What is the best way to get feedback from customers?
Talk to a handful of recent customers about what they actually did, and send everyone else a survey of three questions or fewer, timed just after a moment that matters (the first result, a renewal, a cancellation). Then log the exact words, act on one theme, and tell the people who raised it. Pick the method by what you need to learn:
| Method | Best for | Size for a small B2B team | Watch out for |
|---|---|---|---|
| 20-minute call | Why people bought, what nearly stopped them, the words they use | 5 to 8 customers a month | Answers can be more flattering with you on the line |
| Three-question email | A quick read across every customer | Everyone past a milestone | Quiet customers don't answer |
| Product market fit survey | Whether a segment would miss you | About 40 responses | Needs active users, not one-off buyers |
| Lost-deal or cancellation call | Objections and gaps you can fix in the pitch | Every loss you can get on the phone | Ask about events, not blame |
| Watching usage | Where people get stuck | 3 to 5 sessions or screen shares | Small samples; treat as clues |
The sizes are our suggestions, except the 40 responses, which comes from Superhuman's method in our guide to the product market fit survey. That post covers the four-question survey in full, so this one focuses on calls, short emails and turning answers into sales material.
How do you ask questions that get honest feedback?
Ask about one thing at a time, ask open questions before multiple choice, and ask what happened rather than what people would do. The first two rules come from Pew's survey methodologists, the third from usability research.
The Pew Research Center's guide to writing survey questions shows how much the format changes answers. After the 2008 US presidential election, 58% chose the economy as their top issue when it was offered in a list, but only 35% named it when asked an open question. In the open version, 43% gave an answer that wasn't on the list at all. A multiple-choice feedback form tells customers what you think the options are; an open question tells you what they think.
The same guide warns against double-barrelled questions that ask about two things at once, against agree-disagree statements (less educated and less informed respondents are more likely to agree with such statements, which Pew says is sometimes called acquiescence bias), and notes that social desirability bias can be greater when an interviewer is present. It also says that closed questions placed before an open one make people much more likely to mention the earlier ideas, which is why the open question goes first.
Jakob Nielsen of Nielsen Norman Group made the behaviour point in a 2001 article, updated since: pay attention to what users do, not what they say, and don't believe predictions about future behaviour. Nielsen and Jonathan Levy found a 0.44 correlation between users' measured performance and their stated preference across 113 pairwise interface comparisons, and Nielsen advises collecting preference data only after people have used a design.
| Instead of | Ask | Why |
|---|---|---|
| How satisfied are you with our onboarding and support? | What was the hardest part of getting started? | One topic, open answer |
| Do you agree the reporting saves you time? | Walk me through the last time you pulled a report. | No agree-disagree, real event |
| Would you use a Slack integration? | Where do you share these results today? | Past behaviour, not a prediction |
| Why did you leave? | What happened in the week you decided to cancel? | Events instead of blame |
How long should a customer feedback survey be?
Our suggestion: short enough to answer on a phone in under two minutes, which means three questions, one of them open. In a web survey experiment published in Public Opinion Quarterly in 2009, Mirta Galesic and Michael Bosnjak varied the stated length between 10, 20 and 30 minutes. The longer the stated length, the fewer people started and finished, and answers to questions later in the questionnaire were faster, shorter and more uniform. Your best answers come early, so put the question you care about first.
Question 3 is worth keeping: the answer names your real competitor, which may be a spreadsheet or a person rather than another vendor. Send one reminder after three days, then stop.
Does a low response rate make the feedback useless?
Not by itself; what matters is whether the people who stayed silent differ from those who replied. The American Association for Public Opinion Research (AAPOR) says on its response rates page that the link between response rates and survey quality has become much less clear, and that some surveys with less than optimal response rates have shown the least bias. Its Standard Definitions page describes nonresponse error as what happens when nonrespondents differ from respondents in ways that matter to the survey.
