The post-mortem

It's typical nowadays to hear about investigations when big mistakes are made by public and private organisations.

This is an exercise we can apply to our own lives when things go wrong and our decisions could have contributed. A post-mortem is a tool we can use to improve future decision-making, examining failures but also successes.

It can not only help us understand and prevent repeats of mistakes in the context being examined, but across other domains too. For instance, if we identify a fundamental issue with the way in which we assess the balance of probabilities (let's say we are too optimistic and under-weight negative outcomes), this learning can be applied broadly across our life.

I planned to write this edition from a plane to Madrid. Instead, I write from London with the pain of missing out on significant plans with friends and family. The simple explanation is my passport wasn't returned on time from an embassy processing a business visa for a different upcoming trip.

However, when something doesn't go to plan and the consequences are significant, I look to see where my own decisions could have been the root cause or exacerbated the consequences or costs. This note is my own post-mortem.

My own personal experience

Plan + decisions: Attend embassy appointment on 13th Sep, pay for express turnaround. Passport due back 15th or 16th, and fly to Madrid on 16th or 17th for two weeks. Decide to not pay for flight until I get passport back due to the uncertainty over which of the two days it will arrive on.

Outcome: At appointment am told express turnaround is unavailable, and I will get passport back at 430pm on 22nd. Very annoyed as missing a trip, but no financial cost.

New plan + decisions: Plan to fly on 23rd, but when looking at flights find 23rd is expensive and 22nd and 24th offer better value. Given trip already been cut short, prefer 22nd to maximise time abroad. Will be cutting it fine with a pickup at 430pm for a flight at 745pm (airport 1 hour away from pick-up point). See key risk as a delay in pickup, but estimate that with up to a 45 min delay I can still make it. This time book flight in advance, with rationale that the cost of flight upon passport pickup 3 hours before take-off will be prohibitively high, and will also be too late for online check-in, so likelihood of making flight falls considerably when factoring in check-in desk queue. At this point see risk of passport not being delivered at all on the 22nd very low, given the amount of extra time the embassy has given itself to clear the backlog.

Outcome: Get indication by email at 1pm that there are processing delays. Arrive at 430pm to pick passport up with suitcase in hand. Told at 440pm that passport is not ready, with no indication of what day in future it will be. Flight lost. Arrive home an hour later and attempt to change flight, but am unable to as had checked-in already earlier.

(Additional context: I unexpectedly found myself moving house the same day of the anticipated flight. This necessitated being up at 430am on 3 hours sleep, finishing packing a house, moving house and then making it to the embassy in time for 430pm, whilst simultaneously completing a busier than usual work day. Time-pressed is an understatement)

Key failings

Not conducting a pre-mortem weeks in advance - I didn't spend much time considering the possible outcome of not getting my passport back at all. Considering this may have affected my decision about organising the trip, buying the ticket before receiving the passport, the date of departure, and a decision to check in on the morning of the flight.

Gambling without properly considering the odds - I gambled that the passport would be ready on the 22nd without considering the odds. I hoped for the best which clouded my assessment of the non-zero likelihood, and I didn't gather additional information to try and estimate it (e.g. speaking to others who had gone through process before, reading forum discussions). On the basis of this, the 24th would have been more sensible departure date, as it would allow for a one day delay in getting the passport back (FYI, it also wasn't ready the following day).

Decisions when tired, stressed and under time pressure - A key failing which wouldn't have changed the outcome, but increased the financial impact, was the decision to check-in online, which I did at 11am on the morning of the flight. My plan had been to wait until I'd made a call reminding the office that I would be arriving at exactly 430pm, but I had been on hold for 15 mins and cancelled the call to join a meeting. At this point I was starting to get nervous that I was falling behind schedule, and feeling the effects of tiredness and lack of food that morning. With a couple of minutes gap before the virtual meeting, I checked-in. In more optimal conditions, I would have considered the value of waiting until later in the day, and likely decided that the risk of possibly being slow with airport check-in if I couldn't do it online, was worth it.

All of this has been taking place against the backdrop of an extremely busy and stressful couple of weeks in my life. Taking the full day off from work would have given me a clearer mind and wouldn't have necessitated the lack of sleep.

Learnings

There are multiple learnings from the post-mortem which can be codified into rules. I'll state one context specific one, and one generalised, fundamental learning which overlaps with other domains, and let you identify the others.

Context specific - Don't plan a trip whilst not in possession of a passport.

Generalised - Assign numerical probabilities to events where there is uncertainty, and only take actions upon a consideration of those estimates. Whilst the probabilities of a delay or not getting the passport back would be (very) subjective, in the absence of them my hope and (over) confidence that I would get the outcome I wanted clouded my judgement. Even rudimentary estimates written down would likely have made me pause and re-think certain decisions.

Implementation idea

The idea is to simply consider this exercise in your own life when things do not go according to plan.

In my opinion, generalised learnings are the most useful. Context specific learnings are valuable if these are situations we expect to find ourselves in again. However, many decisions are one-shot games where the stakes are high and we do not have the luxury of previous experience.

If we look for generalised learnings to create rules we can apply across domains, we can find ways to improve decision-making in future scenarios that we face for the very first time.