How to prepare an incident status message
Write an incident status message that explains confirmed customer impact, what has changed since the last update, and when customers will hear from you.
Free app. Runs on your Mac with the Claude or Codex plan you already have.
What an incident status message is
An incident status message tells customers what a service problem means for them, what has changed since the last public update, and when they can expect to hear more. If report exports are delayed, describe what customers see when they request one.
In outage communication, the response team establishes what happened and a qualified incident communications owner approves what can be shared. Use the incident timeline to verify public claims, then prepare status page update drafts and a support reply. Keep a publication record so the next writer can build on what customers actually saw.
What it does for the business
- Help customers decide what to doCustomers can plan around delays when they know which tasks are affected and when to expect more information. An approved workaround lets eligible customers keep working.
- Give support a current answerA short approved reply gives agents language they can use when a ticket matches the incident scope. If a report falls outside that scope, they can send it to the incident lead with the missing detail instead of automatically assigning it to the known problem.
- Make approval decisions easierA reviewer can focus on open decisions when the evidence is easy to find. For example, they can check whether a proposed workaround is ready to share and carry forward approved language whose evidence still holds.
- Carry the update across a shift changeThe next owner can see what customers were told and when they expect another update. They can refresh the draft using the recorded sources and open questions.
How to build an incident status message, step by step
Find the last published update and set a cutoff
The next message starts with the last update customers could read.
What to do
- Record the incident identifier, affected service components, regions or environments, and the customer tasks reported as impaired. Name the communications owner and who can confirm changes in state.
- Record the last published wording, its publication time and promised checkpoint. Give the owner any conflicting wording from the status page or support replies.
- Set an explicit cutoff with a time zone. In your notes give each event its observation time and the time the report arrived. An account report received just now may describe a failure from an earlier window.
- Label each source as an observation, customer report, working theory, or approved public statement. If sources disagree, ask the person responsible for the relevant evidence to confirm the scope and time before using the disputed claim.
Why
A clear cutoff tells the reviewer which facts belong to this update. The last published message shows which statements may need correcting and which commitments customers are still waiting on.
Decide what the evidence supports
Use the incident state you can support at the cutoff.
What to do
- Create a claim register, a list of the statements you might publish. Record evidence, observation time, who can see the information, and proposed wording for each claim. Express customer impact as a delayed, failed, or unavailable action within the confirmed scope.
- Use Investigating while cause or work is unresolved and Identified when the cause is confirmed. Use Monitoring when responders are observing the effects of an evidenced mitigation. Use Resolved after the authorized owner confirms the agreed recovery checks and addresses any remaining impact.
- Ask the service owner to verify recovery across the affected customers and time window. A successful export confirms that request succeeded. To make a broader recovery claim, also check the affected export queues and any regions still showing delays.
- Use the owner's promised update time or the incident update cadence from your incident communication plan. If the deadline passed, flag it to the owner and draft an update with the current facts. If there is no commitment, propose a checkpoint for approval.
- Include a workaround only after its owner approves the exact steps and limits. Send unresolved questions to the named owner, especially when a draft would otherwise imply a restore time, data-loss assurance, or security conclusion.
Why
A mitigation can improve new requests while earlier requests remain queued. Describe both facts so customers who still see delays can tell whether their experience fits the update.
Draft the status update and the support reply
Every channel should tell the same story about customer impact.
What to do
- Start the status page message with a clear title and supported state. Explain who is affected and what customers are seeing. Write the affected action in terms customers recognize from your product or service.
- Explain the material change from the last published message. For example, if new exports are completing and earlier requests still face delays, say both when the evidence supports them. If a checkpoint arrives with unchanged findings, say the investigation continues.
- Write the next update time and time zone as a clear commitment after owner approval. Add exact workaround instructions when approved, including any affected group for whom the workaround is unavailable.
- Write a concise support reply a person can use in a conversation. Lead with the issue reported, relay the same scope and next update time, and keep any approved instructions consistent with the public page.
- Keep earlier drafts labeled as historical and keep internal evidence and owner questions outside the public text. For an open incident, list the recovery checks still needed before a resolution message could be written.
Why
A public update serves readers with different symptoms, while a support agent answers a specific report. Consistent scope and timing help customers recognize the same incident across channels.
Check the public copy against its sources
Check the exact words customers will read before the draft leaves your hands.
