How to predict the future by pretending it's the past
When facilitating workshops for projects doing a Risk Pre-Mortem has got to be one of my favourites. I like it because it makes one of the traditionally more difficult and dry subjects (risk management) easy and fun and gets very high quality information out of a project team. It’s a workshop that I highly recommend carrying out as part of the kick-off activities at the start of any project. It’s based on the psychology of how we make decisions and the structure has been tried and tested on numerous projects. Its an idea I came across and developed into a practical workshop format that not only helps to create team buy-in but generates top quality information to put the project on the best foundation possible. It’s also a fun workshop to do.
When starting up any project, it is generally considered good practice to put together a ‘risk register’ where we look ahead to foresee all of the things which could go wrong, estimate their likelihood and impact and take preparatory steps to deal with them before we begin. A good idea, but where do the entries on a risk register come from? How do you get them? And most importantly, why do we think that this information is accurate?
As a project manager, you might try thinking about the risks yourself. Not the best option but better than not doing anything. You might ask your project team. A good start but what will you ask them; ‘give me a list of all the things that might go wrong in this project’? Sounds obvious but its probably the worst way of getting what you want.
Asking people about the future makes them stressed and disempowered.
Being asked to think about the future and predict what might happen (in relation to something new) is inherently stressful and disempowering.
It’s stress inducing because the future is unknown and we much prefer certainty. The future often seems scary compared to the known and rosy past. There can also be any number of work related stresses or worries which become triggered when we are asked to look into the future in relation to our work.1
It’s disempowering because predicting the future is something we are inherently bad at. We are influenced by all manner of unconscious cognitive biases and our performance is rarely if ever given any rigorous checking, testing or feedback. The further away in time we are asked to predict, the less accurate our ideas become. Most fundamentally, the future is outside of our control (though we can influence it, sometimes). Psychologically we know this, it is the untameable unknown and our awareness of this does not create in us a feeling of power or of being in charge.
So if we ask a project team to start thinking about the future and what to do about it, we place this group of people into a stressed and disempowered state, tasking them with doing what they do not like and cannot do well. Not the best psycho-emotional state from which to get quality information.
A further, more social disadvantage of asking people directly about risks is that suggestions are likely to involve someone present not doing what they should have done so may involve either a) blaming someone else or b) someone giving themselves extra work. Either way, there is a disincentive to speaking up and speaking honestly. There might also be a cultural inhibition in the organisation about voicing concerns or worries due to being perceived as being negative or unsupportive.
Let's say you press ahead anyway and the group come up with a list of suggestions. The traditional method piles on the agony by asking this group to estimate the likelihood of each event happening. You might use a scale such as from 1-10. Is this event as likely as a 7 or an 8? This one a 3 or a 4? Do you know the difference? I never have. (There are two great articles on the futility of this method here2 and here3 )You might then ask people to do the same with the likely impact. By this point, the answers you are getting are pretty much pure conjecture. If you then ask for mitigations and then repeat the above estimations for the impact and likelihood after your theoretical actions have theoretically been implemented, you are essentially putting a guess upon an estimate and have gone through the wardrobe and stepped into the land of Narnia. No wonder most people don’t like doing risk estimations and why we all generally are not terribly good at it. That’s a perfectly rational reaction. But there is a better way.
Forget the above and pretend the future is the past.
Instead of thinking about the future, a great way to surface potential risks for your project is to carry out a project ‘pre-mortem’ workshop. A post-mortem is something carried out after the patient dies, to find out what went wrong. A pre-Mortem on a project is carried out before the project starts, to find out why it died. That sounds strange; it works like this.
In a pre-mortem we (you, the project team and anyone else invited), imagine that we are in the future, looking back at the project. The project has finished and it was a complete, unmitigated disaster. It was awful. We are independent of the organisation and the project and are compiling the final report that the organisation can learn from. We have perfect information on the project; everything that did and didn’t happen and why. Our job is to compile what went wrong and recommend what the organisation should have done to avoid the series of missteps and disasters.
