Company Announcements That Do Not Get Ignored

You spent four hours drafting the announcement. Legal reviewed it twice. It went out Tuesday at 4:52 p.m. By Thursday, three people had asked questions answered verbatim in paragraph two, a manager in support had told her team the opposite of what it said, and the view counter — if you had one — would have read 34%.

Nobody involved was careless. The announcement failed for structural reasons: it competed for attention against a hundred other messages, buried its point under three paragraphs of preamble, went out once in one channel, and treated "sent" as "communicated." These are the four standard failures, and each has a known fix.

This is a mechanics guide: how to structure an announcement, choose its channels, repeat it without spamming, run the comments, and check whether it landed — with three worked examples you can copy. It assumes a functioning channel system; if your company has no agreed home for announcements at all, fix that first with an internal communications strategy and come back.

Why announcements get ignored

Worth thirty seconds on the mechanism, because the fixes follow from it.

An employee at a mid-size company processes somewhere between dozens and hundreds of workplace messages a day across chat, email, and meetings. Attention is triaged in about two seconds per item: Does this affect me? Do I need to do anything? Now? Any message that does not answer those questions in its first visible line gets deferred — and deferred, for internal comms, means dead. There is no second session with an announcement.

Layered on top is trained skepticism. If the channel that carries real news also carries the parking garage repainting schedule, employees learn the channel's average value and read accordingly. And if leadership's announcements habitually open with throat-clearing — "As we continue our journey…" — employees learn to skip openings, which is fatal when you finally lead with the point.

So the craft reduces to four disciplines: front-load the point, mark the significance honestly, put it where it persists, and repeat it in more than one form. Everything below is elaboration.

The anatomy of an announcement that lands

Use this structure for anything that matters. It is boring on purpose — the structure should be invisible so the content is not.

HEADLINE — the news itself, in one line
TL;DR — 2–3 sentences: what is changing, when, who is affected
WHAT'S CHANGING — the details, in short paragraphs or bullets
WHY — the honest reasoning, one or two paragraphs
WHAT THIS MEANS FOR YOU — by audience if needed; explicit actions + deadlines
TIMELINE — dated steps, if there are stages
QUESTIONS — where to ask, who owns answers, when an FAQ/AMA happens

Rules that make it work:

The headline is the news, not the theme. "Expense reports move to Ramp on Nov 1" beats "An update on finance tooling." Test: if someone reads only the headline, are they materially informed? Write the headline last, after you know what the message really says.

The TL;DR is not optional above 150 words. Most readers will read only it, by design. Do not resent this; write for it. A good TL;DR carries the change, the date, and the affected group: "Starting January 5, all customer-facing teams move to a 4-day support rotation. Engineering and G&A are unaffected. Nothing is required from you before December: your lead will bring rotation drafts to team meetings the week of Dec 8."

"What this means for you" is the section people actually search for — write it even when the answer is nothing. "No action needed — this only changes how finance closes the books" is a gift. When impacts differ by group, split it: For managers… / For ICs… / For anyone hiring… If you cannot write this section, you do not understand your own announcement yet; the drafting failure is the useful early warning.

The "why" must contain the real reason. One honest paragraph — including the tradeoff, if there was one — outperforms five paragraphs of framing. "We chose X over Y because Z, and we accept that this costs W" is the sentence that builds trust. Employees have a finely calibrated detector for reasoning that has been sanded down to nothing; a why-section that could justify any decision justifies none.

Length caps: 300–400 words in the announcement itself; everything else links out. FAQ, policy text, and methodology live in a linked doc for the minority who need depth. The announcement is a front door, not the building.

Timing is a real variable. Tuesday–Thursday mornings beat Friday afternoons for anything you want read and discussed (the Friday 4:52 p.m. send is read, correctly, as burying). One hard rule: never let significant news land while managers are as surprised as their teams — pre-brief managers 24–48 hours ahead with the draft, the reasoning, and likely questions, so the "wait, what does this mean?" conversations that start within minutes have informed answers.

Choosing the channel (and the significance tier)

The channel question resolves cleanly if you separate two properties: where the announcement lives and how people are told it exists. Conflating them is the root of most channel mistakes.

The announcement itself needs a persistent, findable home — a company feed post that can be pinned, commented on, and found by search in March when someone asks "wait, when did the expense policy change?" Chat is a terrible home (it scrolls away and cannot be the record) but a fine doorbell: a one-line pointer with a link. Email works as a doorbell for populations who live in email. All-hands are amplifiers and context-setters, never the primary record — a third of the company is on PTO, in a customer call, or half-listening.

Then tier by significance, and be disciplined about the tiers, because over-alerting minor news is how you deafen people to major news:

Tier	Examples	Home	Doorbells	Manager pre-brief
1 — Major	Org changes, strategy shifts, compensation policy, layoffs	Pinned feed post	Chat + email + all-hands segment + newsletter	Always, with talking points
2 — Standard	Tool changes, process updates, notable hires, launches	Feed post	Chat pointer + newsletter	If action required of teams
3 — Minor	Office logistics, small policy clarifications, reminders	Newsletter item only	None	No

The counterintuitive discipline is tier 3: most announcements should not be standalone announcements at all. Batching minor news into a weekly roundup (the internal newsletter is built for exactly this) protects the signal value of the announcement channel — so that when a standalone post appears, its mere existence says "this one matters."