What to do
- Read each draft sentence next to its entry in the claim register. Confirm service names, scope, timing and promised checkpoints. Pay attention to broad words such as all and fully, which can turn a narrow observation into a wider claim.
- Compare the draft to the prior published copy and check each channel against the others. If you are narrowing a previously published impact statement, confirm that the revision is supported and that readers can tell what changed.
- Ask a second reviewer to check the public text and the claims behind it when one is available. Check that the copy excludes customer records, credentials and private investigation details. Resolve or qualify uncertain claims and retain the questions for the owner.
Why
A regional improvement can read like service-wide recovery if you drop the region name. Comparing the wording with the evidence catches that scope change.
Give the owner the current message and open decisions
The owner needs the draft, its cutoff, and the outstanding decisions in the same handoff.
What to do
- Identify the current copy-ready draft, its last evidence cutoff, its proposed publication owner, and the exact decision still needed if publication is on hold. Put the evidence link alongside each question so the owner can reply directly.
- Check for material changes with the incident lead before publication. Refresh the draft and its promised checkpoint if you learn that scope, mitigation, or the owner's expectations have changed.
- Record the text, time, channel and commitments after publication. Pass that record and the outstanding checks to the next owner.
Why
Accurate text can age quickly. The cutoff tells the owner where to resume, and the publication record gives the next responder the same baseline customers have.
Common mistakes
- Using the newest report as the current stateCheck when the reported event happened before using it to change the public description.
- Treating deployment as proof of recoveryUse the agreed recovery checks and obtain the owner's confirmation before publishing a resolution.
- Publishing a speculative causeKeep working theories in responder notes until the cause is confirmed and approved for public use.
- Letting reply templates driftReview queued support replies when the scope, workaround or checkpoint changes.
- Moving a missed deadline without explanationTell the owner the checkpoint has passed and prepare a truthful current update using the current time.
What a good result looks like
- A customer can identify the affected action and confirmed scope.
- The state, progress and remaining impact agree with the supporting observations.
- Each channel carries the same facts and next checkpoint, with the time zone stated.
- Each material change from the last published update is clear.
- The owner can identify the latest draft, its cutoff and the decisions still open.
And then what? Using the result
- Give the qualified communications owner the current draft and open decisions.
- After approval, have the authorized publisher post the update and give support the matching reply.
- Record the published copy and assign the promised checkpoint to the person covering it.
- Use each new checkpoint to verify changes in customer impact and revisit any withheld recovery claim.
- After closure, review missed checkpoints and revise the communication plan where the handoff broke down.
When to run it
- When an incident first affects customersDraft an initial update from the authorized incident report so customers know what you are investigating and when to expect the next message.
- When a mitigation starts to show resultsExplain which affected actions are improving and which customers still face delays while the response team verifies recovery.
- When support reports symptoms outside the published scopeBring those reports to the incident lead for verification and prepare revised public wording if the evidence changes the confirmed scope.
- When responsibility passes to the next shiftHand over the last published wording, the draft in progress, and the time of the next promised update.
Questions people ask
How do you write an incident status message?
Start from the last published update, verify current customer impact, and draft the supported state, material change and next checkpoint. Have the communications owner review the copy before it is published.
What should you say if the cause of an outage is unknown?
Say that you are investigating and share the customer symptoms and scope you can verify. When an authorized incident report establishes the issue and impact is still being assessed, describe that uncertainty plainly and give the next approved update time.
How often should you update customers during an incident?
Follow the communications owner's commitment and give a time zone for the next checkpoint. Publish an update when it is due even if the message is that the investigation continues, and tell the owner when an earlier checkpoint has already passed.
When should an incident update say resolved?
Use Resolved once the authorized owner confirms the agreed recovery checks and reviews any remaining impact. If exports remain queued, have the owner assess closure against the recovery criteria and state any remaining impact.
How should a support reply differ from a status page update?
A support reply should apply the same confirmed facts to the problem the customer reported, while the status page covers the current incident as a whole. Give the same checkpoint and approved workaround in both.
What if the previous public update overstated the impact?
Correct the affected claim in the current update using the latest verified scope. Record the earlier wording and the evidence for the correction internally so that later writers can explain the change consistently.
More jobs in Developers
Prepare incident status messages, on your own Mac.Free app. Runs on your Mac with the Claude or Codex plan you already have. Your files stay on your computer.
Download for Mac