In the spirit of constant learning and improvement, the purpose of retrospectives is to analyze the current sprint and change the team's behavior accordingly in the upcoming one. In this article, I will share a few ideas and challenges that my team and I faced while doing the "retros," and, more importantly, while trying to implement our decisions in subsequent sprints.
The unbiased choice
First, let's consider the choice that must be made. How should we change? How should we pragmatically select the most important things to focus on in the fairest and most balanced way? Team members always have something to say when asked what should be kept or done away with.
It is common practice for each team member to have an equal number of votes — for example, 50 — that he or she can divide among different suggestions. Typically you will observe that most members place the largest number of votes on their own proposals. This reflects a natural bias one has toward his or her own suggestions. ("With all due respect, I am the smartest guy here.") To neutralize this tendency, I suggest limiting the number of votes members can place on their own suggestions (e.g., 25%). In this way, a team member is forced to deliberate others' opinions and select the best ones based on clear and unbiased consideration. Having done that, we should take the top five suggestions as those that we need to adopt in the next sprint.
The difficult part is over. But is it really? Not quite.
The change that holds
One sprint later, we realized that in the heat of the sprint, we didn't exactly implement all the changes we had agreed to adopt. The pattern repeated itself in the next sprint. What was happening? Why couldn't a group of well-intended developers adopt five changes right away?
The answer is old habits
. This is actually not very different from adopting any habit in other areas of our lives. A behavior becomes habit the minute it sinks under our consciousness and becomes automatic — our second nature. This process takes time and practice. Trying to adopt, drop, or change five new behaviors every two or three weeks, in addition to all the other numerous tasks we need to focus on, is hardly an achievable goal.
So what do you do? Pick only one change! With so many things to do in a sprint, adopting one change has a higher chance of success than adopting several. And even if you select only one, there is a chance that it will be forgotten (don't underestimate the strong inertia of habit). To improve chances for the successful adoption of a change, I recommend writing it down in big letters and putting it in a very visible place (e.g., the team board). This will be a constant reminder to the team of the change it is working on adopting now. You may even keep a few of the previous changes on the board for several sprints (a kind of circular buffer of improvements) until the older ones become part of the team's natural behavior.
Finally, do not forget to review your retrospectives periodically. Determine what is working and not working in the improvement process itself.
To summarize, consider the following takeaways:
- Put most of your votes on others' improvement suggestions.
- Try to adopt one change at a time.
- Review your retro meetings periodically.