Do You Do Post Mortems

Published on Friday, 20 December 2013 10:21:56    Written by Marc
This blog is probably more of an open question than the traditional entry where we try to provide you with experiences that one of us had. We normally try to do our best to share some ideas, tips, and warnings that we either came across one day, or completely fell into the hole on another day. There is nothing like the pain of real live experience to be able to relate to something.

Conducting post mortemsWhy do we do post mortems?

So, back to the title of the blog, “Do you do post mortems?” First, I would ask why we would do post mortems in the first place. Well, if you do a project and it does not go as planned, do you know where something went wrong? Did something go wrong at the beginning? Maybe all was OK and the outcome was the best you could have had. Perhaps it was the planning that was not perfect but the outcome was great. Then again, could the outcome have been better?

Why should we do post mortems?

If you do not do post mortems, chances are you cannot identify areas where you could have improved during the process. Was it with one supplier, an item that was delayed for a reason that could have been prevented, something everyone forgot about, or that was not budgeted? In order to become better, you need to do post mortems, to analyse what went good and what went wrong, with the obvious idea of improving what went wrong and replicating what went good in the future. Because without post mortems, you will simply have nothing to measure (good or bad) and later you will be back at a new project without ever improving your work. However, how many companies actually take time after a project to stop and do a detailed post mortem? Sadly, not many. Most companies are so busy and tied up that once a project is done, they are already late for the next one. In addition, this is the sad part, because going from project to project without doing any post mortems is like driving without having a destination in mind. You will get somewhere eventually, just maybe not where you would have liked to go.

How do we integrate post mortems into our project?

How can you insure that you will make post mortems part of the process? I found that there are many ideas on conducting post mortems and a number of web sites that offer information, training, and templates. Many offer valuable information and in theory, they are mostly good. However, after doing various projects for well over 20 years I found that many of these theories are hard to apply in real life.

Many of these theories originate from various consultants who are happy to help companies for a fee and if they can make the post mortem process complicated, well they can better justify their work. Not that we have any problems with consultants, its only that too many times some of the guidance they provide is simply not practical for most companies struggling to balance between improving themselves and surviving. If you are a company that is struggling to try and increase profit, chances are that doing lengthy post mortem process will not be your priority. You would probably be right.

So, coming back to the question above, how can we make post mortems considering all the other challenges and work we already have to deal with? Well, first make it a part of the process. If the post mortem is an integral part of the project, there are more chances that it will be looked at than if it is a simple element that you promise yourself to tackle once the project is completed. Making the post mortem process as a part of the project will insure that you take notes during the process, not only at the end when everybody forgot about things and are working on another project somewhere else.

Also, make sure to appoint someone to be responsible to follow the post mortem process, take notes, and ask questions to the people on the project. Having a person that is responsible for this process will go a long way in avoiding that the post mortem gets forgotten mid-way in the project.

What are the components of a post mortem?

Before you start the ongoing process, you need to clearly define what you want to measure and what would you consider to be a success once the project is completed. Remember that quality metrics are subject to interpretation, so metrics that you can measure (i.e. put a number on it) are better. Although many metrics are easy to measure (such as time and cost for example), other may be more difficult. For example, if one of your metrics happens to be quality of something, try to define exactly what you could measure as part of determining what the quality is rendered by the project. Is it the number of defects, the cost of redoing some work of poor quality? If you were building a new construction, what would you consider as metrics related to quality that can be measures without interpretation?

If you are building a new warehouse for example, your metrics of time (when the project will be completed and ready) is easy to determine. Same for the cost metric (how much did the project cost in the end). In the case of quality metrics, you might establish a number of defects and cost associated with fixing them. Finally, do not forget other metrics, which may not be about time, cost, or quality per say, but sort of a mix of them all. For example, you might establish an energy consumption target as a metrics, if you wanted your warehouse to have a certain level of energy efficiency. You might aim for a LEED certification status. These are all metrics, which can be established in the beginning of your project and reviewed in your post mortem.

How do I gather the lessons learned from project participants?

Once your project is completed, it is always good to seek additional information from the project participants. Some will opt to have a face-to-face meeting with all participants in a same room; others will do conference calls, or ask participants to submit information on templates that are provided to them. Regardless of the process used, the important thing to remember is that you need to gather information from the project participants before the project is 100% complete. Do it at the very tail end of the project but before people leave for other projects. Show them the metrics you used and ask them to rate the project. This is where templates can sometimes come in handy (especially in theory) if (in real life) you can convince people to take the proper time to fill them in carefully and not in a rush simply to get it over with.

You can then compile the information and bring it to a general meeting or conference call to discuss the results. Take time to ask people what could have been improved. This can sometimes justify having a face-to-face meeting with all participants because they might share information verbally that they would not take the time to share on a template. However, speaking from my own experience, if there were areas of personal friction on the project, these will unlikely come up at a general meeting, as people prefer to say nothing that to start a confrontation in a group. Because of this, I tend to prefer getting some ‘off the record’ information by talking one-on-one with key participants when asking what could be improved from a human resources perspective. As we all know, projects are made by people and the team will make or break a project depending on how well they work together. This is an area where there is always something to improve but is also the most difficult area to get information.

Let’s Create Our Post Mortem!

Finally, once you received all information from participants, compiled info, debriefed with people, issued recommendations, and discussed them with the participants, you need to share the information within the company. Some will say that you need to share the information with all the company, but in practice, you only need to share the information with those that might need it, such as other project teams. In today’s world, people are already overwhelmed with information, emails, reports, templates, charts, and memos that if you try to share your post mortem info with everyone, most people will simply not read it. Make sure your post mortem conclusion and recommendation and clear and concise. Put them in bullet form on one single page and send them to people that will have usage for your recommendation. Ideally, only make one or two recommendations each time since overloading people with recommendations is often useless, people tend to forget long lists.

In summary, companies could benefit more from doing post mortems, but implementing this process has to be done taking into consideration the workload of people. Make it part of the project process, find metrics that are easily measurable and propose a few clear recommendations. Over time, you have a good chance that the process of doing post mortems could evolve into something that is standard in your company. The more it is seen as bringing benefits to the people, the more they will accept to spend some time to be involved and this is really the key of success for post mortems.