How to Create a Survey in 8 Steps: Free, Anonymous, and Shareable by QR Code
From a blank page to answers you can act on: one goal, a handful of honest questions, a free form, a link or QR code people can open on a phone, and a short read of the results.
A good survey has one goal, ten questions or fewer, wording that does not lead people to an answer, an anonymity setting that suits the topic, and a share link that opens on a phone. You can build one free in a form builder in about twenty minutes. The eight steps below take you from the question you want answered to the results.
What This Guide Covers
- How to pick the one decision the survey informs
- Question types, the bias rules and 20 questions you can copy
- How to build it in a free form builder, or in Google Forms
- What anonymous really requires, and anonymous vs confidential
- Sharing by link, embed and QR code, with the print-size rule
- Free survey tools compared, limits read October 2026
Building a quiz rather than a survey? The AI quiz maker is the better starting point.
TL;DR Quick Summary
Create a survey in eight steps. 1. Write down the decision the answers will inform. 2. Pick the type: customer feedback, NPS, employee pulse, event, research or course evaluation. 3. Write closed questions first and one open question, with no leading or double-barrelled wording. 4. Put easy questions first and sensitive ones last, and keep it under five minutes. 5. Build it in a free form builder. 6. For an anonymous survey, ask for no name or email and require no login. 7. Share the link, embed it, or print the QR code at least 2 cm wide. 8. Check the response rate and where people stopped, then compare answers by one question.
Build your survey free →Table of Contents
Eight steps from the goal to the results, then the free tools and a word on survey apps.
- Step 1: Decide What You Will Do With the Answers
- Step 2: Choose the Survey Type
- Step 3: Write the Questions
- Step 4: Put Them in Order and Keep It Short
- Step 5: Build the Survey Free in a Form Builder
- Step 6: Make It Anonymous
- Step 7: Share It by Link, Embed or QR Code
- Step 8: Read the Results and Act on Them
- Free survey tools compared
- When you need a survey app instead
- Frequently asked questions
Step 1: Decide What You Will Do With the Answers
Before you write a single question, write one sentence that finishes "When the results are in, we will decide…". That sentence is the survey's goal, and every question has to earn its place against it. If you cannot finish the sentence, do not send the survey yet: people notice when nothing changes after they answer, and they stop answering the next one.
A goal like that also tells you who to ask (people who already visit), when to ask (right after a visit, while it is fresh) and what you do not need to ask. The cafe has no use for age or job title, so they stay out.
Step 2: Choose the Survey Type
The type sets the length, the questions and how often you send it. Pick one; a survey that is half NPS and half market research does neither job well.
| Survey type | What it answers | Typical length | When to send |
|---|---|---|---|
| Customer feedback | How was it, and what should we fix? | 3 to 8 questions | Straight after a purchase or visit |
| Net Promoter Score (NPS) | Would they recommend you, and why? | 2 questions | Every quarter, or after a key moment |
| Employee pulse | How is the team doing this month? | 5 to 10 questions | Monthly or quarterly, same questions each time |
| Event feedback | Was it worth attending, and what next time? | 5 to 8 questions | Within a day of the event |
| Market research | Would people want or pay for this? | 8 to 15 questions | Before you build or launch |
| Course evaluation | Did it teach what it promised? | 6 to 12 questions | At the end of a course or module |
Market research surveys need extra care, because people are poor at predicting what they would buy. If you are testing an idea for an app or a product, pair the survey with the interview and sign-up tests in how to validate an app idea. And if the survey is meant to find where customers get stuck, a customer journey map tells you which moment to ask about.
Step 3: Write the Questions
Most bad survey data comes from the wording, not the tool. Use closed questions (choices, ratings, scales) for anything you want to count, and keep one open question for the reasons behind the numbers.
| Question type | Looks like | Use it for | Watch out for |
|---|---|---|---|
| Multiple choice | Pick one of four options | Facts and clear preferences | Missing options; add "Other" when you are not sure you listed them all |
| Rating scale | 1 to 5 stars | Satisfaction with one thing | Rating two things in one question |
| NPS (0 to 10) | How likely are you to recommend us? | Loyalty, tracked over time | Changing the wording between rounds, which breaks the trend |
| Likert scale | Strongly disagree to strongly agree | Agreement with a statement | Statements that are all positive, which nudges agreement |
| Yes / no | Did you find what you came for? | A simple fact | Questions where the honest answer is "it depends" |
| Open text | What is one thing we should change? | Reasons and ideas you did not think of | More than one or two per survey; people skip them |
Four ways a question biases the answer
Each of these pushes people toward an answer or makes the answer impossible to read. Fix them before you build anything.
