Announcing the survey

A survey that arrives with no warning gets poor participation. And participation is not a vanity metric: it is what decides how many results you will be able to publish. Every branch that does not gather enough responses gets hidden by the confidentiality threshold — and there you lost the area, not the survey.

This page is material to use, not theory.

The four things to say

1. Why it is happening. Not «because it is that time of year». If there is a concrete reason —internal communication came out badly last year and something was done about it— say it. People answer differently when they believe someone will read.

2. That it is anonymous, and why they can believe it. The most neglected part, and the one that changes answers most. The word «anonymous» is not enough: you have to explain the mechanism. See What to tell them about anonymity, which is written to be forwarded.

3. How long it takes and the deadline. A concrete date. «Over the coming weeks» mobilizes nobody.

4. What happens afterwards. The cheapest and most effective commitment: say when results will be shared. And keep it — next year's participation depends almost entirely on whether anything happened this time.

What not to do

Do not send it by email alone if half your people do not use email. If you have operations on a floor, a shop or a plant, the kiosk exists for that. A survey that only reaches desk workers measures desk workers.

Do not promise changes you will not be able to make. It is worse than not asking: next time, people already know.

Do not communicate it from HR only. When the announcement comes down through direct management, participation goes up. When it arrives as an institutional email alone, it reads as paperwork.

Do not let managers push too hard. A manager standing next to the kiosk is the fastest way to ruin perceived anonymity, however impeccable the system is.

A template

To adapt, not to copy verbatim:

We are measuring workplace climate, and we need you to take part.

From [date] to [date] you can answer a survey about what it is like to work here. It takes around [X] minutes.

It is genuinely anonymous. The system records that you responded —so we know how many people took part— but it keeps no link between you and your answers. Nobody can see what you wrote: not your manager, not HR, not the leadership team. On top of that, results for small groups are not shown, precisely so nobody can be identified by elimination.

What it is for. Results are read per area, so what comes out of this can be worked on where it belongs. We will share the results on [date] and tell you what was decided.

How to respond. [You will get a link by email · Tablets are available at [place] during [hours]].

If you have questions about the survey, write to [person].

While it is running

Watch participation per branch, not the overall figure. The global number can look fine while an entire area sits at zero — and that is exactly the one you are about to lose.

If a branch is lagging, a reminder helps. What does not help is that reminder coming from its direct manager, person by person.