StewaraGuidesSafety
Why a church emergency alert can't just be a text message.
At twenty to ten on a Sunday morning, nearly every phone in your building is on silent, because you asked them to be. That is also the hour you are most likely to need your safety team in the next ten seconds.
The hour when nothing gets through
Picture an ordinary Sunday. It is 9:40 in the morning and the service started ten minutes ago. The nursery is full. The teenagers are in the basement. Two greeters are still at the door with a stack of bulletins.
And nearly every phone in that building is on silent. Your own announcement slides asked for it, ten minutes before the countdown ended. The ushers set theirs before the prelude. Parents silenced theirs so a text would not go off during the prayer. The worship team left theirs in a box backstage. That silence is not a fault in your church. It is a room full of considerate people doing exactly what you asked them to do.
Now something happens. It does not matter what: someone collapses in the third row, a stranger tries the nursery door, the county puts out a tornado warning. The person who noticed needs the safety team, the nurse, and two elders, and they need them now.
So they do the natural thing and send a group text. And it lands — correctly, politely, silently — in forty pockets.
That is the whole problem, and it is worth saying plainly: an emergency alert system is not a messaging feature. It is a system that has to work at the exact moment messaging has been deliberately switched off.
What a text message is actually promising you
A group text feels sufficient because it always seems to work. You send one, people reply, life goes on. But nothing about a text message was designed for urgency. It is an ordinary notification, so it obeys a phone's ordinary rules, and a phone's ordinary rules are written to protect its owner from being interrupted.
Here is what a normal text or push notification does not do:
- Make a sound on a phone whose mute switch is on.
- Make a sound on a phone in Do Not Disturb, or in a Focus mode.
- Tell you who received it and who did not.
- Know who is in the building right now and who is at home with a cold.
- Say what the person receiving it should actually do.
Critical alerts: the one thing phones let you make louder
Phone makers know about this problem, and they built a narrow door for it. On an iPhone it is called a critical alert: a notification that is allowed to play a sound even when the phone is on silent and even when Do Not Disturb is on. Android has an equivalent idea: a notification channel the system treats like an alarm clock rather than a message.
Three things about that door are worth understanding before you shop for anything, because they explain most of what you will hear from vendors.
- You cannot simply switch it on. Apple grants critical alerts by written application, one app at a time, reviewed by a person. The developer has to describe the safety situation, who triggers an alert, and why an ordinary notification is not enough.
- It is only for genuine safety. Not marketing, not a reminder that pledges are due, not "your order has shipped." A vendor who uses it for anything else risks losing it.
- The person still has to say yes on their own phone, once. It is a separate permission prompt, on top of ordinary notifications. A church cannot make a member's phone loud without that member agreeing to it.
Reaching the right people, not simply everybody
The second thing that separates an alert system from a mass text is not how loud it can be. It is who it decides to be loud at.
The person who most needs to know that something is happening in the building is a person in the building. The member at home with a cold cannot help, cannot get there in time, and does not need a siren. Waking that person's phone with an alarm teaches them, gently and permanently, to turn your alerts off; a person who has turned your alerts off is a person you have already lost for the real emergency.
So a well-built system knows roughly who is on the property at this moment, and treats those people differently from everyone else. On campus: the full alarm. Somewhere else: a quieter heads-up, or nothing at all, depending on what happened. Someone who moved to another state two years ago and never left the church directory: nothing, ever.
That is a harder thing to build than a broadcast, and it is the single feature that decides whether people keep your alerts turned on.
Making it slightly hard to send
Any control that can alarm an entire building should be a little awkward to operate. This sounds backwards until you picture a single tap on a phone in somebody's back pocket sending a lockdown alert to two hundred people in the middle of a sermon.
The usual answer is a deliberate gesture: press and hold for a couple of seconds, rather than tap. It is hard to do by accident, it is easy to do under stress, and it gives the sender a moment to change their mind. That is the same reason a fire alarm sits behind a little glass panel instead of being a doorbell.
The other half is telling the sender what will happen before it happens: which alert type they picked, roughly how many phones are about to make noise, and who will be left alone.
How we do it
How Stewara does it
Stewara has a screen called Emergency alerts. It is not a paid add-on and it is not part of any plan tier. Every church on Stewara has it, and who may send an alert is decided by role, not by what the church bought. Life-safety was ruled out as something to sell, deliberately.
Sending is meant to be deliberate. You choose what is happening from a short list, you see a preview of who is about to be alarmed, and then you press and hold a button that reads Hold to send this alert. There is no single tap that alarms a building.
Alerts are scoped to the property. The screen puts it in its own words: an alert "Alarms the phones that are on campus right now. People who are somewhere else are left alone." After it goes out, the sender sees the delivery split — how many phones got the full alarm, how many got the quieter caution, and how many got a text — rather than a hopeful "sent."
On critical alerts, the part worth being exact about, Apple has approved Stewara's application. The written safety case described the situations a church actually faces: a lockdown, severe weather arriving in the middle of a service, an emergency in the kids' wing. Approval is not a switch we throw on our end, so the capability reaches phones in an app update. And as above, each person still approves critical alerts on their own phone, once. Nobody at the church can do that for them, and that is the right way round.
Until a church's phones are on that update, alerts ride at the level iOS calls Time Sensitive, which breaks through a Focus mode but not the physical mute switch. That gap is exactly why the rest of the design carries weight rather than being a nicety: if the Stewara app is open, the alert screen makes the alarm sound and speaks the announcement out loud, through the mute switch. If the app is closed and the phone is fully silenced, the alert arrives as a silent banner and the noise starts the moment it is opened. The on-campus tiering and the text fallback are what cover the distance in between.
The questions to ask, whatever you end up using
You do not need Stewara to have a good answer here, and plenty of churches will land somewhere else. What no church should do is assume the group text is covered. Whatever you are looking at — a dedicated alert product, a feature inside your church software, or a paging service — ask these, and ask for specifics rather than adjectives:
- When a phone is on silent and Do Not Disturb, what exactly happens to your alert? "It's a high-priority notification" is not an answer to that question.
- Does each person approve the loud alerts on their own phone? The right answer is yes.
- After I send one, can I see who received it and who did not?
- Can I reach only the people on the property, and leave everyone else alone?
- How hard is it to send one by accident?
- What happens if the building's internet is down?
- How do we run a test that does not frighten anybody?