Delivery window alerts
Setting this up takes about a minute. The slow part is not the software — it is deciding what to call yourself, because the handle ends up on a poster and cannot be changed afterwards.
What this actually looks like
A missed delivery costs two journeys and a refund conversation. A channel per round tells everyone on it that the van is an hour behind, that the road is shut, or that today’s drop is now tomorrow — once, instead of forty times.
How you would set it up
- Create a channel per round or per depot and give out the code when a customer books.
- Send the window in the morning and the exceptions as they happen.
- You get how many phones could be reached and how many opened it — not a delivery receipt, because this product keeps no per-device record and will not invent one.
Draft, schedule, send — and what cancelling cannot undo
Composing and sending are deliberately different permissions, so the person or script that writes an alert can be forbidden from publishing it. A scheduled alert becomes visible at its time without anything having to run at that minute, which means a server that was briefly down delivers late rather than never. Cancelling is honest about its limits: it stops an alert reaching phones that have not fetched it yet and does nothing at all to the phones that already have it. An alert cannot be unsent.
Draw a circle without knowing who is inside it
You place a centre and a radius on a map, and the circle travels attached to the alert. Every subscriber’s phone receives it and compares its own position locally, then decides whether to show it. No subscriber position is ever sent to us, which is why this is a design fact rather than a promise: there is no column anywhere that could hold one. Two consequences follow deliberately. A geofence only ever narrows — it cannot add somebody who did not subscribe. And it fails open: a phone that cannot work out where it is shows the alert rather than hiding it, because missing an evacuation is worse than seeing one meant for the next street.
Three numbers, and the one we will not print
You get: how many devices could be reached at the moment you sent, how many phones reported opening the alert, and how many acknowledged it. What you do not get is a delivery rate, because there is no per-device delivery record anywhere in this product — that absence is exactly what makes sending to four million people cost the same as sending to four. Reads and acknowledgements are self-reports from handsets that chose to send one, and they are labelled as such rather than dressed up as receipts.
Questions, answered plainly
Can I unsend an alert?
No. Cancelling stops it reaching phones that have not fetched it yet, and does nothing to the ones that already have it. Compose as a draft and read it back before you send; that habit is worth more than any undo button we could offer.
Do you know where my subscribers are?
No. The circle is sent with the alert and each phone compares its own position locally. We hold no subscriber location, so there is nothing to leak, subpoena or lose. A phone that cannot determine its position shows the alert rather than hiding it.
Why is there no delivery rate?
Because there is no per-device delivery record anywhere in the product, and that is a deliberate design decision rather than a gap. A phone’s inbox is a query over sent alerts, which is what makes the cost independent of audience size. We would rather show three numbers we can stand behind than a fourth we invented.
Do subscribers need an account?
No. They install the app, subscribe to a handle, and that is the whole of it. Only the person running the channel signs up for anything.
Say it once. Every subscribed phone gets it.
Free up to 100 subscribers, no card to start, and the handle is yours permanently.
Start a channel