Related guide: Community engagement planning: from the decision to a working plan
Good community survey questions start from the decision you are making and only ask about what the community can genuinely influence. Each question should be neutral, specific and about one thing, and the survey should mix closed questions (ratings, rankings, multiple choice) that you can count with open questions that let people tell you what you did not think to ask.
Below you'll find example questions for six common types of decision, leading questions rewritten as neutral ones, guidance on demographic questions, and an honest answer to how many survey responses you need. All the examples are generic, so adapt them to your project and community.
Start from the decision, not the question list
The most common reason a consultation survey disappoints is that it was written before anyone agreed what the survey was for. Before drafting a single question, write down three things:
The decision. What will be decided, by whom and when? "We will adopt a park concept design in March" is a decision; "hear what people think about the park" is not.
What is open to influence. Which parts of the decision can community input genuinely change? Which are fixed by budget, legislation, safety standards, technical constraints or earlier decisions?
What you need to learn. For each part that is open, what does the decision maker need to know? Priorities, preferences between options, concerns, local knowledge, how people use the place today.
Every question in the survey should trace back to something on that third list. If you cannot say how an answer will be used, cut the question. Cut questions about things already decided too: asking people to rate what they cannot change creates expectations you cannot meet. Explain the fixed parts in the survey introduction instead.
A survey is also only one method: good at reaching a broad group and counting preferences, weak at exploring trade-offs or reaching people who don't do online forms. Your community engagement plan should say what the survey is for and what other methods sit alongside it.
How do you create a community survey?
Once the purpose is clear:
Write a short introduction. Say what decision the survey informs, what is and is not open to change, how long it takes, and how and when results will be reported.
Draft questions against your "what we need to learn" list. One or two questions per item is usually enough.
Order the questions. Start with easy questions about the topic, put the most important ones early, and demographics at the end.
Cut it down and test it. Remove anything you would not miss, then ask a few community members to complete it and tell you what confused them. Check it on a phone and with a screen reader.
Plan the other ways in. Paper copies, phone support, translations and assisted completion widen who can respond.
Decide how you will analyze it before it opens. It will change how you word the questions.
Community survey question examples
The examples below are grouped by the kind of decision the survey informs. Each set mixes closed and open questions and ends with an "anything else" question, which you should always include: it catches issues you did not anticipate. Replace the bracketed text with your project details.
Park or open space upgrade
How often do you currently visit [park name]? (Most days / Weekly / Monthly / A few times a year / Never)
What do you mainly use the park for? Select all that apply. (Walking / Play with children / Dog walking / Sport / Relaxing / Events / Passing through / Other, please specify)
Please rank the following possible improvements from most to least important to you: shade and seating, play equipment, lighting, paths, toilets, planting, dog off-leash area.
The draft concept shows [feature]. How well would this meet your needs? (Very well / Quite well / Not very well / Not at all / Not sure)
Is there anything else you would like us to know about the park or the upgrade?
Transport or street changes
How do you usually travel along [street name]? Select all that apply. (Walk / Bike or scooter / Bus / Car / Delivery or work vehicle / I don't use this street)
How safe do you feel using [street name] at the moment? (Very safe / Fairly safe / Not very safe / Not at all safe / Not sure)
We are considering two layouts for the street. Which would you prefer? (Option A / Option B / No preference / Neither, please explain)
During construction, what would help reduce the impact on you? Select all that apply. (Advance notice of work dates / Access maintained to my property / Work outside business hours / A single contact person / Other, please specify)
Is there anything else we should consider about the street changes?
Place plan or town center
What do you value most about [town center] and would not want to lose?
How satisfied are you with the following in [town center]? (Very satisfied to very dissatisfied, for each of: places to sit and meet, shops and services, walking routes, parking, events and activities, cleanliness, feeling of safety at night)
Which of these directions for the next ten years best matches what you would like to see? (Direction 1 / Direction 2 / Direction 3 / A mix, please explain)
If you could change one thing about [town center], what would it be?
Is there anything else you would like to tell us about the future of [town center]?
Housing or growth strategy
Which of the following housing issues matter most to you or your household? Choose up to three. (Cost to rent / Cost to buy / Choice of housing types / Housing for older people / Housing for young people or first homes / Quality of new development / Other, please specify)
The strategy proposes focusing new housing in [areas]. How much do you support or oppose this approach? (Strongly support / Support / Neutral / Oppose / Strongly oppose / Not sure)
What is the main reason for your answer to the previous question?
If more homes are built in your neighborhood, which local services or infrastructure would most need to improve?
Is there anything else you would like to say about housing and growth?
Service change
Have you used [service] in the last twelve months? (Yes / No / Not sure)
How would the proposed change to [service] affect you or someone you care for? (It would make things better / No real difference / It would make things worse / Not sure)
Please tell us more about how the change would affect you.
