LinkedIn posts about burnout and mental health: what 468 show

Posts about burnout and mental health do work on LinkedIn, and the data on how they work is unusual. Across 468 mental health posts from 40 creators, the strongest ones use native video 37.0% of the time against 10.3% for the weakest, almost never share an external article (6.5% against 38.9%), and, uniquely among the topics we studied, avoid questions in the hook. The harder question is not how to write one. It is whether to publish it at all, which depends on something no engagement metric can tell you.

Is it okay to post about mental health on LinkedIn?

Yes, with a caveat that has nothing to do with propriety. The audience is receptive: the median post in this topic drew 135 reactions and 17 comments, and 17 is the highest comment median of any topic we broke out, against a cohort median of 12. Readers do not scroll past this material. Many of them reply at length, and a good proportion of the replies in the posts we looked at are people describing their own situation.

The caveat is that LinkedIn is a professional network where your name is attached to everything permanently, and the same post can read as candour to one reader and as instability to another. That is not fair and we are not going to pretend otherwise. It is just true, and it is better to decide about it deliberately than to discover it in a hiring conversation two years later.

The useful frame is not “should I share this”. It is “what is this post for”. A post that offers a reader something usable, written from a settled distance, does something for other people. A post written in the middle of the difficulty, asking the feed to hold it, is doing something else, and it is the version most likely to cost you something. Neither is dishonest. They are different acts.

What do 468 mental health posts show?

We isolated the mental health topic inside our 12,988-post cohort: 468 posts from 40 creators, covering burnout, stress, therapy, wellbeing, and health at work. Then we ranked those posts against each other, comparing the topic's own top decile (46 posts, 8 authors) with its own bottom half (234 posts, 26 authors). Eight authors is our minimum threshold for publishing a segment figure, so treat the top-decile column as indicative.

The engagement distance is large. The median top-decile mental health post drew 344 reactions and 105 comments. The median bottom-half post drew 79 reactions and 6 comments. A 105-comment median is the highest of any group in this series, which tells you what kind of response this material produces when it lands: not applause, conversation.

Why do questions fail in this topic?

This is the finding we did not expect. Across our whole cohort, a question mark in the hook is a mild positive: 19.0% of top-decile posts against 14.2% of bottom-half ones. Inside the mental health topic it inverts. Questions appear in 8.7% of the strongest posts and 13.2% of the weakest, and question-word openers appear in 2.2% against 11.5%, a five-fold difference in the wrong direction.

The likely reason is what those questions are. “Do you ever feel overwhelmed at work?” is not a question, it is a lead-in to a lecture, and readers of this topic in particular have seen it a thousand times. It asks the reader to raise their hand about something private in exchange for advice they have not yet been given a reason to want. The posts that work in this topic tell you something first.

Note the tense. “Some years ago” does a lot of quiet work: it signals that the writer is on the other side of the event, which is what lets a reader engage without feeling recruited into a crisis.

Where is the line between honest and performative?

Everyone can feel this line and almost nobody can describe it, which is why the advice on this topic is usually just “be authentic”. That is useless. Here is a more concrete version, based on what actually differs between the posts that land and the posts that make readers wince.

The last row deserves its own sentence. The fastest way to make a sincere post read as a performance is to end it with a service. If you sell coaching or wellbeing consulting, the post about your own burnout is the one place where the call to action does the most damage, because it retroactively reframes everything above it as setup. Say nothing. The people who want your help will find the button.

And the completeness test is the one to actually run before publishing: if nobody comments, is this post still doing its job? If the honest answer is no, it may be a message to a friend rather than a post.

What happens to your professional reputation?

We cannot measure this, so what follows is judgment rather than data, clearly marked as such. Three things appear to be true in practice.

First, the audience effect is asymmetric. A well-written post about recovering from burnout builds unusual loyalty with people who have been there, which is a large group. The people it makes uncomfortable mostly do not say so; they just quietly re-file you. You will see the upside in your notifications and never see the downside.

Second, the risk sits with real-time posts, not with the subject matter. A retrospective post about a difficult year reads as perspective. A post written on the worst day reads as a request, and a reader who cannot help you feels awkward rather than moved. Both are honest. Only one of them is a good use of a public professional feed.

