DEV Community

Hive80-lab
Hive80-lab

Posted on

The 4 Messages That Save a Small Team During an Incident (Templates Inside)

When your service goes down, the fixing gets all the attention — but what customers remember is the silence. A team of 10 without comms templates spends the first 20 minutes of every incident deciding how to say "we're on it."

Here are the four messages every small team (1–50 people) actually needs, ready to copy into your runbook tonight. Brackets = fill during the bad day.

1. The first customer ack (within 15 minutes)

We're aware of an issue affecting [plain-words symptom — "sign-in", not "auth service"]. We're on it and will post an update by [time +30 min]. No action needed from you.

Two rules: name the symptom the customer sees, and always promise the next update at a specific time.

2. Status page updates (every 30–60 min)

[HH:MM] Still working on [issue]. Progress: [one sentence of real progress]. Next update: [time].

Same three slots every time. If nothing changed, say so — "no change yet, next update 10:30" is a good update. Customers forgive slowness; they don't forgive discovering nothing was happening.

3. The exec update (five lines, no jargon)

Issue: [one line]. Customers affected: [group]. Money impact: [orders/logins]. Fix: [one line]. Next update: [time].

Five lines. If it's longer, you've started defending instead of informing.

4. The resolution note

Resolved as of [time]. What happened: [plain words]. What we're doing so it doesn't repeat: [one concrete action]. If anything still looks odd, reply here — we're watching until [time].

It's the only message some customers will ever read, so it can't lean on earlier updates. And "we're watching until X" is how a bad day becomes a story about how you handle bad days.

Why this matters more at small scale

Enterprise teams have comms staff. A team of five has you, mid-incident, typing under pressure. Templates aren't bureaucracy — they're the difference between a 2-minute update and a 20-minute freeze.

The same logic applies to postmortems: one screen, impact + timeline + contributing factors + exactly three actions with owners and dates. A postmortem that names a person as the cause is the last honest postmortem that team writes.


If you'd rather not assemble this from scratch, I packaged it:

Launch week: 30% off any paid kit with code HIVE-LAUNCH30.

Top comments (0)