Customer Feedback Surveys That People Actually Finish

A half-finished survey tells you who had time to quit. A finished one tells you what the client thought. Length and wording and the invite decide which result you get. The survey best practices below come from AAPOR and Typeform and Kantar and the Australian Bureau of Statistics.

Two rates people mix up

A response rate is the share of people who finish among the people who got the invite. Typeform defines it that way. Divide completions by invites. Typeform’s example is 45 completions out of 100 invites. That rate is 45 percent.

A completion rate is about people who start and then finish. Typeform’s question-count finding is a completion rate. Kantar’s 5 to 30 percent band is a response rate. Use the matching number when you judge your survey completion rates.

Kantar says an acceptable response rate is between 5 percent and 30 percent. It calls anything above 30 percent excellent. Kantar says there is no universal line for a good rate. Audience and length and design and the day of the week all move it. Typeform says 33 percent is a common benchmark. Treat that as Typeform’s benchmark. It is not a law.

AAPOR says a very small response can miss whole types of people. The estimates can then lean toward whoever had time to answer. A high response rate survey is useful because the missing voices are fewer. It is useful when the questions are ones those people can answer.

Ask for feedback you will use

AAPOR says a survey fits when you need feelings or thoughts or behaviors and you lack current data on the topic. Ask whether the objective is specific. Ask whether another survey has the answer. Ask whether a conversation or a record would serve you better.

AAPOR’s ethics code bars surveys that exist to sell or to raise money or to produce a result you chose in advance. A customer feedback survey can ask what worked. It becomes a pitch when the questions push a purchase.

Write the purpose where the client sees it first. Typeform says a welcome screen can name the purpose and who you are. It says to share how long the survey should take. The Australian Bureau of Statistics says a respondent’s grasp of the task and their reason to finish shape the answers you get. Put help next to the question it explains. Repeat an instruction in the same words each time it applies.

Tell them what you will do with the answers. Typeform says that transparency belongs on the invite and on the welcome screen. AAPOR says to follow the privacy rules that cover personal information. A public file of answers can identify a client even after you remove a name. AAPOR says demographic details plus the answers can do that.

How long people stay

Typeform’s research says surveys with more than six questions have a completion rate under 50 percent. The same research says forms with six or fewer questions had the highest completion rates. Forms that took under a minute had a completion rate 15 percent higher than longer forms.

Kantar says there is no exact time limit. It recommends under 12 minutes for surveys and prefers a survey under 10. The concern is sharpest on a phone. A notification or a queue can pull the person away. Kantar says to get them through before patience runs out.

A short client feedback form can follow Typeform’s question count. A deeper customer satisfaction survey can run longer and should stay inside Kantar’s clock. Say the time on the welcome screen. Typeform says a progress cue shows how much remains. A line such as a few questions left can help. Skip a promise you cannot keep.

Typeform’s research ties email-first forms to a completion rate 9 percent higher than forms that ask for email after other questions. Typeform says to avoid a story about why that happens. If you need the address, test email first against email last on your own list.

Customer survey questions that one person can answer

AAPOR says each question should be specific and cover one concept. Kantar uses this example of a double question. “How satisfied are you with the product’s quality and pricing?” Quality and price can diverge. The person is forced into one score. Ask them apart.

The Australian Bureau of Statistics uses a gym example that joins a frequency question with a duration question. It says people make more mistakes on these questions or skip them when concepts point in different directions. It says to avoid vague words. Its examples are a frequency word and the word “average.” Skip acronyms and jargon and double negatives. Use a precise term. “Screen time” is broader than the concept if you mean time watching shows.

Closed questions give a list. AAPOR says they lighten the work of inventing an answer and are easier to count. They can steer the answer. Open questions use the client’s words. They need coding before you can count them. Kantar says open questions are harder on a phone because of the small screen and the keyboard. Avoid a run of them. One optional open question at the end can be enough. Kantar’s sample line is “Is there anything else you’d like to share?”

Build the answer list with care. AAPOR says the options should not overlap and should cover the reasonable choices. Include “don’t know” or “does not apply” or a neutral choice such as “neither agree nor disagree” when those fit. Put them in a logical order. In a self-administered survey, people tend to pick the first option they see. In an interviewer survey, they tend to pick the last. Rotate the order for half the list and the reverse for the other half if you need to check that pull.

Kantar wants a balanced scale. A five-point scale from “Very satisfied” to “Very dissatisfied” has a midpoint and matching poles. A scale packed with praise is harder to read. A 7 out of 10 can mean different things to different people. If you will report the highest score or the lowest score, Kantar says a yes-or-no question or a multiple-choice question may give clearer data. Kantar says a bank of similar statements should stay at or under 12. Repeated grids invite straight-line answers and random clicks. On a phone, long grids feel tedious.

Kantar says to offer “Not applicable” and “I don’t know” and “Prefer not to answer” as separate choices. A blended “I don’t know / Not applicable” blurs the meaning. AAPOR says a required answer raises the chance someone leaves. Allow a skip on a hard or sensitive item. Code a blank as a blank. AAPOR says that blank needs a different value from “none of the above” or “I don’t know” or “I prefer not to answer.”

Ask people about an experience they had. Kantar says awareness of a brand is not the same as using its service. A filter question can hide the follow-ups. Typeform says logic can route the next questions from a prior answer. The person who bought one service sees questions about that service.