Third, seniority changes the calculation in both directions. If you lead a team, your post about exhaustion is also a message to people who report to you, and they will read it as information about their own workload. That can be exactly the right thing to publish. It is never a neutral act.

The practical version: write the post, then wait a day before publishing it. Almost nothing in this category is time-sensitive, and a draft that still reads well tomorrow is usually the retrospective version rather than the raw one.

One more thing, since it decides more of this than any writing advice: consider who is actually in your audience. A founder posting about exhaustion is read by investors and customers. A person job-hunting is read by hiring managers who will draw conclusions they will never tell you about. A tenured employee at a stable company has a wider margin than either. None of that determines whether you should publish. It determines which version of the post you are choosing to publish, and that decision is better made in advance than discovered afterwards.

Why is video so dominant in this topic?

Native video accounts for 37.0% of the strongest mental health posts and 10.3% of the weakest, one of the widest format gaps we found anywhere. The mirror image is the collapse of shared links: 6.5% at the top, 38.9% at the bottom.

Both make sense for the same reason. This is a topic where the reader is deciding whether to believe you. A face saying a difficult sentence carries information that text cannot fake as easily, and an article somebody else wrote about workplace wellbeing carries none at all. Sharing a study about burnout with the caption “important read” is the single most common weak post in this category, and it is a category where borrowed material is particularly hollow.

If you are not going to film, the next best thing is a text post with the same properties: first person, specific, past tense. Format guidance generally is in the video posts breakdown, and the case against reflexive link-sharing is in the external links analysis.

What does the tone data say?

Two smaller signals in the same within-topic comparison are worth reading together. Emoji appear in 21.7% of the strongest mental health hooks and 1.3% of the weakest. At least one hashtag appears on 93.5% of the strongest and 56.4% of the weakest, while four-or-more stacks run 37.0% against 42.7%, which is close to flat.

The emoji figure is not an instruction to decorate a post about depression. It is a proxy for register. The posts that land in this topic are written the way a person talks, informally, in the first person, at 17.4% against 14.5%. The posts that do not land are written in the institutional voice: measured, general, tagged for reach, and about the subject rather than about anything that happened.

The hashtag split says something similar from the other direction. Strong posts tag themselves so the right readers find them. Weak posts tag themselves instead of saying anything. The difference is not the number of tags, it is whether the post underneath needed writing.

The practical read on all of this: write it the way you would say it to one person who already knows you a little. That single instruction produces most of the profile above. It gets you first person, it gets you a scene instead of a survey question, and it keeps you off the institutional register that produces the awareness-week post nobody finishes.

Should companies post about mental health?

This is a different question from whether a person should, and the honest answer is that most corporate mental health content on LinkedIn is doing nothing. The announcement pattern sits near the bottom of our exemplar set: programme launches, awareness days, partnership news. They are legitimate activities. They are just not posts, in the sense that nobody's feed is improved by them.

What does work, if a company wants to say something on this subject, is the operational version. Not “we care about wellbeing” but “we moved on-call to a two-week rotation and here is what happened to sick days”. A policy with a consequence attached is information. A statement of values is a poster.

The second thing that works is letting a named person write it. Every finding in this article favours first-person, specific, video, past tense, and a company page can do none of those things convincingly. If your organisation wants to be credible here, the useful move is to support a manager who is willing to describe what they changed and why, and then get out of the way of their phrasing. The general case for individual accounts over branded ones is in the personal branding guide.

The third thing, and the one almost nobody does: publish the thing you got wrong. A company post describing a workload problem it created and then fixed would be genuinely novel on this platform, which is exactly why it would travel.

What kinds of mental health posts are there?

Not all of these carry the same risk or the same return. Sorting them helps, because most people writing on this topic have only considered one of them.

• The retrospective. Something that happened to you, at a distance, with the part you changed. The highest-return and lowest-risk version, and the one the data most supports.
• The operational. What you changed about how your team works: on-call rotas, meeting-free days, the policy you rewrote. Almost no reputational risk, and it is the version that other managers copy.
• The correction. Pushing back on a norm, like performative overwork or sleep bragging. Argument-shaped, so it earns comments, as the example above shows.
• The witness.Writing about someone else's experience, with their consent. Handle carefully and never as a lesson you extracted from their difficulty.
• The announcement. A programme, a fundraiser, an awareness day. These sit low in our exemplar set. They are corporate messages about a personal subject and the feed treats them accordingly.
• The real-time. Written during. Sometimes the most important thing a person publishes. Also the one to sleep on first.

