How to design polls your group will actually respond to
A group poll gets answered when it asks one question, offers three or four options, opens when members have a free minute, and closes before you need the answer. Everything that drives response rate down is a design choice, not a member problem.
TL;DR
Design a group poll around four decisions: ask exactly one question, keep it to three or four options (two for a yes/no), set a fixed open and close window rather than leaving it running forever, and keep results private to the manager unless the poll is a genuine shared decision. Survey-methodology research on respondent "satisficing" shows that the longer and more cluttered a question, the more people pick the first plausible answer or skip it entirely. A poll is for one decision or one signal; the moment you stack a second question onto it, you want a check-in instead.
Why do most group polls go unanswered?
Because they ask too much. The most common failure is not apathy; it is a poll that is really three questions wearing one coat: "Can you make Saturday, and if not when suits you, and shall we move it to the hall?" Members open it, see work, and close it.
The underlying mechanism has a name. Stanford's Jon Krosnick has spent three decades documenting survey "satisficing" — the well-evidenced tendency for respondents to give the first answer that is good enough rather than the most accurate one, and to do it more the harder you make them work. Every extra option, every clause in the question, every scroll, pushes a few more members from a considered answer to a lazy one or to no answer at all.
For a coaching group this matters twice over. A skipped poll is missing data, and missing data on a Friday is a roster you do not have. The fix is not a reminder nudge; it is a poll that was designed to be answered in one tap.
How many options should a poll have?
Three or four for a decision. Two for a yes/no signal. Above five, response rate falls because the member is now reading a list and comparing entries instead of tapping the obvious answer.
When you genuinely have six candidate session times, do not put six options in one poll. Run two rounds: a coarse poll that narrows the field to the two most popular, then a clean head-to-head between the survivors. Two short polls outperform one long one almost every time, because each one is answerable in the three seconds a member will actually give it.
One more rule on the options themselves: make them mutually exclusive and exhaustive. "Yes / No / Maybe" leaves the manager guessing what "Maybe" means. "Yes / No / Not sure yet, I'll confirm by Thursday" tells you what to do with the answer.
Does the order of the options matter?
Yes, more than people expect, and this is where Krosnick's research earns its place in a coaching toolkit. In self-administered polls that members read on a screen — which is every poll in a coaching app — there is a measurable primacy effect: the options near the top get chosen slightly more often, simply because tired or rushed readers stop at the first acceptable one. (In polls read aloud, the bias flips to the last option heard, a recency effect, but that is not the case you are designing for.)
The practical takeaway is opinionated and we will defend it: do not list options in the order that flatters the answer you want. If you put "Yes, attending" first because you are hoping for a full room, you are baking a small upward bias into your own attendance data and then trusting it. Order options neutrally — alphabetical, chronological for times and dates, or worst-to-best for a rating — and let the answer be the answer.
When should a poll open and close?
Open it when members have their phone in hand and a minute to spare. For most adult groups that is early evening, not the morning. A poll fired at 08:00 lands in the same commute scroll where it gets buried; the same poll fired at 18:00, when the working day is done and the member is on the sofa, gets answered before the kettle boils.
Close it before you need the answer, not at the moment you need it. If your roster has to be set on Saturday morning, a Friday-midnight close gives you no slack to chase the two people who did not respond. A Monday 18:00 open and a Friday 12:00 close hands a weekend group its roster a day and a half early, with the whole of Friday afternoon to follow up by name.
Inside Bitir a poll can be scheduled to open and close automatically on a recurring weekly cycle, so the manager sets the rhythm once and never has to remember to post it. The auto-close is the part that earns its keep: a poll that never closes is a poll members feel no urgency to answer.
Should poll results be public or private?
Default to private. A poll about how members feel, whether they are struggling, or whether they will turn up should go to the manager alone. The manager can then act on the pattern without exposing any individual.
Make results public only when the poll is a shared decision the group is entitled to see — picking a session time, choosing the next topic, voting on a social. Those are the cases where seeing the running tally is the point, and where a member changing their answer to join the forming majority is a feature, not a contamination.
The trap is making everything public because it feels more democratic. Public-by-default polls quietly pressure members toward the visible answer, which is exactly what you do not want when the poll is meant to surface honest signal. Bitir keeps poll responses private to the manager unless you explicitly choose to show the tally, which is the right default for the same reason member posts are private to the manager by default: honesty survives where visibility is opt-in.
What does a well-designed poll look like in practice?
Take a worked example. Priya Nair runs a 9-person leadership cohort in Bristol — a 10-week programme for newly promoted managers at a mid-sized NHS trust. She runs two standing polls. The first is a Monday 18:00 attendance poll for Thursday's live session: one question, three options ("In the room" / "Joining online" / "Can't make it this week"), closing Wednesday noon, results private so nobody feels watched for missing a week.
The second is a topic poll she opens in week four to choose the back-half curriculum: four options, chronological order, results public because the cohort should see the choice being made together. Her attendance poll runs at a 94% response rate; the topic poll hit 100% because, she says, "people answer when their vote actually changes something."
The contrast is the whole lesson. Same group, same app, two different visibility settings — chosen on purpose, because the polls do different jobs. A coach who makes both public, or both private, gets one of them wrong.
If your "poll" keeps wanting to ask three questions and collect a reflection, you have outgrown the format. That is a weekly check-in, and it has its own design rules — see 30 weekly check-in questions coaching groups actually answer for the three-question structure that replaces a bloated poll.
Questions we're asked about group polls
How do I use polls in a coaching group?
Ask one question per poll, keep it to three or four options, give it a fixed open and close window (Monday evening to Friday lunchtime works for most weekly groups), and keep results private to the manager unless the answer is genuinely a shared decision. A poll is for a single decision or a single signal, not a survey. If you find yourself stacking questions, you want a check-in, not a poll.
How many options should a group poll have?
Three or four for a decision, two for a yes/no signal. Beyond five options, response rate falls because the member has to read and compare a list instead of tapping the obvious answer. If you need six dates or six session times, run two rounds: a coarse poll to narrow the field, then a second poll on the two survivors.
Should group poll results be public or private?
Default to private. A poll about how members feel, whether they will attend, or what they are struggling with should go to the manager only. Make results public only when the poll is a shared decision the group is entitled to see, such as picking a session time. Public-by-default polls quietly pressure members toward the visible majority.
When should a poll open and close?
Open it when your members are most likely to have their phone in hand and a free minute, which for most adult groups is early evening, not first thing in the morning. Close it before you need the answer, not at the moment you need it. A Monday 18:00 open and a Friday 12:00 close gives a weekend-match group the roster a day and a half early.
Does the order of poll options change the result?
It can. In polls members read on a screen there is a small primacy effect, where options near the top are chosen slightly more often by rushed readers. Order options neutrally — alphabetical, chronological, or worst-to-best — rather than putting the answer you are hoping for first, so you are not biasing your own data.
Run your group polls inside Bitir
One question, scheduled to open and close on a weekly cycle, private to the manager by default — with a public tally when the decision genuinely belongs to the group.
Start Your Group