How is this different from a mass notification system?

Mass notification systems push short one-way messages to a contact list or a geographic area. They are built to interrupt, and they do it well. They cannot hold a forty-minute briefing, take questions, or run in several languages at once. The two are complementary: the alert tells people something happened, the briefing explains what to do about it.

Do residents need to install an app?

No. A listener opens a link or scans a QR code and hears the audio in their phone's browser. There is no download, no account, and no registration step. That matters during an incident, when asking thousands of people to install software is itself the obstacle.

Is this a replacement for radio or land mobile radio?

No. Land mobile radio, P25, and FirstNet exist to keep responders talking to each other on infrastructure built for that job, and broadcast radio reaches people with no connectivity at all. This sits somewhere else entirely: it is a way to talk to the public at length, on the devices the public already carries.

Emergency communications: the briefing layer

Q: Does this work when the cell network is down?

No. Phone-delivered audio needs a working data connection at both ends. In a total communications blackout nothing internet-based reaches anyone, and broadcast radio remains the only medium that does. This approach is for the far more common case where the network is congested, patchy, or partially restored rather than gone.

Every jurisdiction has a way to send an alert. Almost none has a way to hold the conversation that comes next — the forty minutes where officials explain the evacuation route, the shelter list, which roads are open, and what happens to people whose houses are gone.

Alerts are built to interrupt. A wireless emergency alert is a few hundred characters, delivered once, in one language, with no way to reply. That is the correct design for "leave now." It is a hopeless design for "here is the plan."

An emergency briefing channel is live audio the public can join by opening a link — no app, no account, no receiver — so officials can explain a situation at length, in more than one language, to people who are not in the room.

It is the same structural problem this site is built around. Crowd overflow is what happens when the audience outgrows the reach of the sound system. In an emergency the audience is an entire county, the room is a fire hall with sixty chairs, and the overflow is everybody else.

Start with what this does not do

It does not work in a blackout Phone-delivered audio needs a working data connection. When Hurricane Helene came ashore in 2024 it knocked out more cell sites than any hurricane on record and destroyed well over a thousand miles of fiber in North Carolina alone; in the worst-hit counties, broadcast radio was the only thing still reaching people, and local stations ran as an improvised dispatch service for days. Nothing on this page changes that. If your plan assumes the internet, your plan needs a layer that does not.

Say that plainly and the rest of the argument gets easier, because the total blackout is not the only failure mode — it is the rarest one. Far more often the network is degraded but alive: towers standing but congested, service restored in patches, a Wi-Fi trailer parked outside the shelter, one bar in the parking lot of the community centre. That window opens within hours in most incidents and lasts through the entire recovery, which is where the majority of public communication actually happens.

Why audio survives a congested network

The instinct during an incident is to stream the briefing on social video. It is free, it is familiar, and it is the recommendation in most public information guidance. It is also the heaviest possible way to move a human voice.

Live video needs a sustained, high-bandwidth connection. Speech does not. The Opus codec used for real-time web audio runs anywhere from about 6 kbps up, and adapts downward continuously as a connection deteriorates rather than stalling and rebuffering. On a tower that is standing but saturated, the voice gets through when the video will not — and the person with one bar in a parking lot is precisely the person who needs the briefing most.

The practical framing Keep your alerting system for alerts. Keep radio for the blackout. Add a briefing channel for the long tail — the town hall, the shelter announcement, the daily crew brief, the two weeks of recovery questions that no 360-character message was ever going to answer.

Where it fits

Five situations where the audience is larger than the room, connectivity is imperfect but present, and the message is too long for an alert.

Language access

The obligation with no product behind it

Federal civil rights law requires meaningful access for residents with limited English proficiency. The standard playbook for post-disaster public meetings does not mention interpretation at all. Live translation to each listener's own phone closes a documented gap.

Read the compliance case → Post-disaster town hall

Sixty chairs, four thousand households

The recognised ritual of recovery: officials, FEMA and insurance representatives in a room, answering the same questions for two hours. Everyone who could not drive there hears it live, and the transcript becomes the record.

Why the room is always too small → Shelters & distribution points

Announcements without waking the room

Four hundred cots, no PA, and a 6am announcement about water distribution. Audio to each person's own phone and earbuds reaches the people who need it in the language they need, without shouting the room awake.

Where PA coverage runs out → Crews & volunteers

The daily safety brief, county-wide

Debris crews, mutual-aid teams, and camp supervisors spread across a working area no PA covers. One link, one briefing, delivered in more than one language, to whoever is holding a phone.

