Observability and reliability
Blameless culture
Blameless postmortems ask which conditions made a failure possible, not who to blame. Accountability for fixes stays in place.
A blameless postmortem assumes everyone acted reasonably, with the information and tools they had at the time. The question is not "who broke production?" It is "which design, process, tooling and incentives made this failure possible, and how do we change them?" Blameless is an information strategy. It is not the same as soft.
The big idea
People who expect punishment hide near-misses, tidy up timelines, and blame "user error" for gaps in the system. People who expect curiosity bring the whole story, including the awkward parts. Only the whole story can be fixed.
Aviation and medicine learned this long ago. If reporting a near-miss ends a career, the reports stop. Google's SRE book makes the same point for software: a culture of finger-pointing hides the data you need.
Concept
Blame assigns moral fault to a person. Accountability assigns ownership of a fix. Blameless culture drops the first and keeps the second. Someone still owns each action item, and someone still checks that it is done.
Walking through one sentence, two ways
Here are the same facts, written two ways.
- Blame: "Alice deployed without reading the runbook and took down payments."
- Blameless: "A config change shipped through the standard path, with no canary on the payments shard and no automated check for the missing feature flag. The deploy looked routine in CI, and on-call had no signal until the error rate crossed the page threshold."
The second version names gates you can build: canary coverage, flag validation, and safer defaults. The first version names nothing you can change.
What blameless does not mean
Blameless does not excuse deliberate sabotage, or ignoring known safety rules after coaching. Those are management issues, not postmortem issues. Most outages are different. Competent people were running a system that made the wrong thing easy.
Common mistakes
- Blameless wording that still shames. "An unfortunate oversight by the team" is blame with better manners.
- Blameless, but nothing changes. Blamelessness without accountability is a shrug.
- Reading "no blame" as "no owner." Each fix needs one owner.
Check yourself
- What is the difference between blame and accountability? Answer: blame assigns moral fault to a person. Accountability assigns ownership of a fix.
- Rewrite "Alice skipped the runbook" in blameless form. Answer: "A config change shipped with no canary and no automated flag check, so the gap went unseen." Name the gate, not the person.
Key takeaways
- Blameless culture gets honest data. Accountability gets changes made. Keep both.
- Ask which gate was missing, not who approved.
- Reckless behavior is a management matter, not a postmortem matter.
- Next, the summary and timeline show how to write the record.
Further learning
Google's SRE book chapter on postmortem culture covers what belongs in the document and why.
Postmortem Culture: Learning from Failure (Google SRE book)Blameless postmortems in SRE culture. Incident study that focus on root cause analysis and preventive actions, for culture of continuous improvement.sre.googleGaurav Sen's walkthrough explains why blameless culture exists and how it plays out after a bad change ships.
Related
Blameless postmortems
Learning from failure with timelines, contributing factors and action items that actually get done — without turning people into the root cause.
Owning your code in production
On-call, alerts, runbooks, and the loop that makes you trustworthy — and better at building systems that survive users.
Written and verbal communication
Lead with the point, write for skimmers, and adjust depth for engineers, PMs and executives.