Repetition: the strategy nobody wants and everyone needs

The hardest sell to leadership is that one send is never enough. The instinct is "we announced it already" — but marketing learned generations ago that messages need multiple exposures to stick, and your internal audience is not more attentive than customers. At any given send, a chunk of the company is on vacation, heads-down, or in the wrong timezone. Reasonable planning assumption: any single touch reaches half of its audience; comprehension needs three to five touches in varied forms.

The move that makes repetition palatable is varying the form so it reads as reinforcement, not spam. A tier-1 sequence looks like:

1. Day 0: Manager pre-brief (written, with FAQ seed).
2. Day 1: The feed post — the canonical record — plus chat/email pointers.
3. Day 1–3: Open comment window on the post; comms and the decision-owner answer publicly (see next section).
4. Week 1: All-hands segment — not a re-read, but the why in the leader's own words plus live or pre-collected Q&A.
5. Week 1: Newsletter inclusion with a one-line recap and link.
6. Week 2–3: The operational follow-up: "Rotation drafts are out — here's what happens next," which re-teaches the original news to everyone who missed it, disguised as an update.
7. At the deadline: Reminder scoped only to people with outstanding actions, not the whole company.

Seven touches, one of which is a repeat in form. Nobody experiences this as spam; most people experience it as "the company communicated that well." Meanwhile the FAQ doc quietly accumulates every question asked anywhere, so answers converge instead of forking.

One more repetition rule: update the original post rather than posting corrections downstream. The canonical record gets edits appended ("Update, Nov 4: contractor timeline clarified below"), so latecomers who find it by search read the corrected version, not the archaeology.

Comments, questions, and the silence problem

Should announcements allow comments? Yes, with narrow exceptions — and this is worth being opinionated about, because closed comments are the more common and more damaging default.

Closing comments does not close the conversation; it relocates it to DMs and private channels where wrong answers replicate unchecked and comms cannot see, correct, or learn from any of it. Open comments concentrate the questions where one good answer serves everyone, generate a free FAQ, and — the underrated effect — give leadership a visible occasion to be responsive in public, which compounds trust in every future announcement.

Running comments well:

• Staff the first two hours. Questions cluster immediately after the doorbells; the decision-owner (not only comms) answers in that window. A question answered in twenty minutes by the VP who made the call is worth more culturally than the announcement itself.
• Answer the hard question first and plainly. Skipping the pointed question to answer the softball is noticed by everyone watching — and everyone is watching.
• It is fine to say "we don't know yet" with a date. "Rotation specifics aren't final; drafts come to team meetings the week of Dec 8" beats improvised precision that gets corrected later.
• Move the personal to private, visibly. "That depends on your situation — DMing you" handles individual cases without letting the thread become one.
• Silence is data, not consent. A tier-1 announcement with zero comments almost always means the conversation moved somewhere you cannot see. Prompt it: ask managers to surface their teams' questions into the thread or FAQ, and run an async AMA where questions are collected (optionally anonymously) beforehand — written question collection reliably surfaces sharper questions than raised hands ever do.

Exceptions where comments-off is right: legally constrained situations (layoff logistics, active litigation) and moments of acute personal grief where a comment thread would be gruesome. Even then, always pair the closed post with a stated open path: "Questions to this form or your manager; we'll publish answers Thursday."

For genuinely hard news — reorgs, pivots, layoffs — announcement mechanics are necessary but not sufficient; sequencing and empathy carry more weight, and that is a separate playbook: Communicating Change.

Three worked examples

Skeletons you can adapt. Note what the openings do not do: warm up.

1. Policy change (tier 2)

PTO carryover is changing: 5-day cap starting Jan 1

TL;DR: From Jan 1, up to 5 unused PTO days carry into the next year (previously unlimited). Days above 5 must be used by Dec 31 or are forfeited. This applies to everyone; 78 people currently have more than 5 banked days and will get an individual note today.

Why: Unlimited carryover sounded generous but produced the opposite — people banked vacation instead of taking it, and our actual usage last year averaged 14 of 20 days. A cap plus a use-it push is the standard fix, and we would rather you rest than accrue.

What this means for you: Check your balance in the Time Off room. Over 5 days? Book the excess before Dec 31 — managers have been asked to approve generously this quarter. At or under 5? Nothing to do.

Questions: Comments below, or the FAQ (link). Policy owner: Dana (People).

2. Org change (tier 1 — after manager pre-brief)

Support is moving into the Product org on March 1

TL;DR: The Support team (11 people) moves from Ops to Product, reporting to Priya. No roles are being eliminated and no one's compensation changes. Day-to-day workflows stay the same through March; changes to escalation paths come in April, drafted with the team.

Why: Support sits closest to the friction our roadmap should fix, and the handoff from support insight to product decision has been slow — three steps and two org boundaries. We are removing the boundaries rather than adding a process. The tradeoff we accept: Ops loses its largest team, and we will backfill Ops' on-call coordination (plan by Feb 15, owner: Marcus).

