Root cause analysis

A classic approach to root-cause analysis to ask “why” five times. If you already understand the underlying causes of your problems, then you have already conducted root-cause analysis. However, it’s easy to get stuck on a particular solution. Asking “why” five times may give you new insight. Here’s a real-world example.

Problem: When we start working on a new feature, we spend a lot of time before we can ship it.

Why? Our building process takes too long.

Why? Mainly because automatic tests fail.

Why? Because Engineers rarely cover new functionality with automated tests.

It’s tempting to stop here and say, “Great! We found the root cause. Engineers need to cover the new code with the tests.” While that’s correct, Engineers know they should write the tests which is already part of continuous integration, but they aren’t doing it for a good reason.

Why? Because oftentimes projects take longer than originally estimated. Engineers feel too much pressure and build technical debt to deliver sooner.

Why? You could come up with several root causes here, which may simultaneously contribute to the problem:

1. Because estimations are misused. Originally the scope was too unclear and estimations were too rough, so the business shouldn’t rely on such estimations.
2. Because people prefer cutting corners (building technical debt) instead of adjusting the scope or updating the estimations.
3. Because Engineers aren’t skilled at estimating. Lack of experience and historical data. Disruptions such as incidents, risks, absences, amount of unknown unknowns are not tracked and information is lost for future projects.
4. Because team works not in a “flow” state. People fail to communicate effectively, lack transparency and support.
5. Because team doesn’t have the right KPIs. Team isn’t encouraged to reduce the Cycle or Lead Time over time following the DOD (Definition of Done which requires test coverage).
6. Because team morale is low. What people say about their company? Their leadership? Their colleagues?

Decide on solutions

Don’t forget to converge the thinking process and prepare to act. The outcomes will be lost otherwise. Ask the team:

• What can you do to prevent the problem from happening again?
• How will the solution be implemented?
• What are the risks of implementing the solution?
• Who will be responsible for it?
• How will you follow-up?

Applications

• You can run a quick “5 Whys’s” during a Retrospective without interrupting the flow of the meeting.
• You can set up a separate meeting for the problem to collect various points of view.
• You could sent up a process that requires root cause analyses. Typically, one person takes on this task and prepares a write-up. Then, this white-up is shared and read together.
• Lastly, encourage people to conduct root-cause analysis in the privacy of their own thoughts. You can use it for yourself. A lot of causes are beyond our control, so try to channel the frustration and energy to fixing processes that we can influence. Share conclusions and reasoning with each other. Involve whoever is necessary to fix the root-cause.

Quiz vector created by vectorjuice – www.freepik.com