Improvement Process

As organisations & teams mature or change there is always a need for processes to change too. There is never one process that works for every team all the time. Things need to be changed / adapted over time. We're always looking to change or adapt our processes to better fit our needs as they change.

Retrospectives

As part of our scrum process we have retrospective meetings. We talk about what has gone well in the past sprint and areas we should look to improve or change. Usually improvement suggestions that come out of this meeting are immediately actionable and we take that action straight away. Particularly complex ones may require follow up research or dedicated workshops.

Crazy T's

A perfect example of this is when during one of our Retrospective sessions, we realised we had too many meetings spread throughout the day and our productivity was suffering from the constant context switching and lack of time to properly focus. We agreed to "ban" all non urgent meetings on every Tuesday of every week. As a result, people felt more productive and relaxed so we extended it to both Tuesday and Thursday and this is how the "Crazy T's" were born!

Later on we figured out that because other processes evolved we can no longer keep this a practice. Therefore, right now you can find this in here.

Failure Board

Failure is a part of everything we do or try and we find that totally acceptable. What's important for us is to learn from our mistakes and openly talk about them. For that we have our Failure Board. This is a simple list of failures that have happened and actions that were taken to fix those or to prevent them from happening in the future. Over time this will also become a list of war stories we can laugh about at a later stage 😉

Here is an example of how such a story is presented in our failure board.

Fail 💥	Result 🐛	Actions Taken ✅
While cleaning up "stale" resources in Kubernetes, Ross accidentally deleted a service with a static IP exposing a logging service, which was being directly used from a test application	The test application lost connection to the logging service and logs were lost for a couple of days.	We now keep a clear documentation of active endpoints. We now enforce DNS names for all of our services.