Leading
Avoid
How much did you enjoy our new menu?
Ask instead
How would you rate our new menu?
"Enjoy" assumes they did. A neutral verb lets the unhappy customer answer honestly.
Double-barrelled
Avoid
How satisfied are you with our coffee and service?
Ask instead
Two questions: one about the coffee, one about the service.
A low score could mean either, so you cannot act on it.
Loaded
Avoid
Should we stop wasting money on paper cups?
Ask instead
Would you bring your own cup for a 20-cent discount?
The wording carries an opinion. Ask about the behaviour you actually want to know about.
Forced choice
Avoid
Which do you prefer: the morning or the afternoon menu?
Ask instead
Same question, plus "I have not tried both" and "No preference".
Without an honest way out, people pick at random and your numbers look more certain than they are.
The five questions every feedback survey needs
If you are not sure where to start, these five cover most customer feedback surveys. Our cafe survey uses the first, second and fourth.
- Overall rating: "How would you rate your experience today?" (1 to 5 stars).
- Recommendation: "How likely are you to recommend us to a friend?" (0 to 10). Bain's Net Promoter System counts 9 and 10 as promoters and 0 to 6 as detractors (Bain).
- The main thing: one rating about the thing you are deciding on, such as "How well did our opening hours suit you?".
- The reason: "What is one thing we should do better?" (open text, optional).
- The context: one question that lets you split the answers, such as "How often do you visit?".
20 survey questions you can copy
Grouped by survey type. Swap the business details for yours, and use only the ones that serve your goal from step 1.
Customer feedback
- How would you rate your visit today? (1 to 5)
- How likely are you to recommend us to a friend? (0 to 10)
- How easy was it to find what you needed? (very easy to very hard)
- How would you rate the value for money? (1 to 5)
- Did anything go wrong today? (yes / no)
- If yes, what happened? (open)
- How often do you visit? (first time / monthly / weekly / daily)
- What is one thing we should do better? (open)
Employee pulse
- I know what is expected of me at work. (Likert)
- I have the tools I need to do my job well. (Likert)
- My workload this month has been manageable. (Likert)
- I would recommend this team as a place to work. (0 to 10)
- What one change would make your week easier? (open)
- Which team are you in? (choice; drop it if teams are under five people)
Event feedback
- How would you rate the event overall? (1 to 5)
- Which session was most useful to you? (choice)
- Was the length about right? (too short / about right / too long)
- How likely are you to attend next year? (0 to 10)
- What should we change next time? (open)
- How did you hear about the event? (choice)
Step 4: Put Them in Order and Keep It Short
Order changes answers. People warm up on easy questions and give more honest answers to harder ones once they have started. Put the questions in this order:
- Easy and relevant first. The overall rating is a good opener: quick, and clearly about the thing they just did.
- General before specific. Ask the overall rating before asking about the coffee, or the coffee question colours the overall one.
- Sensitive and personal last. Age, income, team or role go at the end, and only if the goal needs them.
- One topic per screen. On a phone, a short page per topic is easier than one long scroll.
- The open question near the end, and optional. A required essay box is where people give up.
Keep it short, and test it. SurveyMonkey's data puts a 10-question survey at about 5 minutes, with abandonment rising past 7 to 8 minutes (SurveyMonkey). Before you send it, have two people answer it on their phones while you time them. If either takes longer than five minutes, or asks what a question means, cut or rewrite.
Step 5: Build the Survey Free in a Form Builder
A form builder is enough for almost every survey: it handles the question types, branching and sharing, and nobody needs an account to answer. Our comparison of the best form builders tests 12 of them on a phone. The Appy Pie AI online form builder describes building a form in three steps: describe it, customise it, share it. Here is the cafe survey through those three steps.
- Describe it, or pick the Feedback Form type. Either way, compare the draft with your list from step 3 and delete anything you did not ask for. If the draft includes name or email fields, remove them for an anonymous survey.
- Check the question types. The visit rating as stars, the recommendation question as a 0 to 10 scale, visit frequency as a single choice, and the improvement question as optional open text.
- Add one branch. If the visit rating is 1 to 3 stars, show a follow-up: "What went wrong today?". Everyone else skips it, which keeps the survey short for happy customers.
- Write the thank-you screen. One line saying what you will do with the answers, such as "Thanks. We read every answer and post what we change at the counter."
- Preview it on a phone. Answer it twice, once with a high score and once with a low one, to check the branch fires.