If the change goes ahead, what support would help you adjust? Select all that apply. (Clear information in advance / Help to use the new arrangement / Transport or access support / Information in another language / Other, please specify)
Is there anything else we should know before a decision is made?
Community priorities or strategic plan
Please rank these draft priority areas from most to least important for the next four years: [list five to seven areas].
How well does the draft vision statement reflect the kind of place you want to live in? (Very well / Quite well / Not very well / Not at all / Not sure)
If funding is limited, which of these should we protect first? Choose up to two.
What is one thing we could do in the next year that would make a real difference in your area?
Is there anything else you would like to tell us about the draft plan?
Notice the pattern: closed questions have complete answer options including "not sure" or "other", rating scales are balanced, option questions allow "neither", and open questions ask why, which is often what decision makers find most useful.
Leading questions examples: before and after
A leading question nudges people toward a particular answer; other biased questions bundle two issues, assume facts or offer a false choice. Bias is rarely deliberate, but it makes results less credible, because anyone who disagrees can point to the wording. Common patterns, with neutral rewrites:
Leading or biased question | Why it's a problem | Neutral version |
|---|---|---|
Don't you agree that the new bike lane will make the street safer? | Leading: the wording signals the expected answer and makes disagreeing feel awkward. | What effect, if any, do you think the proposed bike lane would have on safety on the street? |
How much do you support the exciting new plan for the town center? | Loaded: "exciting" carries a positive judgment and only support is measured. | How much do you support or oppose the draft plan for the town center? |
How satisfied are you with the cleanliness and safety of the park? | Double-barreled: someone could feel the park is clean but unsafe and has no way to say so. | Two questions: How satisfied are you with the cleanliness of the park? How satisfied are you with how safe you feel in the park? |
How often do you use the library's new online booking system? | Assumptive: it assumes people know about and use the system. | Have you used the library's online booking system? If yes, how often? |
Do you support the proposed modal filter and active transport corridor treatment? | Jargon: many people will not know what these terms mean, so answers are guesses. | Do you support or oppose closing [street] to through traffic, while keeping access for people walking, cycling and living on the street? (Include a short plain description or diagram.) |
Should we upgrade the pool or let it close? | False choice: there may be other realistic options, and the framing pushes toward one answer. | Which of these options for the pool would you prefer? (Upgrade / Repair to keep it open / Replace with a different facility / Close / Other, please specify) |
Agree or disagree: the draft plan reflects community priorities. | Agree/disagree acquiescence: people tend to agree with statements, especially from an authority. | How well, if at all, does the draft plan reflect your priorities for the area? (Very well / Quite well / Not very well / Not at all / Not sure) |
Given the rising cost of maintenance, should the council sell the old hall? | Leading preamble: the framing gives only one side of the argument before asking. | Provide balanced information about the options separately, then ask: What should happen to the old hall? (Options listed, plus Other) |
Rate the proposal: Excellent / Very good / Good / Fair | Unbalanced scale: three of the four options are positive, so results look more favorable than opinion is. | Rate the proposal: Very good / Good / Neither good nor poor / Poor / Very poor |
A useful test: if someone who disagrees with the proposal would object to the wording, rewrite it.
How to write survey questions that people can answer
Beyond avoiding bias, good questions are easy to understand and easy to answer honestly.
Use plain language. Write the way you would explain the project to a neighbor, and replace technical terms with everyday words.
Ask about one idea per question. If a question contains "and" or "or", check whether it is really two questions.
Be specific. "How do you feel about the park?" invites vague answers. "What would make you visit the park more often?" gives you something to act on.
Give complete, non-overlapping answer options. Include "other", "not sure" or "prefer not to say" where relevant, and balance scales with equal positive and negative options.
Keep open questions focused. "What is the main reason for your answer?" is easier to answer and analyze than "Any comments?"
Respect people's time. State how long the survey takes and keep that promise. A focused survey people finish beats a long one they abandon.
Likert scale basics
A Likert scale asks people to rate their level of agreement, support, satisfaction or importance on an ordered scale, usually five points (for example, strongly support, support, neutral, oppose, strongly oppose). A few practical tips:
Label every point in words, not just the ends. Numbers alone are interpreted differently by different people.
Prefer support/oppose or satisfaction scales over agree/disagree statements, which invite agreement for its own sake.
Report the full spread of answers, not just an average. An average of "neutral" can hide a community that is sharply split.
Personal information, accessibility and translation
Only ask for personal information you will use. If location matters, ask for a neighborhood or postal code rather than a full address, and say how information will be stored and used, in line with your organization's privacy policy.
