Incident response time (TTR) is inherently unpredictable and rarely under statistical control, as demonstrated by a control chart analysis of Cloudflare's incident data from 2025. Filtering out irrelevant data, the analysis showed significant TTR variations, particularly for complex incidents. The unpredictability of incidents makes metrics like MTTR ineffective, emphasizing the need for continuous improvement in incident response without expecting full control over the process.
https://surfingcomplexity.blog/2025/11/27/fun-with-incident-data-and-statistical-process-control/