A suggestion box catches the loud voices and misses the quiet ones. I would rather ask every member the same short questions and compare the answers. One link in a newsletter, or on a notice by the desk, is enough.
Use this when
You want to know why attendance is flat, or which fix matters most to members.
Skip it when
You need to know why a person left. That is a different question, and an exit survey fits better. Also skip it if you cannot act on the answers.
What is asked and why
How often and how long set the context. Areas used and ratings for equipment, cleanliness, staff and classes show where members feel the gym. Busy times point to crowding. The wish list lets members vote on changes. One recommend score tracks the mood over time. One open question names the single change that would bring people in more often.
What a member's AI would ask
"How often do you come?" Then "When is the gym too busy?" Then "What one change would make you come more often?" It shows all answers and sends only after a yes. See agent-fillable forms.
Run the survey again after you make changes. Browse forms for gyms. A gym membership application welcomes new members, and a workout log keeps current ones engaged.
Questions
How do I find out which area to fix first?
Download the CSV and average equipment_rating, cleanliness_rating and staff_rating. Count the options in wanted_changes. The lowest score with the most votes is a good place to start.
Can I send it only to members who rarely visit?
Yes. Share the link in a message to that group, then filter visit_frequency for Rarely or A few times a month. Their answers tell you why they stay away.
How do I stop answers after the survey week?
Set close_at to the last day you want replies. Set max_responses if you only need a sample from the members who answer first.