A postmortem only works when nobody is being blamed
A note on Postmortem Culture: Learning from Failure by Google SRE, published on sre.google.
The argument is that people hide information when they expect to be blamed, and hidden information is exactly what you need after an incident. Remove the blame and you get the truth, and the truth is what lets you fix the system rather than the person.
What makes the chapter useful rather than motivational is the specificity about the artefact. A good postmortem has a timeline, an honest impact statement, a root cause that goes past "someone made a mistake", and action items with owners. It names the conditions that made the mistake likely, the misleading dashboard, the deploy process that made it easy, instead of stopping at who typed the command.
Solo consultants tend to skip this on the grounds that there is no team to be blameless towards. That misses the point. The written record is the deliverable, and it is worth more than the fix: six months on, the fix is in the code and the reasoning is nowhere, which is exactly the situation that produced the incident in the first place.
The rest of that book is free online and most of it repays reading.
This is the thinking behind monitoring the paths that carry your business.