Your alert went out. Nobody answered. Now what?

Here is a failure that does not show up in any product comparison, because every product in the category passes the test that would catch it.

A worker misses a check-in. The system detects it correctly and sends an alert to the supervisor, exactly as configured. The supervisor is driving, with the phone in a cupholder on silent. Forty minutes later they see it.

Nothing malfunctioned. Every component did its job. And the worker was alone for forty minutes after the system already knew something was wrong.

That is the gap between a notification and an escalation, and it is the most expensive gap in lone worker safety because it is invisible until it is not.

A notification discharges an obligation. An escalation solves a problem.

A notification is a one-shot event. It fires at a configured recipient and then it is done. From the system's point of view the job is complete the moment the message leaves. Whether a human being ever became aware is somebody else's department.

An escalation has a different definition of done. It is not finished when a message is sent. It is finished when a person acknowledges. Until that happens it keeps working, moving down an ordered list, and the failure of any one attempt is simply an instruction to try the next.

That ordered list is the thing the whole category is named after. An alert roster is long-standing emergency services and military terminology for the sequence of people a call-out reaches, in order, until someone answers. The concept is older than any of this software. What software adds is that it does not get tired, distracted or embarrassed about calling somebody senior at 2am.

What an escalation needs in order to work

Four things, and three of them are organisational rather than technical.

An order, decided in advance. Not a group blast. A blast to six people reliably produces the bystander effect, where everyone assumes somebody closer is handling it. An order means position two knows that if they are being contacted, position one did not answer.

People who consented. Everybody on the roster agreed beforehand to be reachable. A roster assembled from a staff directory is a list of people who will be surprised, and surprised people do not acknowledge quickly.

A real backstop. The last position has to be somebody who will always answer, and in a small operation that is frequently the owner. Organisations find this uncomfortable, which is the point. If the final position on your list is unstaffed, your escalation terminates in silence.

Acknowledgement that means something. A person actively claims the incident. Not a read receipt, not a delivery confirmation, not "the push was accepted by the device". A human saying "I have this."

The free watcher problem

Here is a pricing decision that is really a safety decision.

Most tools in this category charge per watcher, meaning per person who can receive and acknowledge an alert. Think about the behaviour that creates. Every additional name on your escalation roster costs money, so there is a standing financial incentive to keep the list short.

Short lists are exactly the failure above. The organisation that charges itself for depth buys itself a roster with one name on it, and one name is a notification wearing an escalation's clothes.

We make watchers free and unlimited, which as far as we know nobody else does. Put the owner on the list. Put the night supervisor on it. Put the person's spouse on it if that is who will actually pick up. The system should never be the reason somebody was left off.

The record is half the product

Everything above produces a byproduct that turns out to matter nearly as much: an account of what happened. Who was on duty, who was contacted, in what order, at what time, who acknowledged, and how long it took.

For anyone whose obligation came from an insurer, a client contract or a regulator, that record is frequently the actual deliverable. Washington's rule, for instance, carries recordkeeping duties alongside the device requirements. Nobody audits your intentions. They audit the log.

It is also the only way to find out whether your roster works before you need it to. Run the escalation as a drill and read the record. The result is usually instructive and occasionally alarming.

Four questions for any vendor, us included

1. When the alert goes out and nobody responds, what happens next? If the sentence ends at "notifies", stop there.
2. How many people can receive an alert, and what does each one cost? If the answer is per head, understand what that will do to your roster over time.
3. What counts as acknowledgement? A delivered push is not a person.
4. What does the record show afterwards, and can I export it?

We answer those with escalation until acknowledged, free unlimited watchers, an explicit human acknowledgement, and an exportable record. What we do not answer them with is a staffed centre, because there is nobody at our company watching your people. Alerts go to the roster you configure and escalate through your own team. We are not an emergency service. We contact one on your behalf only through automatic SOS on a Garmin inReach, which Garmin Response receives, and outside that we contact nobody but your own roster. That is a deliberate shape rather than a missing feature, and it is a large part of why our pricing looks the way it does.

---

What AlertRoster is, and is not

AlertRoster is a call-out notification and escalation tool. It notifies people who have agreed in advance to be notified, and records what happened. It is not an emergency service. It contacts one on your behalf only when your organization has turned on automatic SOS for a Garmin inReach registered to you: if you miss a check-in while your phone has no signal, the AlertRoster app declares an SOS on that inReach, and Garmin Response, Garmin's emergency coordination service, receives it under your organization's Garmin plan. It is not a fire alarm, a security alarm, or an alarm monitoring service, and it holds no life-safety certification. It is not a replacement for your team's official paging arrangements, and it should not be the only way a call-out can reach your people.

Sources reviewed October 2026. Statutes and rules change; if you are reading this long after that date, check the current text.

About the author

Jim Hankins is the founder and CEO of Cloud Bedrock, LLC and the developer of AlertRoster, which he first launched in 2009. A US Army combat veteran and a 42-year veteran of IT and software development, he has spent much of his career building emergency communications platforms, including as CTO of GEOS, the international emergency response coordination service later acquired by Garmin.