When something goes wrong in the tech world, it is common practice to have a Post-Mortem. The Post-Mortem session aims to look back at what happened, identify one or more root causes, and determine a number of actions to take to prevent the same event from happening again. There are good practices around it to avoid finger pointing and instead focus on finding a solution and moving forward. But in any case, it’s not a good place to be in because the team just experienced a failure.
Overtime, as I experienced Post-Mortems in different circumstances, it became clear that some of these issues could have been identified ahead of time. The question was how to help identify those issues early and take action to avoid or mitigate them. This is where Pre-Mortems come in.
Pre-Mortems are brainstorming sessions about how a project plan can possibly fail. The process is to take the team to a hypothetical future in which the project has failed. It is now up to the Team and the Engineering Manager to ideate what are the possible causes of the resultant failure.
Pre-Mortems are incredibly beneficial as this process taps into the creativity and valuable experience of the team members to identify ways the project plan could fail and ways to mitigate them.Gilles
In my experience, there are 3 elements that can make Pre-Morterm sessions particularly successful:
1. Create a safe psychological environment
Let’s face it, it may be uncomfortable to openly point out flaws or weaknesses in a plan, especially with the people behind the plan in the same room. Some of them could be colleagues or managers. That’s why it is very important for the meeting facilitator to make a few introduction statements to create a safe psychological environment for the participants.
- Explain the purpose of the meeting and the expected outcome: a list of potential flaws or reasons for failure, and mitigation actions for each of them.
- Explain that this is not to criticize anybody or the plan in any way, but this session is only aimed at making the plan stronger and improving the chances of success for everyone.
Even if you are having this session with the same team repetitively over time for different projects, it is important to repeat this introduction so as to keep everyone aligned and keep the energy over time.
2. The person accountable for the project speaks first
As an engineering manager, I am the person accountable for the delivery of the projects that our team tackles. I personally make a point to open the discussion by generating the first idea. This makes everyone in the room more comfortable: if the person accountable criticises their own project plan, anyone can do it too.
Furthermore, as the first speaker, it is important to make a rather big point as this will set the bar for others to follow. If the first person to speak picks a shallow or light element, others may not feel comfortable discussing the hard issues.
3. Keep the creative flow going,
then afterwards go into solution mode
The meeting should have 2 phases. The first phase is the creative one. This is the time when the team generates all the possible ways the project can fail. As a meeting facilitator, it is important to keep the group in that flow of thinking. Whenever someone says something, just write it on the board without any discussion or filtering. Just write it on the board, thank the person for sharing their idea, and keep on encouraging others to generate more ideas.
At some point, the flow of ideas will slow down and the team may be tempted to start discussing the solutions to some of the items. My recommendation is to not jump into solution mode immediately. Keep the group in the creative flow for a little longer to get a few more ideas.
Once the creative phase is over, start discussing each item one by one and provide a preventive measure or action that can be taken to mitigate the risk.
Note that you may end up with a very large number of items to look at. This is usually in direct correlation with the size and complexity of the project. The rule of thumb here is ‘the more the better’! The more possible reasons for failure are identified early, the less likely the same reasons will hurt the project later on. So if the meeting ends with a list with 20, 30, 40 or even more reasons for failure, this is actually great news for the team that will deliver the project.
Here is what a short example of a pre-mortem meeting notes could look like, with actions for each added in the 2nd phase of the discussion:
- The project plan still needs approval from Director X.
- Team Member A will follow up with Director X to try and speed up approval.
- If step D fails, the whole project will fail.
- Move step D forward in the project plan to start and finish it faster to reduce uncertainty over the entire project.
- Supplier M may not have the necessary quantities to supply the increased quantities we need for this project.
- Team Member B will check with Supplier M that they will be able to supply required quantities. If negative, initiate conversation with procurement to find an additional supplier.
This kind of session has been incredibly useful for me and my team; and I can only recommend others to include it in their planning process.
- McKinsey: https://www.mckinsey.com/business-functions/strategy-and-corporate-finance/our-insights/bias-busters-premortems-being-smart-at-the-start
- Stanford (Guy Kawasaki): https://www.youtube.com/watch?v=zN-PT8PYjTo
- Harvard Business Review: https://hbr.org/2007/09/performing-a-project-premortem