One post in our top set shows how honestly the mixed version can be done: pride and grief in the same breath, without either one being resolved for the reader's comfort.

And the highest-engagement mental health post in our exemplar set is not gentle at all. It is a hard, concrete account of consequences.

More hooks from this topic with the numbers attached are on the mental health examples page.

What should you not post?

Not a morality list. A list of things that reliably fail, for reasons the data or plain observation supports.

• The unresolved crisis, published live. It asks strangers for something they cannot give and it is the version most likely to be reread later by people making decisions about you.
• Somebody else's breakdown as your content. Including a colleague, a client, or a family member, whether or not you name them.
• The tragedy with a business lesson stapled on.“And that is why you need a CRM” is the fastest way to lose a reader permanently.
• The awareness-day post with nothing in it. A ribbon graphic and a hashtag on the relevant date, then silence for a year.
• The diagnosis you are guessing at, for someone else.Describing an ex-manager's mental state in public is a legal and human problem, not a writing one.
• The link with “important read” on top.38.9% of this topic's weakest posts are shared articles. It is the default and it does nothing.

How should you respond when someone else posts about this?

Nobody covers this and everybody faces it. This topic produces the longest comment threads in our data, which means most people encounter it as a reader before they ever write one. Three practical notes.

Do not diagnose, and do not recommend. “Have you tried” is the most common reply under these posts and it is almost always unwelcome, because the writer has usually tried it. The reply that works is the one that adds evidence rather than advice: what you saw, what you did, what it cost you. That is the same rule the posts themselves follow.

Do not make it a networking moment. A comment that pivots to your service, your newsletter, or your availability under a post about somebody's worst year is visible to everyone, including the people you were hoping to impress. It is the clearest own goal available on this platform.

And if you know the person, send a message instead of, not as well as, a comment. A public reply is a signal to the audience. A private one is a signal to the writer. They are different acts and only one of them is actually for the person who posted.

The one exception worth naming: if you are a manager and someone on your team posts something difficult, a public comment can be the right call, because it tells everyone else on that team how their own version would be received. Keep it short, keep it warm, and follow it with the private message.

How do you write one of these yourself?

1. Wait until it is past tense. Not because feelings need to be tidy, but because you cannot see the useful part while you are inside it.
2. Open on the scene, not the topic. The strongest hooks in this topic start in a specific room with a specific thing happening. Only 2.2% of them open with a question word.
3. Name the behaviour, not the emotion.“I stopped answering messages on Sundays” carries more than “I was overwhelmed”, and it is what a reader can copy.
4. Keep other people out of it. Unnamed, unblamed, unrecognisable.
5. Cut the offer. Whatever you sell, it does not belong in this post.
6. Run the completeness test. If nobody replies, does the post still do its job? If not, do not publish it yet.
7. Then answer the comments properly.This topic produces the longest replies of any we measured, and a one-word answer to a paragraph of someone's history reads badly. The comments guide covers the general version.

On craft, the mechanics of putting a real scene at the front of a post are covered in the storytelling guide, and the fold constraints are in the hooks study.

What this analysis cannot tell you

These are correlations, not causes. We can say that the strongest mental health posts in our cohort use video more than three times as often as the weakest. We cannot prove the format caused it, and this topic's top decile rests on 8 authors, our minimum threshold, so individual habits move the percentages. The cohort skews toward established creators with at least 1,000 followers across 65 accounts. Engagement counts are lifetime-cumulative at scrape time and post ages vary. All text findings describe the visible hook, not the full body. And nothing here is advice about your health or your employment situation; it is a description of how a public feed responds to a category of writing. Full method in the 34,000-post engagement study.

Writing about a different subject?

Each subject behaves differently, and we compared every topic against its own top decile rather than against the whole cohort: productivity, remote work, hiring, leadership.