Public services
Public consultation survey
Put the proposal in front of people, then ask what they support, what worries them, and who they are answering as.
Why this survey gets sent
“We published the proposal for six weeks and received 11 responses, 9 of them from the same street. Then the decision was taken, and everybody was surprised by it.”
A consultation fails in one of two ways: nobody answers, or everybody answers about something else. Both are wording problems before they are outreach problems — a proposal explained in three hundred words gets read, and questions that separate support from concern produce answers you can put in front of a committee.
What is in it
- The proposal itself, in a text block at the top
- Who you are answering as (single response)
- How much you support it, on a five-point scale
- Which parts you agree with (multiple choice)
- What concerns you, and why (long answer)
- The one change that would improve it (long answer)
- May we publish what you wrote? (yes / no)
- Email, only if you want the published summary
About 3 min to answer
Public consultation survey
About 3 min
How to write this survey
Three hundred words at the top, a link for the rest
The text block at the top is the survey. If it is only a link to a forty-page document, the people who answer are the people with an interest strong enough to read forty pages — exactly the sample that makes a consultation look unanimous and then leaves everyone surprised. Write three hundred words: what is proposed, why now, who it affects, what happens next. Keep the full document one click away for whoever wants it.
Write it without the words that appear only in internal circulars. A proposal described as “optimising the utilisation of the frontage” is answered by nobody; the same proposal described as “removing 12 parking spaces to widen the pavement” is answered by everybody, including the people who will object — and their objection is the thing you were asking for.
Separate support from concern
Support and concern are not opposite ends of one question. A person can be strongly in favour of a change and worried about how it will be carried out, and a five-point scale on its own records that person as lukewarm. Ask for the position on the scale, then ask separately what worries them; the pair is what lets you report that most people want this, and most of them want it done differently.
The multiple-choice question in the middle does the work of a hundred open answers. Break the proposal into the parts people actually argue about — the reason, the timing, the cost, the way it would be carried out, its effect on the nearest neighbours — and let a respondent agree with three of them and not the fourth. You learn where the disagreement is concentrated from a single chart, and the open box stops having to carry everything.
Ask who is answering, and read the results by group
One question about who is answering changes what the results can be used for. Residents, business owners, associations and people who merely pass through want different things from the same street, and a total that mixes them is a number a committee cannot act on. Keep the categories few and obvious, and add “someone else” rather than trying to name every case.
Then read every question split by that answer before you read any total. It is entirely normal for a proposal to be supported by three groups and opposed by one, and it is usually the opposing group that lives closest to it. That split is the finding; the average of the four is a way of not reporting it.
Say what you will publish before they write
Decide before you publish what you will do with the responses, and say it on the page. Two sentences — “a summary of what people said will be published alongside the decision” and “we may quote responses without contact details, and only where you allow it” — raise the quality of what is written far more than any reminder does, because somebody writing for a summary writes an argument rather than a shout.
Ask for permission to publish as its own yes-or-no question, and keep the email separate, optional, and attached to a stated purpose: to send the summary once it is ready. An address requested without a reason is an address people decline to give, and the refusal costs you the whole response rather than only the address.
Reach the people who will not come looking
A consultation that lives only on a website is answered by the people who visit that website. A printed QR code where the proposal physically applies — on the street itself, at the counter, in the entrance of the building — reaches the person who is affected daily and has never once opened the announcements page.
Give it a closing date, put that date in the first paragraph, and leave it open long enough to survive one holiday and one weekend. Four weeks is a reasonable floor; a single week is an announcement dressed as a question, and everyone who finds it too late will tell you so.