How to run a menopause survey people answer honestly
Most employers who have tried to measure this have already run a survey. The results came back thin, or reassuring, and nothing followed from them.
That is not a failure of question design. It is that answering honestly costs the person answering something, and answering blandly costs them nothing. This page is about how to change that arithmetic.
Why an internal survey under-reports
Disclosing a health issue to the organisation that decides your promotion is a career risk. It does not matter how the survey tool is configured. What matters is who the employee believes is on the other end of it, and in an internal survey the answer is the employer.
- The tool sits on the company’s own account, so people assume, often correctly, that somebody internal can see the raw data.
- Demographic questions make a small team identifiable, and people can do that arithmetic faster than the person who wrote the survey.
- It arrives from HR, which is the department the answer is about.
- Nobody has ever seen what happened to the last one, so there is no reason to believe this one is different.
What makes a survey safe to answer
| Design decision | Why it changes the answer |
|---|---|
| No login and no personal identifier | Nothing in the response can be traced back, including by us. No name, no email, no employee number, no IP address. |
| Collected by someone who is not the employer | The organisation reading the results is not the one that decides the promotion. |
| No commission from anyone being assessed | A finding against existing provision costs the collector nothing, so the finding can be reported. |
| Work impact, never symptom severity | Keeps it a workplace assessment rather than clinical data collection, and keeps special category health data out of it entirely. |
| A reporting floor | Nothing is shown for a group of fewer than 10 people, so a small team cannot be reverse-engineered. |
| Sent to everyone | A list picked by age or sex tells people a list exists with their name on it. |
The last one is the most commonly broken. Targeting the survey at women over 45 feels efficient and destroys the trust the exercise depends on.
The reporting floor, and why it has to be in the software
A promise not to report small groups is worth very little if the only thing enforcing it is a person’s judgement at the point of writing a report. It has to be in the system that produces the numbers.
Here, nothing is returned for a group below 10 people. Below the floor the employer is not told the count at all, only whether answers have started arriving, because an employer who can watch the number sit at eight knows exactly how many responses to add themselves. Health-related tallies above the floor are rounded to the nearest five. None of it is a setting that can be turned off.
Getting a response rate you can quote
- Have a senior sponsor send it, not HR and not a survey tool. Who it comes from does more for the response rate than anything else on this list.
- Send it to everyone. Say so in the invitation, so nobody wonders why they got it.
- Publish what the employer will and will not see, before people answer rather than after.
- Put the link on a poster with a QR code where people actually are: canteens, noticeboards, lockers, the back of toilet doors.
- Set a close date about two weeks out and send exactly one reminder in the middle.
- Say what happened afterwards. Nothing raises the second round like visible action after the first.
Do not ask managers to collect or chase answers. A response handed to a line manager is not anonymous whatever the covering note says, and one instance of it undoes the rest.
What to ask, and what never to ask
Seven questions is enough, and under two minutes is the target. Ask what is affecting work, which part of the organisation they are in, what would help most, whether they have already asked for it, what stopped them if not, which of your existing support they have used, and how supported they feel on a one to five scale.
Never ask for symptom severity, a diagnosis, medication, or a date of birth. Severity scales turn a workplace survey into health data collection about identifiable people, which changes your legal position, raises the bar on every question that follows it, and is the fastest way to lose the response rate you were trying to protect.