What this means for you: Support: Priya is holding team and 1:1 sessions this week — invites are out. Everyone else: escalation paths are unchanged until April; you will hear from us before anything shifts.

Timeline: Feb 15 Ops coverage plan → Mar 1 reporting change → April escalation redesign with the team.

Questions: Comments open; Priya and Marcus are answering today. Async AMA Friday — question form (anonymous OK): link.

3. Tool migration (tier 2, action-heavy)

We're replacing Trello + Confluence with one workspace — migration starts Nov 10

TL;DR: Both tools sunset Dec 15. Boards and docs migrate team-by-team starting Nov 10 (schedule linked). Your action: 15-minute onboarding before your team's cutover date, and flag any board or doc you want excluded from migration by Nov 7.

Why: We pay for six overlapping tools and search none of them well; consolidating removes two subscriptions and puts boards and docs under one search. The tradeoff: two weeks of switching cost per team, which the pilot (Design, thanks for guinea-pigging) put at "annoying for three days, fine after."

What this means for you: Your migration date and 15-min session: schedule (link). Old tools go read-only on your cutover date, deleted Dec 15 after export archives are saved (link).

Questions: #tool-migration channel; migration owner: Sam. Daily office hours at 11:00 through Nov.

Sequencing across time zones, shifts, and desks

Everything above assumes an audience that is, roughly, online together. Many companies are not, and the standard failure is announcing on headquarters time and letting everyone else wake up to a 200-comment thread — which reads, accurately, as "you found out second."

For distributed teams, three adjustments:

• Anchor to the canonical post, not the live moment. If the all-hands is the emotional center of the announcement, record it and post the recap within the hour — but let the feed post carry the news, timed so no region gets it mid-night and discovers it eight hours stale. For genuinely major news, a send window at the start of the overlap between your two largest regions is the least-bad slot.
• Run the comment window for 48 hours, not two. Staff it in two shifts so the questions from Singapore get same-day answers from someone authorized to give them, not a "the US team will respond tomorrow."
• Repeat the AMA once per major region rather than forcing one heroic time slot. Written pre-collected questions make this cheap: same FAQ doc, two 30-minute sessions.

For deskless and shift-based staff, the doorbell logic inverts: persistent homes they rarely browse, so the push layer carries more weight — mobile notifications, shift-start huddles with a one-page brief for the shift lead, and printed one-pagers where phones are not on the floor. The manager pre-brief stops being a courtesy and becomes the primary channel: if the shift lead cannot explain the change in ninety seconds, the announcement has not happened for that team, whatever the view counter says.

The pre-send checklist

Tape this next to the publish button. Ninety seconds of checking prevents most of the failure modes in this guide:

1. Does the headline alone inform someone who reads nothing else?
2. Does the TL;DR contain the change, the date, and who is affected?
3. Is "what this means for you" present — split by audience where impacts differ — with explicit actions and deadlines, or an explicit "no action needed"?
4. Does the why state the real reason, including the tradeoff?
5. Have managers been pre-briefed (tier 1, or any tier with team-level actions)?
6. Is it posted in the persistent home, with doorbells that point rather than duplicate?
7. Is the repetition sequence scheduled — not intended, scheduled — including the newsletter recap and the deadline-scoped reminder?
8. Are comments open and staffed for the first window, with the decision-owner on deck?
9. Is there a named question-owner and a linked FAQ that will absorb answers from every channel?
10. Is the timing defensible — mid-week, mid-morning, no region asleep for tier-1 news, and nothing that looks like a Friday burial?

If you cannot check item 3, stop: the announcement is not ready, and no amount of channel strategy will save a message that cannot say what it means for its reader.

Did it land? Close the loop

"Sent" is not the finish line; verified comprehension is. Lightweight checks, in ascending order of effort:

• Reach: views on the canonical post within 72 hours, against your own baseline. A tier-1 post underperforming the average tier-2 post is a fire alarm — re-doorbell it.
• Engagement quality: count questions answered publicly vs. corrections you had to issue. Corrections trace back to drafting failures; log which section failed (usually a missing "what this means for you" case).
• Recall: two factual questions in the next pulse survey. If under half the company can state the new carryover cap a month after two touches, your repetition sequence needs another form, not another send.
• Action completion: for action-carrying announcements, the only metric that matters — track completion against deadline and chase the outstanding individually, never with another all-company blast.

The compounding effect is real: every announcement that is clear, honest, findable, and responsive raises the open-rate on the next one. Channels have reputations, and reputations are built one send at a time.

Next steps: Steal the anatomy template and the tier table into your comms playbook. Pre-brief managers on the very next tier-1 item. Batch this week's minor news into one roundup instead of three posts. And make sure your announcements have a persistent, searchable, commentable home — if they currently live in scrolling chat or email, that is the structural gap to close first. Openbook's Feed room was built to be exactly that home: pinned announcements, comments and reactions, polls for quick sentiment, and a bulletin sidebar for the standing stuff — free to start at openbook.work.