Many users dismissed Spotify's postmortem of its third outage this month as inaccurate and misleading about detection timelines and reliability.
Based on 5 visible X reactions from 6 accounts; directional sample.
Ask a question below.
Published answers will appear here.
It took Spotify a month to publish this postmortem (after I offboarded thanks to poor reliability) and they could not get the timeline right, despite me having an email thread with the team that wrote this postmortem... The postmortem: https://engineering.atspotify.com/2026/7/content-ingestion-and-podcast-video-incident-report
@GergelyOrosz starting the clock at your own alert hides the worst number in the postmortem. detection latency is the gap between the first external report and when monitoring fired. moving the start time to the alert deletes the exact failure the writeup exists to expose.
@GergelyOrosz Gergely Orosz spotted the leak at 17:31. Spotify's engineers were still sleeping at 17:32. the alert system is slower than your morning coffee
@GergelyOrosz moving creator reports from 17 31 to 19 00 makes internal monitoring look more effective than it was
This is Spotify's version of the below timeline. The actual timeline is that "creator reports of episodes not appearing on Spotify" were at 17:31, when I emailed them, before their own alerts fired. I posted the below tweet at 17:35. I hate seeing postmortems change past dates.
The email I sent to the Spotify Podcasts team at 17:31 UTC (my timezone was UTC+2, hence the 19:31 timestamp) So yes, creator reports did not start to come in from 19:00, but much earlier.
Many users dismissed Spotify's postmortem of its third outage this month as inaccurate and misleading about detection timelines and reliability.
Based on 5 visible X reactions from 6 accounts; directional sample.
Ask a question below.
Published answers will appear here.