The power of premonition: on the use of pre-mortems for strategy, planning and risk management

Introduction

Admittedly, Minority Report was a dystopian movie, but psychic premonition would be a fantastic management super power. We are defining and implementing strategies, embarking on new initiatives, and making tough decisions every day. What if we could know the outcome of these in advance with detailed foresight, allowing us to take a course to success every time?

Far too often, we wait for the new strategy to flop, the product launch to fail, the expansion into a new geography to be rolled-back, employee engagement to plummet despite the new ‘transformation program’, our exciting startup acquisition to not achieve the expected ‘synergies’ (was it really just an integration issue?), that major IT project to be cancelled (again), or the new executive we recruited to have to be ‘performance managed’ out of the business.

Then, if we are smart, we conduct a post-mortem, or a retrospective, to work out what went wrong and to hopefully implement some changes to prevent the same failure occurring again.

Instead, we can do a post-mortem in advance, using the ‘wisdom of the crowd’ to try and imagine and explain possible outcomes – good and bad – of whatever new initiatives or management decision we are considering. This is a pre-mortem.

The term pre-mortem was coined by Gary Klein in an article introducing the technique. I was personally introduced to pre-mortems during a course at Stanford, and I’ve had the opportunity to use the technique now a couple of times. I’ll try and share what I’ve learnt. I’ve also provided some links to further reading below, including the Klein article.

How not to do a pre-mortem

1. Don’t start with a group brainstorming

When people are introduced to the idea of a pre-mortem, they immediately lurch towards everyone’s favourite management tool: sticky notes!

Book the big conference room, bring lots of butcher’s paper, markers, and snacks. Add in a high-energy coach, some fun ice-breakers to get everyone’s creative juices flowing, and don’t forget the rubber chicken for extra points. Fun!

These kinds of idea-generation events are always engaging and they are great for techniques like design thinking, but they don’t always work well for the initial Elicitation and Analysis Phase of a pre-mortem.

One of the goals of a pre-mortem is to try and avoid the problem of group think. Starting a conversation from the perspective of failure is intended to liberate people to share their honest insights, but this is not always easy in a group brainstorming environment. Leaving the unsaid, unsaid.

Group think can be a particularly insidious problem in the case of pre-mortems because by the time we get to running the pre-mortem, the strategy, initiative, or decision in question probably has some political capital behind it, possibly at a senior executive level, and it can be hard for someone one to say "the emperor has no clothes".

2. Remember, it’s not risk management

Another trap is to think of a pre-mortem as just risk management planning by another name. The objectives and approach are actually different. The goal of risk management is to identify all of the threats (events) that could occur, and then to establish plans to mitigate these in a very structured way, often linked to compliance requirements for a company or project.

In contrast, in a pre-mortem our approach is to define only one or two outcomes (events), and to try and explain in detail how and why they might occur. And then to extract challenges/opportunities from this analysis to solve. A pre-mortem should be more strategic and holistic than the traditional approach to risk management planning, but might be a useful input into risk management (in the same way that traditional scenario planning can be an input into risk management).

3. Take the right perspective and ask why

The article that introduced the key concepts underpinning pre-mortems 2, talked about the idea of temporal perspective. This is where people will attempt to predict the future (what will happen), but will explain the past, answering why it happened.

People generate reasons more efficiently, and often in more detail, when they conceive of an outcome or event as certain (as having ‘already occured’). This concept is called prospective hindsight in the article, which showed through experimentation that this approach “increases the ability to correctly identify reasons for future outcomes by 30%”.

So rather than ask people, "what might happen?" or "what might go wrong?", the correct framing in a pre-mortem is to examine outcomes that “have already occured” (good and bad), and to ask why they occured.

The right way to run a pre-mortem

A pre-mortem needs a sponsor. Someone who is actively seeking critical input into a key question. This could be concerned with company’s strategy and direction as a whole (in which case the sponsor should be the CEO), or could be related to a new business strategy, an investment decision, the development of a new product, a major project, or some other key new initiative.

The pre-mortem process should be run by an independent facilitator. Their goal is to gain as many inputs as possible from a diverse group of people, and to provide those people with support, and with the guarentee of anonyminity, to elicit frank and fearless input.

It is very important that participation is much broader than the immediate team directly involved in the strategy or initiative in question. A successful pre-mortem will need the participation of group of people with diverse functional responsibilities, and from as many levels in the organisation as possible, and ideally from some stakeholders or experts from outside of the organisation.

