One of the most interesting practices at Etsy is the use of blameless postmortems to learn from mistakes and create processes that prevent them from being repeated. If I could take only one thing from Etsy and implement in other organizations, it would be the concept of a Just Culture:
Having a Just Culture means that you’re making effort to balance safety and accountability. It means that by investigating mistakes in a way that focuses on the situational aspects of a failure’s mechanism and the decision-making process of individuals proximate to the failure, an organization can come out safer than it would normally be if it had simply punished the actors involved as a remediation.
One of the most important aspect of the blameless postmortem process is that it’s an effort to overcome hindsight bias. When we look back on past mistakes, it’s all to easy to make the mistake of thinking that the information we have now is the information we had when the mistake occurred. This makes it extremely difficult to analyze the mistake and prevent similar mistakes in the future.
A big part of a postmortem is constructing a timeline of what happened. When you get to the moment when the mistake occurred, the person whose action caused the mistake is asked to think back and remember exactly what they expected to happen at the time the mistake was made, so that the gap between the expected outcome and the actual outcome can be revealed and the organization can work on preventing it in the future. The key is focusing on what information was available at the time of the incident.
Lately I’ve been trying to keep a journal of the decisions I’ve made at work and what they were based on, with the thought that when I’m trying to evaluate the quality of those decisions in the future, I’ll be able to go back and see what the rationale was at the time. When you look back three months or six months on a decision you made, it’s easy to forget what information you based that decision on, so I’m trying to write it down.
Not only do I often have to explain decisions I made to other people, but I also have to review them for my own good. We’ll see if the journal helps out.
September 2, 2012 at 7:30 am
The process you describe reminds me of the briefing-debriefing process in the Israeli Air Force.
According to the following article, is was inherited from the British and American air forces.
http://ie.technion.ac.il/Home/Users/peterb/team_reflexivity_HRM.pdf
September 2, 2012 at 9:47 am
Please update us in the future. My guess is that the decisions you think are the ones you should document for posterity in your journal will not at all be the ones that are relevant in the future. But, then that may be a sled-fulfilling situation: by going through the process of evaluating which decisions you think might be relevant in the future, you are giving more conscious thought to the ones that meet your criterie, thereby lessening the chance that you’ll make a mistake with those decisions. Heisenberg uncertainty–the act of self-observing will affect the actions themselves.
September 2, 2012 at 10:12 am
@Hanan – When John Allspaw talks about postmortems and the need to avoid focusing on attributing mistakes to “human error,” he talks a lot about airline safety, so I imagine that’s a common ancestor for a lot of this stuff.
@Stan – You are probably right. Since I’m just getting started, I’m haven’t really gotten into writing about decisions that I’m in the process of making but rather those that I made recently and people asked me about. So I guess for today I’m trying to document the past while I can still remember it rather than trying to document the process as it happens.
September 5, 2012 at 10:21 pm
This is standard practice in chemistry and biology labs where you are expected to keep a laboratory notebook handy. It’s useful for post-mortems, but also for remembering details that might some day be relevant, either for duplicating earlier work or for use by the patent attorneys. (The latter insist on numbered pages and a bound notebook so that pages can’t be removed without leaving an obvious trace.)
It’s interesting that most software people don’t do this. I never kept a formal lab notebook, but I used to keep piles of paper notebooks which are still full of surprising ideas, some still interesting, some idiotic. Of course, now I keep dozens or maybe hundreds of little text files around and I rarely delete anything I just slide stuff down. Of course, Lion is keeping older version for me, but old habits die hard.