Feedback form best practices: write one people finish

Most feedback forms fail quietly. Nobody complains about them; people simply open them, answer two questions and leave. The replies you do get come from the most patient customers, not the most typical ones, and the form looks fine because you never see the people who gave up.

This guide is about one thing: getting people to the end. It covers how many questions to ask, what order to put them in, when to use open or closed questions, how to lay the form out, and how to find out where people drop off.

Start with the decision, not the questions

Before writing a single question, write down what you will do with the answers. "Decide whether to extend weekend opening hours" is a decision. "Understand our customers better" is not.

Then list the questions that decision needs. Most forms end up with a handful. Anything that doesn't feed a decision is a question you're asking out of curiosity, and your customers are paying for that curiosity with their time.

A useful test: for every question, finish the sentence "If most people answer X, we will...". If you can't finish it, cut the question.

How many questions

Fewer than you think. For a customer feedback form sent after a purchase, visit or support conversation, aim for three to seven questions that can be finished in about two minutes on a phone.

Every question adds a small chance that someone stops. Those chances add up. A form that is twice as long does not get twice the insight; it gets fewer finished replies, and those replies skew towards people with time to spare.

If you have more to ask, you have three options:

• Rotate. Keep three core questions every time, and add one or two from a longer list that changes each month.
• Split. Send a short form to everyone, and a longer follow-up only to people who opt in.
• Use logic. Show a question only to people it applies to. Someone who didn't use delivery doesn't need the delivery questions.

The order that works

Order matters as much as length. A reliable shape:

1. One easy rating first. "How was your visit today?" on a 1–5 scale. It takes a second, commits the person to the form, and gives you a number to count even if they stop afterwards.
2. The most important open question second. "What's the main reason for your score?" People have the most energy at the start, so put the question you most want answered while they still have it.
3. Specific ratings next. Two or three short ratings on the areas you already know matter.
4. An optional open question. "Anything else you'd like to tell us?" Optional, so nobody quits here.
5. Details last. Demographics, contact permission, booking reference. These feel like admin, so they go at the end where a drop-off costs you least.

Avoid starting with personal details. Asking for a name and email before anything else signals that the form is about you, not them.

Open or closed questions

Closed questions (ratings, multiple choice, yes/no) are quick to answer and easy to count. Open questions (a text box) are slower to answer and harder to count, but they're where you learn things you didn't already know.

The best forms use both, in a pattern: a closed question to measure, followed by an open one to explain.

Two practical rules for open questions:

• Keep required open questions to one or two. A form with five required text boxes feels like homework.
• Ask a question with an answer. "Any comments?" invites "no". "What's one thing we could do better?" invites a specific reply.

Don't avoid open questions because they're hard to analyse. That problem has a solution; see how to analyse open-ended survey responses.

Writing each question

Small wording choices change completion and answer quality.

• One idea per question. "How fast and friendly was our service?" can't be answered by someone who found it fast but rude.
• Plain words. "How easy was it to book?" beats "How would you rate the ease of the reservation process?"
• Label every point on the scale, or at least both ends. "1 = Very poor, 5 = Excellent" removes guesswork.
• Keep scales consistent. If one question runs 1–5 and the next runs 1–10, people slow down and make mistakes.
• No leading questions. "How much did you love our new menu?" tells people what you want to hear.
• Give an out. Add "Didn't use this" or "Not sure" where it's honest. Forcing an answer produces made-up answers.

One question per screen or all at once

There are two main layouts, plus a third that's becoming common.

All questions on one page (classic). People can see the whole form, judge its length and skip around. It suits short forms, desktop users, and people who like to know what they're getting into. The risk is that a long page looks daunting before anyone starts.

One question per screen (step-by-step). Each screen asks a single thing. It feels lighter, works well on phones, and lets logic route people without them seeing skipped questions. The risk is that people can't tell how long the form is, so always show progress ("3 of 6").

Chat-style. Questions arrive one at a time as messages. It suits casual audiences and short forms, and can make an open question feel like a conversation rather than a text box.

There is no universally right answer. A sensible default: classic for forms of three or four questions, step-by-step for anything longer or mainly answered on phones. If you're unsure, try both on the same audience for a few weeks and compare completion.

Designing for phones

Most feedback forms are opened from an email, text message or receipt link, which means a phone. Check every form on one before sending it.

• Big tap targets. Rating buttons should be easy to hit with a thumb.
• No tiny dropdowns. For four or five options, show them all as buttons.
• Short text boxes that grow. A huge empty box is intimidating; one that expands as people type is not.
• No horizontal scrolling. A 0–10 scale must fit on a narrow screen.
• Fast loading. A form that takes several seconds to appear loses people before question one.

Where people drop off, and how to measure it

You can't fix drop-off you can't see. Track three numbers for every form:

1. Views. How many people opened it.
2. Starts. How many answered at least one question.
3. Completions. How many reached the end.

Then look at drop-off by question: of the people who reached question 3, how many reached question 4? This is a conversion funnel for your form, and it usually points straight at the problem.

Common patterns and what they mean:

Change one thing at a time and compare completion over the next few weeks. If you change three things at once, you won't know which one helped.

Before you send it

Run through this list:

• Can you name the decision each question feeds?
• Is it finishable in about two minutes on a phone?
• Is the first question an easy rating?
• Are no more than two open questions required?
• Does each question ask one thing, in plain words?
• Does logic hide questions that don't apply?
• Have you tested it on a phone yourself?

For a worked example built on these rules, see our hotel guest feedback survey.

Doing this with Userforms

In Userforms you can describe a form in a sentence and edit what it builds, or start from a survey, recommend-score or signup template. Each form can be classic, step-by-step or chat-style, with logic, calculated fields and scoring.

Every form has its own views of completion and drop-off, and the Pro plan adds a conversion funnel showing where people stop. Open answers are filed into topics with the tone of each mention as they arrive, so long text replies don't sit unread.

You can start for free, or see what each plan includes on the pricing page.