We’re doing the same thing as before aren’t we? So why does this simple imaginative exercise make such a massive difference (and trust me, it really does)? The answer lies in our psychology and the cognitive trick that this exercise pulls on us.
The science bit
The background to a pre-mortem dates back to research done in 1989.4 It found that “prospective hindsight—imagining that an event has already occurred—increases the ability to correctly identify reasons for future outcomes by 30%”.
Karl Weick5 argues that it works because of a cognitive quirk: “we find it easier to imagine the detailed causes of a single outcome than causes of multiple outcomes” and I think that this gets to the core of why taking a ‘pre-mortem’ approach works so well. Reframing the future as the past frees us to think more creatively and confidently. The work and idea was then developed further by the psychology Gary Klein, who specialises in how we make decisions. Klein says that doing a pre-mortem exercise “liberates people who might otherwise be afraid of looking like they’re not a team player. "Now, everybody is being asked to think about failure. So instead of looking like a bad teammate, you’re pulling in the same direction as everyone else”6.
Doing this exercise brings a further, project-lifetime benefit. It sanctifies being negative as a positive. Instead of looking like a ‘moaner’, it reframes looking at potential problems in a positive, creative and fun way, which means that people get used to thinking about risks and pitfalls and have a context for voicing them which is entirely positive.
The Method: Ideas and Refining
So how do we do this in practice? Within the context of a group knowing the above, start your pre-mortem by asking your group to record all of the things which might have gone wrong with this project, from their future perspective. This uses their inside knowledge of the organisation and environment (as well as any subject specific expertise), but in a blame free way. When a big initial list of everything has been created, you can then move on to sorting and refining.
Initially ask people to start categorising these items, in turn, into Wildcards (ones the group thinks they have little control over), Likely ones, and finally Serious ones. This bit I have copied from an article on the ‘riskology' site 7 but I have reversed the order. This is good for stimulating discussion while avoiding the ‘is this a 7 or 8 likelihood’ type debates.
At this point, your wall will likely be covered in a lot of sticky notes. They’ve been sorted into categories but probably there are still too many to deal with in any detail. We need some way to further refine these ideas and draw out what truly are the main concerns of people. The next stage then is to prioritise. Get people to pick the top three, then the next three and then another three. It’s explained by telling them that the organisation they are doing this for has very limited resources, so what will they put into their report if the organisation could only deal with three areas - what would they be? This really helps people to focus on what they as a group think is most important. When they have done this, tell them that some more resources have been found and they can pick 3 more. Then some more resources can be found and another three can be picked. Leaving you with nine prioritised risks.
The final stage gets people into smaller groups to work on a number of these prioritised tasks between them, the exact number depending on the group size. Each group is tasked with answering four questions for each risk: a) what might cause it to happen, b) are there any early warning signs that it might be happening, c) what might prevent it from happening and d) what might lessen its affect if it did happen. The group is also requested to put down who should ‘own’ this risk by stating the person who would be most qualified to confirm that a satisfactory plan for this risk is in place.
At the end of this, we have nine top risks, chosen and worked on by the project team itself. If you don’t think that the nine chosen are enough, then the workshop can be repeated focussing on others in the list until everyone is satisfied that everything needed has been covered.
I wouldn’t want to start any project without doing this. The best effect of these kind of activities isn’t just that the group has ownership and ‘buy-in’ but that it has understanding of what might be ahead of them and what to do about it. It’s not the PM’s risk register it’s the project team’s risk register. It is this that will help to generate quality information not just from the workshop but over time as the project rolls on. Worth doing for that alone.
SwiftWorks provides specifically designed workshops to help support projects and service reviews. Download a free workshop plan for this pre-mortem workshop.
You can download a pdf of this article for reference here: http://bit.ly/PreMortemRiskManagement
1 For the purposes of this article I’m talking about the ‘general world’ not mathematical predictions of regular systems. We know where the sun will be in the sky in six month’s time because it is a simple and repeating pattern, though it is still predicting the future.
2 Problems with scoring methods and ordinal scales in risk assessment.
3 Risk Analysis and Ordinal Risk Rating Scales—A Closer Look.
(scroll down almost to the bottom for this quote)