The same survey in Google Forms: use a rating question for the stars, a linear scale set from 0 to 10 for the recommendation question, and "Go to section based on answer" on the rating to send low scores to a follow-up section. Leave email collection off.
Step 6: Make It Anonymous
An anonymous survey is one where nobody, including you, can tell who gave which answer. That is a stronger promise than most surveys keep, so check each of these before you call yours anonymous:
- No name, email or phone fields. Not even optional ones; an optional email field turns some answers into identified answers.
- No login to respond. If people must sign in to answer, the tool knows who they are, even if your results sheet does not show it.
- No hidden identifiers. Personalised links, tracking parameters, or a link sent from an email tool that tags each recipient can all tie an answer to a person.
- No identifying combinations. Team, role, location and start date together can point to one person. In small groups, drop them or merge them into broad bands.
- Check what the tool records. Look at the export: if it has an IP address or device column, the survey is not anonymous to whoever holds the export.
In Google Forms, leave email collection off and do not turn on "Limit to 1 response": with that setting on, Google says users must sign in to their Google Account to fill out the form, even though their usernames are not recorded unless you collect emails (Google Forms Help).
Anonymous vs confidential
Anonymous
- You cannot link any answer to a person
- No name, email, login or identifying combination
- Best for staff surveys and sensitive topics
- You cannot follow up with anyone
Confidential
- You know who answered, but you do not share it
- Answers are reported only in groups
- Lets you follow up on a serious problem
- Needs a stated promise and someone trusted to keep it
Only say it if every point in the list above is true. If the tool records something you cannot switch off, call the survey confidential instead and say who will see the answers.
Step 7: Share It by Link, Embed or QR Code
Most people will open your survey on a phone, so share it in a way that works there. There are three ways, and many surveys use two of them.
Share link
Paste it into an email, a message, a social post or a receipt. The quickest way to reach people you already have contact with.
Embed on a page
Put the survey inside a web page, so people answer without leaving your site. Useful on an order confirmation page or a feedback page.
QR code
For people in the same room as a printed thing: a table card, a receipt, a poster, a product box, a slide at the end of a talk.
For the embed, the survey sits best on a page with one job. What a landing page is explains why, and the landing page builder is one way to make that page.
How to create a QR code for a survey
- Copy the survey's share link from the form builder's share step. Many form builders also generate the QR code for you there.
- Download the code as a PNG or SVG. SVG stays sharp at any print size.
- Print it with a short line that says what happens: "Scan to rate your visit, 1 minute". A bare code gets fewer scans.
- Size it for the distance. Make the code at least 2 cm (about 0.8 in) wide, and about one tenth of the scanning distance: a poster read from 1 metre needs a code about 10 cm wide (Scanova).
- Scan the printed code yourself with two different phones before you print a batch.
The same codes work inside apps too, for check-ins, menus and feedback; QR codes in mobile apps covers that side.
Step 8: Read the Results and Act on Them
Read the results against the goal from step 1, in this order:
- Response rate. Answers divided by the people who saw the survey. A low rate means the results may only reflect the keenest or the angriest.
- Where people stopped. If many started but did not finish, look at the question just before the drop; it is usually too long, too personal or required.
- The headline numbers. The average rating and the NPS: promoters (9 and 10) minus detractors (0 to 6), as a percentage of all answers.
- One cross-tab. Split the answers by your context question. This is where the decision usually shows up.
- The open answers. Read them all, and tag each with a theme. Count the themes; quote one or two.
Here is how the cafe might read its cross-tab. The figures are illustrative, to show the method, not results from a real survey:
| How often do you visit? | Answers | Average rating | Would like a 7am opening |
|---|---|---|---|
| Daily | 42 | 4.4 | 71% |
| Weekly | 88 | 4.2 | 38% |
| Monthly or less | 61 | 3.9 | 12% |
Overall, only about a third want an earlier opening. Split by visit frequency, 71% of daily customers do, and they are the ones who visit most. That is the decision from step 1, made by one table. Export the answers as CSV or send them to a spreadsheet so you can make the same split yourself, and then tell people what you changed: a sign at the counter saying "You asked, we open at 7am from Monday" is the best way to get answers to the next survey.