Put a general question before a specific one on the same topic. AAPOR says the specific question can prime the general one. Start with easier items. Save sensitive items for after that. AAPOR says to disclose a sensitive topic at the start or in the moment before those items and to allow a skip. Put overall satisfaction before a question about one staff member or one fee.

AAPOR notes that many U.S. residents speak limited English or none. Most national U.S. surveys offer English and Spanish. Offer the languages your clients use. These sources do not set a language list for a local service business.

When you send it

Kantar says an invite at a fitting day and hour works better than a blast. The day of the week can change the response rate. Typeform says the time of year can change engagement. AAPOR says to contact people more than once and in more than one way and at different times of day. It calls for reasonable efforts with people who have not answered and with people who refused.

These pages do not say to send a survey one hour after checkout or three days after a project. Pick a moment when the client can recall the work. Say that moment in your own test. Keep the question wording stable if you want to compare this month with next month. AAPOR says a wording change can look like a change in opinion. A move from phone to web can do the same on personal questions.

Incentives without turning the survey into a sale

AAPOR says incentives can be money or something else. Typeform says a coupon or a freebie or a chance at a prize can raise completions. It lists softer offers such as “Be the first to know” or a subscriber line. Put the offer on the welcome screen or in the invite. Typeform says to highlight the benefit to them there.

You can tell them you will use the feedback. Say what you will do. Then do that thing on the next job. AAPOR’s ban on selling surveys means the questions stay about the experience. The incentive is the thank-you. The survey is the questions.

Test before you send the list

The Australian Bureau of Statistics says testing with real respondents is required to learn whether questions are understood. Following a guide is not a substitute. AAPOR says a pretest can use cognitive interviews with people like the ones you will survey. A pilot checks the invite and the survey and the cleanup of the file.

Kantar says to take the survey yourself. Ask whether each question makes sense. Ask whether it repeats. Ask whether you would finish it and give a considered answer. Rewrite or cut what fails.

After the survey is live, Kantar says to watch dropout by question. If more than 2 percent of people leave on one question, treat that question as a problem. They may find it hard or dull. Open questions and Likert banks deserve a close look. Kantar describes a click check. An answer option that fewer than 3 percent of people pick adds reading time and weak data. Kantar says the average miss-click rate on multi-choice questions is about 2 percent.

Survey design tips you can apply this week

1. Write one objective. If a sales email would do the job, skip the survey.
2. Draft six or fewer questions for a short pulse. Stay under 10 minutes if the study must run longer.
3. Ask one concept at a time. Use words your clients use.
4. Give a balanced scale or a clear list. Add a real opt-out. Leave hard items optional.
5. Put purpose and time and the incentive on the first screen.
6. Show progress. Route follow-ups from the service they bought.
7. Send the invite at a time you can explain. Send a second contact to people who have not finished.
8. Pretest with a few real clients. Take it on a phone.
9. Code blanks apart from “don’t know.” Compare this round with the last round with the same wording.

You can host that customer feedback survey as a link with Jotform. The client opens the link. You read one set of replies.

Pitfalls

You ask two things at once. Quality and price need two questions. Kantar and the Australian Bureau of Statistics both flag the double question.

You require every item. AAPOR says a forced answer raises the chance someone leaves. Kantar says a forced guess hurts the data.

The list is a wall. Kantar says to shorten lists and to group rare choices. A long grid on a phone invites straight-line clicks.

You change the question and then compare months. AAPOR says the wording is part of the measure. Keep it stable when you want a trend.

You treat a thin set of answers as the market. AAPOR says a thin response can miss types of clients. Report how many people were invited and how many finished.

The survey is a pitch. AAPOR bars surveys used to sell. Ask about the work they bought.

You hide the use of the answers. Typeform says to state how you will use the data. AAPOR says personal details can identify someone in a shared file.

You blast the list at one hour. Kantar says timing and the day matter. AAPOR says to vary the contact.

Trade-offs

Six questions versus a 10-minute study. Typeform’s completion finding favors the short form. Kantar’s clock allows a longer instrument if you stay under 12 minutes and prefer a survey under 10. A pulse after one job can be six questions. A review once a year can be longer if every item earns its place.

Closed lists versus open comments. Closed lists take less time to count and can steer people. Open comments carry their words and take longer to read. Kantar says one optional comment at the end limits the burden.

A coupon versus a plain thank-you. Typeform says a tangible incentive can raise completions. AAPOR includes non-monetary incentives. A coupon is a tangible offer. A clear purpose is a non-monetary reason. Test which one your clients answer.

Email first versus the welcome screen. Typeform’s research links email-first to a higher completion rate. A welcome screen carries the purpose and the time. You can show the purpose and then ask for email. Test the order on your list.

One invite versus several. AAPOR says to try more than once. That effort covers people who have not answered and people who refused. AAPOR calls the effort reasonable.

Web versus a call. AAPOR says online surveys can cost less and can feel more private on sensitive topics. A call can help a person through a hard question. Mail works when you have an address and the path through the form is simple. A complex skip pattern fits a screen better than paper.

What you walk away with

You reduce survey abandonment by asking fewer and clearer questions and by letting people skip what they cannot answer. You learn who was invited and who finished. You keep the wording stable so a change in the score means a change in the work.

The sources leave the exact send time to you. They give you the length and the question rules and the invite habits. Use those. Then read the replies before you change the offer.

If you want one link for the customer survey questions, Jotform can hold the form. The judgment about what to change stays with you.