Overview

When something goes wrong, a clear failure analysis presentation turns a setback into organizational learning. The pptx.dev API's failure-analysis narrative structures your post-mortem into a professional deck.

The Failure Analysis Structure

1. Incident — outline the failure timeline and impact
2. Root Cause — identify root causes with evidence

Example: Service Outage Post-Mortem

{
  "title": "March 15 Outage — Post-Mortem",
  "narrative": "failure-analysis",
  "colorScheme": "slate-gray",
  "fontScheme": "calibri",
  "slides": [
    {
      "layout": "title-center",
      "content": {
        "heading": "Service Outage Post-Mortem",
        "subheading": "March 15, 2026 — 14:30 to 17:45 UTC"
      }
    },
    {
      "layout": "number-3x-title-center",
      "content": {
        "heading": "Impact Summary",
        "items": [
          { "value": "3h 15m", "label": "Total Downtime" },
          { "value": "12,400", "label": "Affected Users" },
          { "value": "$45K", "label": "Estimated Revenue Loss" }
        ]
      }
    },
    {
      "layout": "list-4x-vertical-title-center",
      "content": {
        "heading": "Incident Timeline",
        "items": [
          { "heading": "14:30 UTC", "body": "Monitoring alerts fire for API latency spike" },
          { "heading": "14:45 UTC", "body": "On-call engineer begins investigation" },
          { "heading": "15:30 UTC", "body": "Root cause identified: database connection pool exhaustion" },
          { "heading": "17:45 UTC", "body": "Service fully restored after connection pool resize and deploy" }
        ]
      }
    },
    {
      "layout": "list-3x-title-center",
      "content": {
        "heading": "Root Causes",
        "items": [
          {
            "heading": "Connection Pool Limit",
            "body": "Default pool size of 20 was insufficient for new traffic patterns after product launch."
          },
          {
            "heading": "Missing Auto-Scaling",
            "body": "Database connection pooler had no auto-scaling configured."
          },
          {
            "heading": "Delayed Alerting",
            "body": "Latency alerts had a 15-minute delay due to aggregation window."
          }
        ]
      }
    },
    {
      "layout": "list-3x-title-center",
      "content": {
        "heading": "Action Items",
        "items": [
          { "heading": "Increase Pool Size", "body": "Raise default to 100, add auto-scaling. Owner: Platform team. Due: March 22." },
          { "heading": "Fix Alerting", "body": "Reduce aggregation window to 1 minute. Owner: SRE. Due: March 19." },
          { "heading": "Load Testing", "body": "Add load test for 3x traffic before every launch. Owner: QA. Due: March 30." }
        ]
      }
    }
  ]
}

Tips for Post-Mortem Decks

1. Be blameless. Focus on systems and processes, not individuals.
2. Quantify impact. Duration, affected users, revenue loss — concrete numbers drive urgency for fixes.
3. Use timelines. A vertical list layout shows the incident progression clearly.
4. End with action items. Every root cause should have an owner and a due date.
5. Use slate-gray — it is professional and serious without being alarmist.

Next Steps

• Explore strategic advisory decks for recommendations
• Add charts with data visualization
• Browse all narratives