Put up two posters labeled 'proud' and 'sorry'. Team members write down one instance per sticky note. When the time is up have everyone read out their note and post it to the appropriate poster. Start a short conversation e.g. by asking:
Did anything surprise you?
What patterns do you see? What do they mean for you as a team?
What could sub-groups improve when interacting with others? Source:
Identify sub-groups within the participants that interacted during the iteration, e.g. developers/testers, clients/providers, PO/developers, etc. Give participants 3 minutes to silently write down what they think their group did that negatively impacted another group. One person should be part of one group only and write stickies for all groups they don't belong to - 1 sticky per issue.
Then in turn all participants read their stickies and give them to the corresponding group. The affected group rates it from 0 ('It was not a problem') to 5 ('It was a big problem'). Thus you get insights and shared understanding about problems and can select some of them to work on.
Brainstorm what behaviors to keep, drop & add and pick the top initiatives Source:
Divide a flip chart into boxes headed with 'Keep', 'Drop' and 'Add'. Ask your participants to write concrete proposals for each category - 1 idea per index card. Let them write in silence for a few minutes. Let everyone read out their notes and post them to the appropriate category. Lead a short discussion on what the top 20% beneficial ideas are. Vote on it by distributing dots or X's with a marker, e.g. 1, 2, and 3 dots for each person to distribute. The top 2 or 3 become your action items.
Gauge participants' satisfaction with the retro on a scale from 1 to 5 in minimum time Source:
ALE 2011, Corinna Baldauf
Put sticky notes on the door with the numbers 1 through 5 on them. 1 is the topmost and best, 5 the lowest and worst.When ending the retrospective, ask your participants to put a sticky to the number they feel reflects the session. The sticky can be empty or have a comment or suggestion on it.