A guide to conducting a post-mortem analysis

In the world of IT, disaster is inevitable. Whether you’re a start-up that suddenly stops working, or the entire Google suite goes down due to an intern’s error, no venture is completely bullet-proof. However, what separates a mature from a less experienced organization is how you deal with the aftermath of the problem.

I’m not talking about a public apology or refunding angry users, but about how to internally deal with the next steps once the issue is solved. Junior organizations might play the “blame game” and look for someone whose unknowing neglect or proactive, well-motivated, actions led to the problem. This isn’t the best solution.

A mature company learns from disaster and actively invests in preventing the issue from appearing ever again. However, even if things do not go wrong, you can always make them go better!

You learn a lot during any project and it’s a great practice to highlight, record, and act on any learnings. To do that, it’s best to invest in a so-called “post-mortem,” the topic of today’s article. Stick around to know how to best set up your organization and team to deal with future crises!

What is a post-mortem analysis?

A post-mortem analysis is a process that helps identify the causes of a failure and how to prevent it in the future. It’s different from a retrospective, as it’s not a part of regular Agile Scrum ceremonies and is always focused on a specific event.

Steps for running a successful post-mortem analysis

For a successful post-mortem analysis, you should consider the following steps:

1. During a brainstorming with the team, pinpoint the project highs and lows that stuck out
2. Limit the selection to the items that the team can draw most learnings from, i.e. by voting on the most eye-catching topics
3. Analyze the factors behind major successes or downfalls of the selected topic
4. Recap, draw conclusions, and outline steps for improvement and best practices for future projects

The process will be similar when the post-mortem framework is applied to analyzing a disaster that happened to your product. Here you basically have one “bad” topic and the whole goal of the meeting is to identify what led to the catastrophe, how it could have been prevented, and what to do in the future.

Benefits of post-mortem analysis

If you’re still on the fence about implementing a post-mortem analysis, here are some benefits that might sway your opinion:

• Learning from failures and successes — By developing a growth mindset you can drive continuous improvement. This way your team’s future projects can be delivered quicker, more efficiently, and with a lesser risk of failure and/or missing the agreed deadline
• Uncovered vulnerabilities in your system — This is more relevant to in-house or internal products, but also presents opportunities with external clients. If you uncover a security fault, you have an opportunity to fix a security glitch!
• Enhanced team communication — Collaboration and morale increases when you celebrate wins and share honest feedback. Similar to a review and retro meeting, this is a great opportunity to flag specific team member’s accomplishments and findings
• Improved efficiency — You can Increase your profit margin, project cost estimates, and resource allocation by pinpointing process improvements and best practices. This is especially important when the learnings are not in the coding but at the process level

Best practices for implementing a post-mortem analysis

This article refers to a specific meeting that will likely run from one to two hours every now and then. Because you’ll only run it on occasion it’s important that you have a sense of the best practices so that you get the most out of your meeting. Try to implement the following:

1. Choose a moderator who can set the meeting goals, timebox certain agenda topics, and make sure that the conversation is focused. The meeting moderator will also be responsible for collecting and documenting action points
2. Find the tool that works for your team to run the meeting. You will likely need to support a hybrid environment with remote team members. Miro is great as a virtual whiteboard with sticky notes for any attendee to use
3. Follow a clear agenda that includes a recap of the project, and a review of the outcomes and results. This will help to recap everything so that the brainstorm can correctly identify all the positives and negatives of any project
4. If the topic focuses on a particular outage, prepare a detailed timeline of the events that led to the product going down. This way it can be analyzed and future countermeasures can be discussed effectively
5. Focus on the facts and the process, even if there clearly is a person to blame. If so, this person should never have been in a position to cause an outage and the procedures need to be updated in order to prevent similar mistakes in the future
6. Make sure to close the meeting with clear action points, people assigned to complete them, and some kind of oversight that will see the agreed action through

Final thoughts

As a PM, it’s important that you don’t just do the same thing day in and day out without reflecting. Just like with agile, you want to work towards improving processes. Because of this, it’s vital to be able to recognize not only what you do well, but also what you failed to achieve.

By implementing post-mortem analysis into your product team you can better understand your failures and take active steps towards preventing them in the future. Good luck!

Featured image source: IconScout

LogRocket generates product insights that lead to meaningful action

LogRocket identifies friction points in the user experience so you can make informed decisions about product and design changes that must happen to hit your goals.

With LogRocket, you can understand the scope of the issues affecting your product and prioritize the changes that need to be made. LogRocket simplifies workflows by allowing Engineering, Product, UX, and Design teams to work from the same data as you, eliminating any confusion about what needs to be done.

Get your teams on the same page — try LogRocket today.