The pre-mortem will then proceed through two phases:

1. Elicitation and Analysis (obtaining our visions of the future)
2. Design and Prototyping (taking action to change our destiny)

Phase 1: Elicitation and Analysis

Step 1. Prepare cases

The pre-mortem should only consider a small number of possible outcomes, typically just: ‘things go well’ (the success case) or ‘things go badly’ (the failure case), and sometimes also ‘things go strangely’ (the unexpected case).

It is possible to define multiple success or failure cases, for example a first success case where we merely meet our revenue targets, and a second case where we exceed them by 10×.

These cases are a kind of prediction that we will attempt to define in a concrete way, but it is not important that these predictions are ‘right’. They are just prompts to elicit input from our participants. Our goal is to give people the right temporal perspective and to be open to an unexpected future (in the future where we exceed our targets by 10×, why did that happen?).

For each case being considered the facilitator should create a short text to act as a prompt for the participants. The text will describe the outcome in question in an open-ended way, but that none the less makes the outcome that we are examining clear, and expresses it as one that has already occured. For example, this text could be in the form of a press-release.

Or as an alternative example, for a positive case, it could be in the form of an email to staff, such as the following.

Step 3. Request input

We should ask participants to provide their input in the form of a narrative or story. This format provides the context for each participant’s insights. For example, the instructions provided to participants might be as follows, with the appropriate text prompt corresponding to their case attached.

Step 3. Analyse input

Once written responses have been collated from each participant, the faciliator, along with a smaller team of participants has the task to analyse this input and to extract the key factors that will correspond to the challenges or opportunities to be addressed. This analysis could be conducted as a group workshop.

It is generally useful to first review the inputs to identify common themes that can be used to categorise the factors, for example, a theme might be "Competitive threats" or "Experience & capability".

Then a list of factors related to each theme should be extracted and these should be assessed as to the extent they are likely to drive the outcome in question: strong drivers are factors that are more likely to occur and more likely to influence the outcome, and weak drivers are the opposite.

An example of a factor might be "Product team lacked experience interviewing customers or conducting usability testing so customers found our mobile app hard to use".

Once this analysis is complete and compiled, it should clearly answer the question: why did we succeed (or why did we fail), and it should do it in a way that expresses a concrete set of imporant factors that drove success or failure.

For example that "we failed because customers found our mobile app hard to use, because our team lacked experience …".

This list of ranked factors are the challenges and opportunities we will take into Phase 2.

Phase 2: Design and Prototyping

This is where we begin to take our destiny into our own hands attempt to change the future.

Each significant opportunity or challenge (or set of closely related opportunities and challenges) should be assigned to a small team of 6-12 people to tackle. Their task is to create a solution to address the challenge.

There are different ways they could do this, but a design thinking-style approach would be ideal. I have (or will have soon) a couple of articles on design thinking. Design thinking is more a mind-set and set of principles than a prescriptive process, and there are many variants.

The objective of each team is do create a prototype solution that they can test and iterate on until they can validate it fully addresses the opportunity or challenge in question. This is a ‘prototype’ in the very broadest sense, and could be a process change, a new operating model or team structure, a change of strategy, a new external partnership or supplier engagement, a new tool or product, a different way of communicating with stakeholders, a new set of project priorities, etc, etc.

Each designed solution should be presented back to the Pre-mortem Sponsor and the wider group of contributing participants for their feedback.

Wrapping up

When we are embarking on new strategies and initiatives, we naturally proceed for a position of optimism and potential cognitive bias: we have worked hard on the planning, we have expended time and resources getting to this point, and we want alignment and support from our teams and from the broader organisation.

By tapping into the wisdom of the crowd and working backward from an assumed negative or positive future outcome, we can broaden our perspective and we can stimulate dissenting views on the real challenges and opportunities that will influence success. We can stress test our plans and develop and create solutions to address the challenges that emerge.

Pre-mortems are a low-cost, high-impact way to see into the future, and to improve our plans and help to ensure our strategies and major initiatives are more successful. Why not give it a try, and maybe you could be a hero, just like Tom Cruise!

Further reading

1. Perfoming a Project Premortem, Gary Klein, Harvard Business Review, September 2007.
2. Back to the future: Temporal perspective in the explanation of events, Deborah Mitchell, Journal of Behavioural Decision Making, January/March 1989.