A board sends a survey, gets nineteen responses out of two hundred and ten homes, and concludes the community does not care. Usually the community cared and the survey was unanswerable.
Surveys fail on design far more often than on apathy. They are too long, they ask about things the board cannot change, they arrive with no deadline, and they disappear afterward without anyone hearing what the answers were. Each of those is fixable, and fixing them is cheaper than another round of guessing what owners want.
Ask about decisions you are actually going to make
The first filter is the hardest one. For every question, ask what your board would do differently depending on the answer. If there is no answer that changes anything, cut the question.
This sounds obvious and it eliminates about half of the typical draft. "How satisfied are you with the community?" changes nothing. "The board is considering keeping the pool open through October at an added cost of roughly $40 per home for the year. Would you support that?" changes something.
Two things follow from this. Surveys are best run before a decision, not after, which means the timing is set by your budget or project calendar rather than by when somebody remembers. And a survey is input, not a vote. Your directors hold a fiduciary duty to the association, and that duty does not transfer to whoever filled in the form. Say so in the introduction, in one sentence, so nobody reads a 60 percent preference as a binding instruction.
Write questions people can finish
Photo by Tima Miroshnichenko on Pexels
Practical rules that hold up across community sizes:
Eight questions or fewer. Ten if one of them is optional and open-ended. A survey that takes more than four minutes gets abandoned in the middle, and a half-finished response tells you nothing.
One idea per question. "Are the landscaping and the pool maintained well?" produces answers you cannot read, because half your respondents mean one and half mean the other.
Give the cost with the choice. Residents will say yes to a dog park and no to a $12 monthly increase for a dog park. Only the second answer is useful. Pairing the option with its number is also the single best way to reduce the number of people who show up angry later.
Include a comment box, and read it. The structured questions tell you the distribution. The comment box tells you the thing you did not know to ask about.
State the deadline and the length up front. "Six questions, about three minutes, closes Friday the 28th" outperforms a bare link every time.
Get it in front of everyone
Response rate is mostly a distribution problem. Send it through the channel with the best coverage, not the most convenient one, and plan to run it on paper as well as electronically unless nearly every household has given you an email address. Boards that only have email addresses for two thirds of their households cannot have a representative survey electronically, no matter how good the questions are.
Announce it once when it opens and once about three days before it closes. The second reminder is where much of the final count comes from. If you have an annual meeting coming up, running the survey two or three weeks before it gives you something concrete to discuss and tends to help attendance; our annual meeting and quorum guide covers the rest of that preparation.
Read the results without fooling yourself
Photo by Lukas Blazek on Pexels
Report the raw counts, not just percentages, and report the response rate alongside them. "62 percent support" reads very differently when it means 31 of 50 responses out of 210 homes, and residents who work with numbers will do that arithmetic whether or not you show it.
Watch for the two predictable skews. People with a grievance answer at a higher rate than people who are content, so an unhappy result may be an unhappy subset. And the households you never reached are usually the same households missing from every list, which means your quiet third stays quiet. Research from the Foundation for Community Association Research is worth reading for national context on how association residents actually see their boards, but it is not a substitute for knowing your own response gap.
Close the loop
Publish what you found and what you decided, including the parts where you went against the majority and why. This is the step boards skip, and skipping it is what kills the next survey.
The mechanics are not hard to run in-house. HOA-OS Community plans include a form builder with 14 field types, so a survey with cost-paired choices and a comment box is a form rather than a spreadsheet exercise, and the same plans carry community voting with configurable quorum and an anonymous ballot option for the times you need a vote rather than a poll. Communities start with a 30-day free trial and no credit card. See how it works.
A survey you do not report back on is a survey you only get to run once.