For a small customer base, you can see the size of that risk with simple arithmetic. A sample example, Northwind Ltd, has 60 customers and emails them all:
| Figure | Arithmetic | Result |
|---|---|---|
| Replies | 18 of 60 | 30% response rate |
| Happy among replies | 15 of 18 | 83% |
| Happy if every non-responder is happy | (15 + 42) / 60 | 95% |
| Happy if no non-responder is happy | 15 / 60 | 25% |
The real figure sits somewhere between 25% and 95%, and the 83% headline says little about which end. Narrow it by phoning five of the 42 who didn't reply, starting with accounts whose usage has dropped; the signs in our guide to spotting a customer about to churn tell you which ones.
What should you ask on a customer feedback call?
Ask about the moments around the purchase and the last time they used the product, in their words, and say little yourself. Keep it separate from selling; if you run regular account calls, our check-in that isn't a sales call explains why mixing the two backfires.
Record the call or take notes with their permission. Siela's meeting notes record on your own device and pull out decisions, owners and follow-ups, which leaves you free to listen. Afterwards, paste the exact phrases into a feedback log with four columns: customer, quote, theme, and what you did about it.
How do you use customer feedback to sell?
Sort every quote in the log into one of four piles, and each pile has a sales job.
| Pile | Example quote (sample) | Sales use |
|---|---|---|
| Trigger | Our second sales hire quit and the pipeline sat untouched for a month | The opening line of cold outreach to companies in the same spot |
| Objection | We nearly didn't buy because setup looked long | A line in your demo and pricing page that answers it before it's asked |
| Result | We stopped rebuilding the weekly report by hand | A testimonial, with written permission |
| Request | I wish it exported to our accounting tool | Build it or not; either way, tell the customer, which reopens the conversation |
Two rules apply to the result pile. Ask before you quote anyone publicly. And if you tidy their words, the FTC's Endorsement Guides say an endorsement may not be presented out of context or reworded so as to distort the person's opinion or experience; quotation marks signal exact words. Our guide on getting customer testimonials covers the asking and the FTC's review rules.
What to do this week
Pick your moment
Choose one trigger, such as 30 days after signup or the week after a cancellation, and list every customer who passed it in the last 90 days.
Send the three-question email
From your own inbox, plain text, one reminder after three days.
Book five calls
Three happy customers, two quiet ones. Use the script.
Fill the log and sort the piles
Rewrite one outreach opening line using a trigger quote, and answer one objection on your site.
Close one loop
Fix one small request and email everyone who asked for it.
Frequently asked questions
What are the best ways to get feedback from customers?
Short calls with a few recent customers give you reasons and wording; a survey of three questions or fewer gives you breadth; calls after lost deals and cancellations give you objections. Watching people use the product shows where they get stuck, which they often can't tell you. Combine one call method with one written method.
How do you ask a customer for feedback by email?
Send it from a named person's inbox, in plain text, with no more than three questions and one of them open. Say how long they have been a customer, ask what was harder than it should be, and promise you read every reply. Send one reminder after three days.
What questions should you ask for customer feedback?
Ask about real events: what made them start looking, what nearly stopped them buying, and the last time they used the product. Avoid agree-disagree statements, questions that cover two topics, and predictions such as whether they would use a new feature, because what people say they will do often differs from what they do.
What is a good response rate for a customer feedback survey?
There isn't a number that guarantees good data. AAPOR, the American Association for Public Opinion Research, says response rate information alone isn't enough to tell how much nonresponse error a survey has. What matters is whether non-responders differ from responders, so check a few silent customers by phone, especially those whose usage has dropped.
How many customers should you interview?
Our suggestion for a small B2B team is five to eight calls a month, which fits around other work. Keep going until new calls stop producing new themes. For a scored survey such as the product market fit test, aim for around 40 responses, the point at which Superhuman's founder said results start to be directionally correct.
Sources
- Pew Research Center: Writing Survey Questions (checked 4 Nov 2026)
- AAPOR: Response Rates (checked 4 Nov 2026)
- AAPOR: Standard Definitions (checked 4 Nov 2026)
- Nielsen Norman Group: First Rule of Usability? Don't Listen to Users (Jakob Nielsen) (4 August 2001, updated 17 January 2020)
- Galesic and Bosnjak, Effects of Questionnaire Length on Participation and Indicators of Response Quality in a Web Survey, Public Opinion Quarterly 73(2) (Summer 2009)
- FTC Endorsement Guides, 16 CFR Part 255 (eCFR) (checked 4 Nov 2026)