Make sure the survey works with screen readers, keyboard navigation, large text and small phone screens, and offer paper, phone or in-person options. Where your community speaks other languages, translate the survey or offer assisted completion, and check translations with community members rather than relying on machine translation alone.
Should a community survey ask demographic questions?
Yes, but only what you will use, and always as optional. Their main job in a consultation survey is to check who you heard from against who is affected. If you won't use an answer for that, or to compare views between groups, don't ask.
Ask what matters for this decision. For a playground, children's ages may matter; for a bus route change, disability or car access may matter.
Make every demographic question optional, include "prefer not to say", put them at the end and explain why you are asking.
Use respectful, inclusive wording consistent with your organization's standards, and let people self-describe where categories might not fit (for example, gender).
Use broad ranges for age and income rather than exact figures.
Ask about connection to the area, such as live, work, study, own a business, visit, which often matters more for consultation than personal characteristics.
Then use the answers. Compare respondents with the groups in your stakeholder analysis. If a group is missing or thin, adjust outreach while the survey is open, and say in your report who you did not hear from.
How many survey responses do I need?
The honest answer: for most consultation surveys, there is no magic number. Sample size calculators assume a random sample, where everyone had a known chance of being selected. A consultation survey is not one. People choose to respond, usually because they care, are directly affected or happened to hear about it. So margin-of-error formulas don't apply the way they would to a properly designed poll, and a large number of responses does not make results representative on its own.
Instead of chasing a target number, ask two better questions:
Did we hear from the groups affected by the decision? Check responses against the stakeholder groups in your plan. A smaller set of responses that covers every affected group can be more useful than a large one from a single part of the community.
Can we explain who we did not hear from, and why? A credible report is open about gaps and describes what was done to address them.
If the decision genuinely needs results that represent the whole population (for example, a measure of community satisfaction you want to track over time), use a survey with a random sample designed with a research provider. The two do different jobs: the random sample estimates what the population thinks, and the open survey lets everyone who wants a say contribute ideas, concerns and local knowledge.
Finally, response counts are not votes. Consultation informs a decision; it does not decide it by headcount. Majority opposition among respondents matters, but so do the reasons people give, the views of groups less likely to fill in surveys, and the policy context. Say this in the survey introduction, and report results in a way that does not imply a referendum.
Survey response rate: how to get more people to respond
For an open consultation survey, a "response rate" is hard to calculate because there is no fixed list of people invited. Track instead which groups responded and how that changed as you added outreach. For a random-sample survey, response rate matters and your research provider should report it.
What helps people respond:
Make the relevance obvious: lead with what is changing and how it affects people, not with the name of the strategy.
Go to where people already are: letters to affected households, pop-ups at the site, schools, libraries, community groups and local businesses, as well as online channels.
Work with trusted community organizations to reach groups who are less likely to respond to official channels.
Show that previous feedback made a difference.
Analyzing the responses
For closed questions, report how many people answered each one, the full spread of answers and differences between groups. For open questions, code responses into themes, record how often each came up and from which groups, and choose quotes that represent each theme fairly. Our guide to analyzing consultation feedback walks through the process in detail.
Then close the loop: tell people what you heard, what was decided and how their input shaped it. A clear community engagement report is what makes the next survey worth answering.
If you run surveys across many projects, CE Canvas keeps each project's decision, stakeholder groups and methods in one record, so survey questions trace back to what is open to influence and reports show who you heard from against who you set out to reach.
Frequently asked questions
How do you write good community survey questions?
Start from the decision and ask only about what the community can genuinely influence. Write each question in plain language, about one idea, with neutral wording and balanced answer options. Mix closed questions you can count with open questions that ask why, and always finish with an "anything else" question. Test the survey with a few community members before it opens.
What is a leading question?
A leading question is worded in a way that nudges people toward a particular answer, such as "Don't you agree the new bike lane will make the street safer?" A neutral version asks what effect, if any, people think the change would have. Leading questions make results less credible because anyone who disagrees can fairly point to the wording.
How many responses do I need for a community survey?
For an open consultation survey there is no fixed number, because respondents choose to take part and sample size formulas assume a random sample. What matters is whether you heard from the groups affected by the decision and whether you can explain who you did not hear from. If you need statistically representative results, use a random-sample survey designed with a research provider.
Should a community survey ask for demographics?
Ask only the demographic questions you will use to check who you heard from or to understand how views differ between groups. Make every demographic question optional, include "prefer not to say", use respectful wording and broad ranges, and put them at the end with a short explanation of why you are asking.
How long should a community survey be?
As short as it can be while still answering what the decision maker needs to know. Every question should trace back to a part of the decision that is open to influence. Tell people how long the survey takes before they start, test it to check the time is accurate, and cut any question whose answer you would not use.