Free Survey Tools Compared
If you only need a survey, you do not need to pay for one. Here is what each free plan allows, read from each vendor's own pricing page on October 1, 2026. Limits change, so check the page before you commit to a tool for a long-running survey.
| Tool | Free plan, as the vendor states it | Branching on free | Worth knowing |
|---|---|---|---|
| Google Forms | Free with a Google account | Yes, go to a section based on an answer | The simplest choice for most surveys. Results go straight to Google Sheets. |
| Appy Pie AI online form builder | Unlimited forms and submissions on the free plan (product page, April 2026) | Yes, conditional logic | Generates the survey from a prompt. File uploads, payments and branding removal are on paid plans from $8 a month. |
| Tally | Unlimited forms and submissions within fair usage guidelines | Yes, conditional logic and calculations | Generous free plan; can close a form at a response limit or date. |
| SurveyHero | Unlimited surveys, questions and responses, 1 user, up to 30 days per survey | No, skip and display logic is on Business | Built for surveys rather than forms; the 30-day limit matters for always-on surveys. |
| SurveyMonkey | Unlimited surveys, 10 questions per survey, limited responses per survey | No, skip logic starts on Standard ($149 a month billed monthly) | The exact free response limit is shown only after sign-up. |
| Typeform | A limited number of responses each month; the page does not give the number | No, conditional logic is on Basic ($39 a month, or $28 billed yearly, 100 responses a month) | One question per screen, which suits short surveys. |
Sources: Tally, SurveyHero, SurveyMonkey and Typeform pricing pages, read October 1, 2026. Appy Pie AI figures are from the online form builder page, last updated April 21, 2026. Google Forms has no pricing page; it is part of a free Google account.
When You Need a Survey App Instead of a Form
A form is the right tool for most surveys. A survey app earns its place when the same people answer again and again, and you want to reach them on their phones without sending a new link each time:
- Recurring pulse surveys for staff, where the same questions run every month and you want the trend.
- Push reminders that bring people back to answer, rather than another email.
- NPS tracked over time for members or customers who already use your app.
- Feedback inside the experience, asked at the moment it happens rather than days later.
The Appy Pie AI survey app builder page lists push notifications for reminders, publishing to Android, iOS and as a progressive web app, and a results dashboard with exports. If you already run an app, our guide to app retention strategies shows where in-app feedback fits. If you only need answers once, stay with a form.
Frequently Asked Questions
What are the 5 questions to ask in a survey?
For a customer feedback survey: an overall rating, how likely they are to recommend you (0 to 10), a rating of the one thing you are deciding on, an optional open question asking what you should do better, and one context question such as how often they visit, so you can split the answers.
How do I create a basic survey?
Write down the decision the survey will inform, choose five to ten questions that serve it, put easy questions first, build it in a free form tool such as Google Forms or a form builder, test it on a phone, and share the link. Read the results against the decision you wrote down.
Can I use Google to create a survey?
Yes. Google Forms is free with a Google account, supports common question types and branching to a section based on an answer, and sends results to Google Sheets. For an anonymous survey, leave email collection off and do not turn on Limit to 1 response, which requires respondents to sign in to Google.
Is there a survey template in Word?
Word has templates for printed questionnaires, but a printed survey has to be typed up by hand afterwards. An online form collects the answers into a spreadsheet for you, works on a phone, and can be shared by link or QR code, so use Word only when people must answer on paper.
How many questions should a survey have?
As few as the goal allows, and usually ten or fewer. SurveyMonkey found a 10-question survey takes about five minutes, and abandonment rose on surveys taking more than seven to eight minutes.
What is the difference between a survey and a quiz?
A survey asks for opinions or facts and has no right answers. A quiz tests knowledge or sorts people into results, so it needs scoring. If you want scores or personality results, start with a quiz maker rather than a survey form.
How do I get more people to answer my survey?
Keep it under five minutes and say so at the top, ask right after the experience, make it work on a phone, explain what you will do with the answers, and tell people afterwards what you changed. A reminder to people who have not answered also helps.
Can respondents edit their answers after submitting?
It depends on the tool and its settings. Google Forms has an option to allow response editing, but it relies on a link or sign-in that ties the answer to a person. For an anonymous survey, assume answers cannot be edited, and keep the survey short enough that people get it right first time.
Your Questions Are Written. Build the Survey.
Describe the survey you need and the form builder drafts it from your description. Check the questions against your list, add your branch, test it on a phone, then share the link or QR code.
Create a Survey FreeOne Goal, Few Questions, One Table
Write the decision first, ask only what serves it, keep it under five minutes, keep the promise you make about anonymity, put the link where people already are, and read the results by one question that splits them. Then tell people what changed. If the survey feeds a sales page, our guide to creating a sales funnel shows where it fits.
Create a survey free →
Aasif Khan
Head of SEO and Growth Marketing Lead at Appy Pie AI with 17+ years in digital marketing and scalable growth strategies.
Abhinav Girdhar
Founder and CEO of Appy Pie AI, one of the largest no-code and AI app platforms, with 10M+ apps and websites created.