The system is down, the team is busy fixing it, and customers are writing to ask what is happening. The most common mistake is not a wrong message. It is no message at all while everyone is looking for the cause.
A few decisions made in advance let a small team communicate calmly in the first hour.
Decide who approves and who sends
During an outage there are two jobs besides fixing the problem: writing to customers and approving what is written. Give both a name before anything breaks.
- One approver, usually the person responsible for the service, and a deputy.
- One sender who posts to the status page, e-mail or chat, so customers do not receive three different versions.
- One internal channel where facts are collected, so the sender does not have to interrupt the people doing the repair.
If approval takes longer than ten minutes, the message is already late. Agree that the approver may delegate wording and only checks facts.
Send the first message within the hour
Do not wait for the cause. Customers need to know that you know. A first message can be five lines:
Subject: Service disruption – we are working on it Since 09:40 the customer portal is not available. We are aware of the problem and are working on it. Orders already placed are not affected. We will send the next update by 10:30 at the latest. Contact: support@example.com.
It says what is affected, since when, what is not affected, when the next update comes and where to ask. Nothing in it is a guess.
Use a few fixed status words
Choose three or four words and keep their meaning stable:
- Investigating – we know about the problem, the cause is unknown.
- Identified – we know the cause and are working on a fix.
- Monitoring – a fix is in place and we are watching it.
- Resolved – the service works normally again.
Customers learn to read them, and your team does not argue over wording in the middle of an incident.
Keep a steady update rhythm
Agree the interval at the start, for example every 30 minutes for the first two hours and hourly afterwards. Send the update even when nothing has changed: "No new information; next update at 11:00." A predictable rhythm reduces incoming questions more than any long explanation.
Facts first, apology short
Put facts at the top: what is affected, since when, what you are doing. Add one honest sentence of apology and stop there.
- Do not guess the cause or name a supplier until you are sure.
- Do not promise a time you cannot control. Promise the time of the next update instead.
- Do not blame a customer, a supplier or a colleague.
- Say what customers can do meanwhile, such as a phone number or a manual way to place an order.
If your service relies on an AI tool, the same rules apply; see what to promise customers and the team when an AI service fails.
Write a follow-up note after recovery
Once the service is back, send a short note the same day or the next morning:
- When the problem started and ended.
- What was affected and what was not.
- What you are changing so it is less likely to happen again, with a date for the next step.
Keep it to a few lines. If customers lost data or money, say how to claim or check it. The internal analysis is a separate document; a one-page blameless review works well for it.
Common mistakes
- Waiting for the cause before saying anything.
- Different messages from sales, support and the owner.
- Silence between updates because "there is nothing new".
- Long apologies and short facts instead of the other way round.
- No follow-up, so customers never learn that the problem is closed.
How to start
- Write the first-message template today and save it where the whole team can find it.
- Name the approver, the sender and their deputies.
- Try the template in a ten-minute rehearsal and shorten what nobody needed.
Summary
In the first hour of an outage, customers need to know that you know, what is affected and when they will hear from you again. A prepared template, fixed status words, a steady rhythm and one approver make that possible without extra effort during the incident itself.
Sources
- Our related articles linked above.
- The approach is common practice for small service teams and is offered as a practical guide. No client results or market figures are claimed in this article.
