A CSAT screen as respondents see it. Scores of 3 or lower jump to the open follow-up; 4 and 5 skip it.
After the send
A feedback survey that only lives in a dashboard nobody opens is a survey that changes nothing. Push every response to the place your team already works.
Google SheetsEvery response appended as a new row
SlackA message per response in the channel you pick
NotionDatabase rows with your own field mapping
HubSpotCreates or updates the contact record
KlaviyoSubscribes the respondent to a list
n8nSigned webhook, no n8n credentials neededEvery integration is per-user OAuth, so you connect your own account and the data lands in your workspace. All of them are on the free plan. See all integrations.
Analytics
The completion rate tells you whether the survey works. The per-question drop-off funnel tells you which question broke it, and it is almost always the one you were least sure about.
Q3 CSAT: post-delivery
2,418
Views
1,902
Submissions
78%
Completion rate
41s
Average time
Example figures, shown to illustrate the layout of the Results tab.
This week
No project plan, no stakeholder workshop. Pick the moment, write four questions, and send it to the next hundred people who reach that moment.
An order delivered, a ticket closed, a trial ending. Write it down as a sentence so the timing is not up for debate later.
One rating, one open follow-up behind conditional logic, one optional demographic, one catch-all. Steal them from the list above.
A public link in the confirmation email, or an embed on the thank-you page. The share and embed options take a minute to set up.
Check the completion rate and the drop-off funnel first, then read the open answers in one sitting. Fix one thing. Then keep sending.
Templates
Preview any template live, then clone it onto your account in one click and change the wording.
How satisfied were you today?
A 1–5 rating and a conditional follow-up. The one to send after a finished interaction.
Feedback · FreeHow likely are you to recommend us?
The 0–10 recommend question and the “why” that makes the score useful.
Feedback · FreeWhich feature do you use most?
Open questions for discovery work, with picture choice for concepts and mockups.
Feedback · FreeWhat's the main reason you're leaving?
Five cancellation reasons and one open box. Short enough to finish on the way out.
Feedback · FreeFive or fewer for a survey sent after a single interaction, and closer to two for a survey sent by email. One rating question plus one open follow-up will answer most of what you need. Every extra question costs you responses, so add one only when you can name the decision its answer would change.
Tie the send to an event rather than to a date on a calendar. CSAT works best within 24 hours of a finished interaction, while the memory is still sharp. A loyalty question like NPS works better 30 to 90 days in, once someone has had enough time with the product to form a real opinion.
CSAT measures satisfaction with one specific interaction, usually on a 1 to 5 rating asked right after it happens. NPS measures how likely someone is to recommend you overall, on a 0 to 10 scale, and it is asked on a cycle rather than after an event. CSAT tells you whether a moment worked. NPS tells you whether the relationship is holding.
There is no single benchmark worth chasing, because the number swings enormously with the channel, the timing and the length of the survey. An in-app prompt after a completed action will always beat a cold email to your whole list. Measure your own baseline in the first month, then change one variable at a time and watch the completion rate move.
Yes. The free plan is free forever, not a trial, and it includes unlimited forms, unlimited responses, all 21 question types, conditional logic and every integration. Typeform's free plan is widely reported to cap you at 10 responses a month, which is the limit most people hit first; pricing checked September 2026, so verify before you decide. Pro is 12 dollars a month and only adds unlimited file storage, branding removal and up to five team members.
Yes, and both are on the free plan. Google Sheets appends every response to a sheet as a new row, and Slack posts a message per response into the channel you choose. You connect your own account with OAuth, so the data lands in your workspace rather than in a shared one.
Get started
Four questions, one link, and every response yours to keep. No response cap waiting to stop you at the worst moment.
No credit card. Unlimited forms and responses on the free plan.