Reaching listeners the room lost → Rumour control

Correcting the record at the speed it spreads

Misinformation is the complaint public information officers raise most. A standing channel residents already know the address of gives officials somewhere to answer it in their own voice, at length, the same day.

What Instant Audio is → Before anything happens

Work out what your PA actually reaches

Most emergency plans inherit a sound system nobody has measured. The calculator gives you the honest coverage figure for a shelter, a staging area, or a community hall before you need it.

Open the calculator →

How it compares to what you already have

None of these replace each other. The point of the table is to show which column is empty in most jurisdictions.

General characteristics. Specific products and deployments vary.
	Mass notification	Broadcast radio	Social video	Briefing channel
Works with no connectivity	No	Yes	No	No
Carries a long explanation	No	Yes	Yes	Yes
Survives a congested network	Yes	Yes	Poorly	Yes
Several languages at once	Limited	No	No	Yes
Audience can ask a question	No	No	Comments	Yes
Needs an app or account	Opt-in list	No	Usually	No
Produces a record afterwards	Send logs	No	Video	Transcript
Reaches people outside the jurisdiction	No	No	Yes	Yes

That last row is quietly one of the most useful. In an evacuation a large share of the affected population is somewhere else — with relatives two counties over, in a motel, or watching from another state and trying to work out whether their parents' street is safe. A geographic alert cannot reach them by definition. A link can.

The app-install problem, and who already proved it

The best-known success story in citizen disaster communication is the push-to-talk app volunteers used to coordinate boat rescues during Hurricane Harvey. Dispatchers and boaters ran the whole operation on one channel, and ahead of the next storm several million people downloaded the app in anticipation. It genuinely saved lives and it deserves the credit it gets.

It also shows the cost of the model. Getting that far required millions of people to find, download, install and learn software in the middle of an emergency — and everyone who did not, or whose phone was too full, or who never heard about it, was outside the system. Anything that opens in a browser from a link starts at the other end of that problem.

What it costs, and why that is not the lead argument

Emergency notification contracts are public records, and the published figures run from a single village paying tens of thousands a year to statewide agreements in the millions across a multi-year term. Many jurisdictions pay nothing directly, because a state or county contract covers them.

So cost is a poor opening argument, and we are not going to pretend otherwise. What those contracts buy is alerting — and alerting is worth paying for. The observation worth making is narrower: the briefing layer is usually the part nobody has budgeted for at all, because until recently the only options were a room with sixty chairs or a social video stream that the people with one bar cannot load. Priced as a flat subscription fee rather than per resident, it tends to be the cheapest line in an emergency communications plan rather than the most expensive.

Verify before you rely on it Procurement figures move, vendor terms differ by state contract, and the regulatory obligations described across this section are summarised rather than quoted. Check current numbers and current rules for your own jurisdiction before building a plan or a budget on anything written here.

Frequently asked

Does this work when the cell network is down?

No. It needs a working data connection at both ends. In a genuine blackout nothing internet-based reaches anybody and broadcast radio is the only medium that does — which is why radio should stay in your plan. This is for the congested, patchy, and partially-restored conditions that describe most of an incident's duration.

How is this different from our mass notification system?

Different job. Notification systems are built to interrupt with a short message and they do that well. They are not built to carry a two-hour town hall, take questions, or run several languages simultaneously. Keep the alerting system; this is the layer after it.

Do residents have to install anything?

No. They open a link or scan a QR code and the audio plays in the phone's browser — no download, no account, no sign-up. Printed QR codes work well on shelter noticeboards and distribution points, where people are standing around with phones and no idea where information is coming from.

Is this public-safety-grade or FirstNet-certified?

No, and you should treat any vendor claiming otherwise with suspicion unless they can name the certification. This runs over ordinary consumer internet on ordinary consumer phones. It is a way to talk to the public, not a replacement for the hardened networks responders use to talk to each other.

Can we set one up during an incident, or does it need planning?

It can be stood up quickly, which is the point — but the useful preparation is deciding in advance who is authorised to speak, what the channel is called, and how residents will be told the address. The technology is the easy part; the communications plan around it is not.

What happens to the recording?

A live transcript can run during the briefing and stay attached to it afterwards. Agencies have to produce after-action documentation and respond to records requests regardless, so a briefing that documents itself removes a task rather than adding one.