The Pre-Mortem: sidestepping disasters before they happen.

Projects fail for many reasons, even after enormous effort and planning. Pre-mortems can help avoid disaster.

Let’s suppose your team needs to decide whether to pursue a project (investment related or not), and it’s time to discuss the risks. What’s the best way to do it? Gary Klein, a research psychologist and currently Senior Scientist at MacroCognition, found that ‘prospective hindsight‘ — imagining that an event has already occurred — increased the chances of identifying the reasons for failure by 30%! In any risk management context, that’s worth learning more about.

Enter the Pre-Mortem.

We’re all familiar with the medical post-mortem in which a medical examiner determines the cause of death. Thanks to my wife’s penchant for all those crime investigation TV shows, I’ve seen my share of Hollywood post-mortems (I really miss Quincy). I’ve also been involved in my share of post-mortems in the business world: examining why a project failed after the fact. With 20/20 hindsight, we figure out where we went wrong – the idea being that we won’t repeat the same mistake twice. The pre-mortem takes this a big step further: it takes advantage of behavioral economics’ lesson that our brains treat the past very differently from the way they treat the (hypothetical) future. It does this by having us to role-play that the project already failed. This is the crucial part of the pre-mortem: that we examine the project with simulated hindsight.

The idea is simple: instead of asking “what could go wrong”, the group leader asks “what did go wrong?” That subtle change makes all the difference. The project team gathers in a room and everyone is asked to take out a sheet of paper. She instructs us – “imagine we’ve gone ahead with the project and we’re one year down the road. The project was such a disaster that each of us is immediately embarrassed if the project is mentioned. We don’t talk about it. Ever. Now, without discussing with others, write down your top 2 or 3 reasons why the project failed.” After everyone has written down their reasons, the group then reviews all the ideas and discussed the causes, risks and prevention methods for each of these scenarios. There are two very important parts to this process:

1. Prospective Hindsight. The brain treats time asymmetrically: we do not think about future hypothetical events the same way as we treat the very concrete past. In the pre-mortem, we take advantage of this fact by supposing the project went ahead and we’re now looking back – into the ‘concrete’ past. By doing this, we apparently come up with more and better examples of what ’caused’ the failure. The brain naturally comes up with causes to the fact (rather than hypothetical possibility) that the project failed. For this reason, it’s very important not to set up a pre-mortem by asking “what could go wrong.” To be effective, it has to be set up as an event that has already taken place and we’re looking back on it. Once we’re in this mode, our brains come up with more examples of what (could have) caused the disaster. Our almost insatiable need for a causal explanation of events is put to a positive use: by simulating that an even has taken place, our brains naturally fill the void with reasons it happened. Our brains do not do this with the hypothetical future, only with the concrete past.
2. Independence From Authority. It’s also important that the team members write down their ‘nightmare retrospectives’ independently of one another. If, instead, everyone discussed their pre-mortem ideas one by one, then the more junior people would be influenced by what the more senior people say, everyone will be disproportionately influenced by the more attractive people and by those with deeper voices, etc. Behavioral studies are clear that this isn’t just a hypothetical possibility: certain random characteristics are more influential than others, and all people are affected by these biases. Having everyone write down their answers independently will limit, but unfortunately not eliminate, the effects of these biases.

In addition to the Pre-Mortem, which looks at risks, I think this can and should be extended to looking at possible reasons why the project might have been a resounding success that weren’t necessarily anticipated when the project was envisioned. Even though the original idea of the pre-mortem was to help overcome the ‘optimist’ effect in which we humans sometimes have a “damn the torpedoes, full speed ahead” attitude about projects, the reasoning behind the two critical ingredients listed above are just as sound for a ‘Pre-Success’ analysis. In this case, the idea is to come up with reasons it was a super success, in hope of finding good strategies to pursue while the project is ongoing.

How to conduct a Pre-Mortem or Pre-Success:

1. In-person meeting of all the stakeholders. The leader sets the stage so that everyone can write their answers independently. Leave enough time for creativity. Once the team is experienced at the Pre-Mortem/Pre-Success, you may have them bring these written reasons for Failure/Success to the meetings, but be sure to have them do so independently.
2. Public collection of reasons. One scribe stands the the white board / easel and records the various reasons so everyone can see all of them.
3. Analysis of preventative measures / Ways to ensure it happens. For the pre-mortem, discuss what can be done starting now to prevent the catastrophic failure. For the Pre-Success, discuss what can be done at this point in time to increase the chances of realizing benefit. For both of them, the actions to be taken are probably not one-time-events, but rather sustained efforts to prevent failure/ensure success. You should leave plenty of time for this step in the process, and come back to it if need be rather than rush it in one meeting.
4. Records your Analyses for Future Projects. In fact, the end of step 2 should include a review of previous pre-mortems (and post-mortems!) to check if any of those causes should be added to this project’s. This is the ‘learn from your mistakes’ step that we often neglect.
5. Go/No-Go. The point of the pre-mortem/pre-success is to help decide whether the project is worth doing. If it’s a no-go, there’s not much more to do. If it’s a go, then the analyses from the pre-mortem/pre-success should guide the project with “critical failure points to avoid” and with “critical success factors”.

After years of reading about cognitive traps because of our need for causal explanations, I’m quite pleased that the idea of the pre-mortem is gaining ground. This is Behavioral Economics put to practical use.

Have any of you tried a Pre-Mortem? If not, does it sound like something you might try? What have been your experiences or what would make this do-able? Please share your thoughts.