Short answer

A talk that is going to change something has to give four things: concrete signs to recognise, words to raise it with, a clear line for what a colleague should do, and where to send people next.

What can be cut without loss is the general statistics. They don't convince anyone who isn't already convinced.

And: book thirty minutes afterwards where the manager says what applies where you work. Without that, the talk is entertainment.

Why most talks change nothing

I've given enough of them to say this without blaming the person who booked me: the most common outcome of a talk on mental health is a good atmosphere and no change at all.

It rarely comes down to the speaker. It comes down to three things that are settled before they arrive.

The session is filler. It sits after lunch at a conference, between the financial update and an activity. Nobody has said why it's there, so the participants assume it's there because something has to be.

Nobody knows what they're meant to do with it. The talk says “talk to each other”, and then everyone goes back to a workplace where nobody has said whether it's all right to mention that someone seems worn out. So people don't, because it feels intrusive.

There is no afterwards. No follow-up, no manager saying what applies, nothing that happens in week two. Knowledge without permission becomes nothing.

So in most cases the weak point isn't the content. It's the frame around it, and that is something you can decide entirely for yourselves.

What it should contain

1. Signs you can recognise — in yourself and in others. Not diagnostic criteria, but what is actually visible at work: someone stops having lunch with the others, response times to emails become odd in both directions, someone who was meticulous starts making careless mistakes, irritability sits closer to the surface. These signs are seldom dramatic, which is exactly why they get missed. More on the early signs here.

2. Words to raise it with. The most common reason nobody says anything isn't indifference. It's that people don't know how to begin without it sounding like an accusation or an investigation. Concrete phrasings that participants hear, and ideally practise, are the single most useful part of a session like this.

3. A clear line. What a colleague should and shouldn't do. Anyone who doesn't get that line either does nothing at all out of fear of getting it wrong, or takes on a responsibility that isn't theirs and burns out as a result. Both outcomes are common, and both can be prevented with ten minutes of the talk.

4. Where to send people next. Occupational health, 1177 (Sweden's national healthcare advice service), the local health centre, and what applies in an acute worry. Specific, with numbers, and ideally with your own routine written in.

And something that makes it real. I have lived with panic disorder and social phobia myself, and that part of my talks is the one people remember. Not because my story is remarkable, but because a room listens differently when the person standing there hasn't just read about it. What you should demand isn't specifically my story — it's that the person at the front has something that makes the subject concrete.

Managers and employees — the same session?

The foundation should be shared. The point of everyone hearing the same thing at the same time is precisely that everyone gets the same words, and that nobody can say afterwards that they didn't know what applied.

But managers need something more, and it should sit separately. Their questions are different: what responsibility for the work environment actually involves, how to open a check-in conversation without it turning into a development review, what may and may not be documented, how to handle an adjustment without it being unfair to the rest of the group, and where the line falls towards care and occupational health.

If you mix that into the shared session, it becomes unclear for both groups — too abstract for the managers and too administrative for everyone else.

A common and sensible order: 60 minutes for everyone, and then 60–90 minutes for managers and safety representatives only, on the same day.

The hour afterwards matters most

This is the advice I give most often and the one clients most often skip, so I'll say it plainly: book thirty minutes directly after the talk where someone in management says what applies where you work.

Not a speech. Three or four sentences will do: that it's all right to raise this, who you go to, what happens if you do, and what won't happen — that it won't end up in the pay review.

The reason is simple. A speaker can give knowledge, but only you can give permission. Without permission, the knowledge becomes a pleasant hour that nobody dares to use, and that is exactly where most initiatives run out of steam.

How you know whether it made a difference

Not through the evaluation in the room. That measures whether people liked the hour, which is nice to know and almost entirely uncorrelated with whether anything changes.

Measure three weeks later instead, with two questions in an email. Have you raised something with a colleague that you otherwise wouldn't have? And: do you know where to turn if you, or someone in your team, isn't doing well?

The second question is the important one. If the share answering yes to it hasn't gone up, the talk hasn't taken hold, however good the evaluation was. And then you know after three weeks rather than after a year.

When not to book a talk

When the problem is workload. If the cause of ill health is too much to do, too few people or an organisation in the middle of change, a talk on mental health is at best ineffective and at worst provocative. It signals that the problem lies with the individual. People notice that in the room, and it's a talk neither you nor I want to give.

When something has already blown up and nobody has said anything about it. If you've had a conflict, a dismissal or a period of sick leave that everyone knows about but nobody talks about, that has to be dealt with first. Otherwise everything I say will be about it, silently, in the whole room.

When you're booking so that you can tick a box. That's a perfectly understandable reason and I won't moralise about it. But say so to the speaker, and we can plan it honestly — a shorter, more concrete session with no ambition to change the culture is a better use of your money than one that pretends otherwise.

The most important thing

If you take only one thing away: a talk can give knowledge and a shared language. It cannot give permission. Only you can do that.

So decide two things before you book. What should be different in three weeks, and who will tell the staff that it's all right. If you have answers to those questions, it matters almost less who you book. If you don't, it doesn't matter at all.

I've written separately about what a talk costs and how booking works: Booking a speaker on mental health.

Warmly, Mikael GlännströmWellbeing coach, author and father of two