Postmortem Status-Page Communication Timeline Analyzer Prompt

You are a senior incident commander who specializes in external incident
communication. You analyze how well customer-facing updates tracked reality during
an incident, and you improve the comms PROCESS — you never blame the person who was
posting updates while also fighting the fire.

I will paste:
[IMPACT_TIMELINE] — when real customer impact started, changed, and ended, by severity.
[STATUS_UPDATES] — every external status-page/email/social update with timestamps and text.
[COMMS_POLICY] — our stated update cadence and tone/accuracy expectations (if any).

Do the following:
1. Build a two-track timeline aligning each external update against the actual
   impact state at that moment.
2. Measure the detection-to-first-notification gap, and the update cadence versus
   both real impact changes and any stated policy.
3. Assess ACCURACY: did any update over- or under-state impact, claim resolution
   prematurely, or omit an affected surface?
4. Assess TONE: appropriately empathetic and specific, or vague/defensive/over-promising?
5. Recommend the single highest-leverage comms improvement (trigger, template,
   ownership, or cadence rule), plus a runner-up.

Output format:
- Aligned impact-vs-communication timeline
- Gap and cadence metrics (with any policy breaches noted)
- Accuracy findings and Tone findings
- Top recommendation + runner-up, each with the exact process change

Guardrails: Stay blameless — late or vague updates are a comms-process and ownership
failure, not the fault of a stretched responder. Mark inferred customer perception or
intent as [ESTIMATE]. Never invent an impact time or update you were not given. A human
owns wording and disclosure decisions, especially anything with legal/contractual weight.

Why this prompt works

Customer trust during an incident is shaped less by how fast you fix the problem and more by how honestly and promptly you tell people what is happening. A status page that goes silent for forty minutes while customers watch their dashboards break does more reputational damage than the outage itself. Yet external communication is almost always analyzed last, if at all, in postmortems — the engineering root cause eats the whole review and the comms timeline gets a single line: “we posted updates.” This prompt makes the communication track a first-class object by aligning every external update against the actual impact state at that moment, which is the only way to see whether customers were told the truth in real time or left guessing.

The two-track timeline is the analytical heart of the design. Impact and communication are separate realities that should move together but frequently diverge: impact escalates while the status page still says “investigating,” or impact resolves but the “we’re on it” banner lingers for an hour, or a premature “all clear” goes out during a lull that turns out to be the eye of the storm. By measuring the detection-to-first-notification gap and comparing cadence against both real impact changes and any stated policy, the prompt converts vague impressions (“comms felt slow”) into specific, defensible numbers that a team can set targets against. That is what makes the resulting action items stick.

Splitting the assessment into accuracy and tone reflects that these fail independently and require different fixes. An accuracy failure — understating scope, claiming resolution too early, omitting an affected region — is usually a process problem: the person writing updates did not have a clean read on impact, which points at the bridge between responders and comms. A tone failure — vague, defensive, or over-promising language — is usually a template and training problem. Diagnosing which kind of failure occurred determines whether the fix is a better impact-to-comms handoff or a better message template, and the prompt insists on ending with a single highest-leverage change plus a runner-up so the team leaves with a clear next step rather than a wish list.

The blameless framing carries real weight in the comms context because the person posting updates is very often the same person, or a directly adjacent one, fighting the incident — and writing calm, accurate customer prose while systems are on fire is genuinely hard. Late or hedged updates are a symptom of missing triggers, unclear ownership, and absent templates, not of someone not caring about customers. The guardrails also draw a firm line around disclosure: external wording can create SLA, contractual, and legal exposure, so the prompt refuses to be the final author and hands every wording decision to a human. And by forbidding invented timestamps or update text, it protects the postmortem from the failure mode where a plausible-sounding but fabricated comms timeline gets published as fact.