Most announcements get eleven likes and no questions, and the owner reads that as approval. It is not approval. It is the sound of a room that has finished reading and has nothing to do with what it read.
Announcing is the most-used and least-examined tool an owner has. Here is what separates the ones people act on from the ones that scroll past.
The test: are you informing, or asking?
Before writing anything, work out which of two things you are doing, because they are not the same post.
An informing announcement changes what members know. An asking announcement changes what they have to do — turn something on, move somewhere, choose between options, show up on a date. The second kind earns attention because ignoring it has a cost. The first kind competes with everything else in their day and usually loses, which is fine, as long as you knew that going in.
Trouble starts when you write an informing post and expect an asking response. That is the announcement that gets eleven likes, and then a month later half the community has never heard of the thing.
Most of it should not be an announcement
The instinct to announce scales badly. An owner who announces everything has taught the room that announcements are ambient noise, and the one that mattered arrives into that habit.
| What you want to announce | What it actually is | Better |
|---|---|---|
| A small new feature or space | A change three people asked for | Tell those three by name; let the rest find it |
| A rule change | A decision that binds everyone | Announce it properly — this is the real case |
| "We just hit a thousand members" | A fact about you | Skip it, or make it about what they built |
| A relaunch | An admission that it died | Do not. Rebuild first, quietly |
The rule of thumb: announce things that bind people, not things that please you. Rule changes, price changes, closures, anything with a date they have to meet. Almost everything else is better as an ordinary post that happens to be good.
The first sentence is the whole announcement
Assume nobody reads past it, because most people will not, and write accordingly.
That means the first sentence carries the change and the date, and everything after it is for the minority who want detail. "From the first of next month, the Monday call moves to 16:00 UTC" is an announcement. "We have been thinking a lot about how to make our calls work better for everyone across our growing global community…" is a warm-up act for an announcement that some people will never reach.
Write the first sentence last. Draft the whole thing, find the one sentence a member would repeat to another member, and move it to the top. If no such sentence exists, you do not have an announcement yet.
Say what changes for them, not what you did
Owners write announcements from the inside, describing the work. Members read from the outside, looking for the consequence.
"We have migrated the resource library to a new structure" is the work. "Your saved links are now under Resources — nothing was deleted" is the consequence, and it is the only part anyone needed. The effort you spent is invisible and should stay that way; a member who has to translate your announcement into its consequence will usually just not bother.
For anything that could read as a loss, say what was preserved. The unasked question behind every change announcement is "did I lose something", and an announcement that does not answer it generates a week of the same question in replies.
Where it goes decides who reads it
Placement does more work than prose, and it is where most announcements are quietly lost.
One post, in the space people actually read, under a person's name. Not a banner nobody sees twice, not a footer note, not a separate announcements area that members visit exactly never. Pin it if it has a deadline, and unpin it when the deadline passes — a pinned post from five months ago teaches people that pinned things are stale, which costs you the next one.
Under your own name matters more than it sounds. An announcement from the brand reads as policy; the same words from a person read as somebody telling you something, and only the second kind gets replies worth having.
Announce once, then repeat on purpose
An announcement is read once, by whoever happened to be there when it landed, and that is a much smaller group than your member count.
So the repeat is not nagging, it is arithmetic — especially if your members are spread across more than one clock, where a single post reaches whoever was awake. Plan two or three touches at different hours before you write the first one, and make each one shorter than the last rather than rewriting it.
What does not work is the anxious re-post — the same announcement again with "in case you missed this!" on top. It reads as low confidence, and it trains people that your first post was optional.
What the software will and will not do
It is worth knowing exactly where the tooling helps, because the choice it puts in front of you is the one this whole post is about.
Mateflow treats a post and a Broadcast as two different things. A post lives in a space and can be answered. A Broadcast goes out on up to three channels — In-app, Email and Push — and the composer gives each one its own content: a subject and a body for email, a short message in-app, one line and a link for push. That separation is the useful part, because those three are read in completely different states of attention. The in-app message is also typed as Announcement, Update, Alert or Reminder, and choosing honestly is the whole discipline — if everything is an Alert, nothing is.
The audience is the setting most people miss. You can send to everyone, to members with a verified email address, or to people active within a set number of days, and narrow further by space or tier. That last option is the one that matters: it is how you announce something to the people who are actually here instead of to a list that includes everyone who left. Two honest limits — a broadcast's status can come back as partial, meaning some of it went out and some did not; and the email channel is not on every plan.
The uncomfortable part is that it will tell you how it went. In-app reports sent and read; push reports sent, delivered and clicked; email reports those three plus opened and bounced — and sent is not the same as delivered, which is its own small education. Most owners never look, which is the only reason the eleven-likes announcement survives as a habit.
The announcements you should never make
Three categories where the announcement itself is the mistake, no matter how well written.
Do not announce a behaviour you want. "This is a place where we all help each other" is a post about a norm rather than an instance of one, and norms are set by what you do visibly, never by being described. The announcement version reads as an accusation that people are not doing it.
Do not announce a removal. A public post about somebody being removed invites a referendum and hands them an audience; do it once, plainly, in private. The exception is when members will obviously notice and speculate, and even then it is one line, not a case file.
Do not announce a relaunch. Reviving a community does not start with a post; it starts with a small group and one reason to come back. Announcing a relaunch to everyone who already left is how you convert a quiet community into a publicly failed one.
Three things not to do
Do not announce something you have not finished deciding. "We are thinking about moving to Thursdays" invites a debate you did not want and cannot cleanly end. Decide, then announce, then genuinely listen — or ask first and call it a question, but do not dress one as the other.
Do not bury the date. Anything with a deadline puts the deadline in the first sentence. A date in paragraph four is a date that will be missed by the people it applies to.
Do not announce and then let it lapse. The thing you announced and quietly dropped is remembered longer than the thing you never announced, and it is a withdrawal from a balance you cannot see. If it died, say it died.
The one you have to get right
There is exactly one announcement where craft is not optional, and it is the last one.
Closing a community is the case where every rule above tightens: one post in the main space under your own name, the date and the consequence in the first line, what happens to their money and their data stated rather than implied, and moderators told privately first. It is also the one people most want to soften, and softening it is what turns a graceful ending into a public complaint.
If you can write that announcement well, you can write any of them — it is the same skill under pressure, which is a reasonable argument for practising on the small ones.
The bottom line
An announcement is not a way of telling people something. It is a request for a slice of attention that you are spending whether or not you get anything back.
So spend it on things that bind people, put the change and the date in the first sentence, write the consequence rather than the work, and pick the channel and the audience deliberately instead of sending everything to everyone. Then go and look at how many people actually read it — the number is usually humbling, and it is the fastest cure there is for announcing things that did not need announcing.