Making Retrospectives Effective in a Long-Running Project

Question

I'm a ScrumMaster for a software firm, and we have a project that is entering its 35th Sprint. While I appreciate the value of Retrospectives, it's becoming progressively more difficult to get good insights from our Retrospective events. We have shortened the meetings to avoid wasting a lot of time, and even the short meetings don't seem to get as much content (or quality content) as they used to.

I'm wondering what strategies the good people of PMX use to continue to get value out of Retrospectives, even for a very long-running project. Are there different formats that you use as a project goes on longer and longer? What insights should we be looking for later in a project? Do you stop doing Retros altogether?

Thanks for any insight!

Answer 1

There are several things that you can do to spice up your retrospectives and keep getting value out of them:

• Have somebody else facilitate the retrospective. E.g. a team members, Scrum master from another team, agile coach, etc. A fresh facilitators can lead to other kinds of discussions, insights and views, topics or improvement approaches.
• Change the setting for the retrospective. Go outside, hop over to a nearby meeting center or facility, visit a museum, have dinner together, anything that can help team members to get out of the daily routine and get refreshed.
• Change the audience: Do a retrospective with stakeholders, customers, people from operations, or any other group of people that the team collaborates with. When the audience changes, the scope and perspective changes along; it becomes a project level retrospective, end to end retro, DevOps retro, Business - IT retro, etc.
• Do a retrospectives on how you have been improving as a team. This includes reflecting on the way that you are doing retrospectives, and also on how the follow up with improvement actions is done.
• Celebrate successes. For a team that's been together for so long there must be things that are going great, so take time to explore them and learn more about things that go well. Find your team strengths.
• And yes, change the retrospective exercise that you use (here's a toolbox with retrospective exercises)

Don't wait until people are getting bored in retrospectives. Change your exercise, setting, facilitator, etc frequently so that your people stay fresh and keep coming up with useful improvement actions. Keep your agile retrospectives valuable!

Answer 2

Mike Cohn blogged on this quite recently.

His comments included trying out different retrospective formats and having a Scrum Master from a different team run the retro.

Mike does emphasise that even a Scrum team that has been together for 10 years gets benefits from retrospectives.

Answer 3

Great Question! While retrospectives are a critical component of continuous improvement, they get become ineffective over time if the same approach is being used each time.

This thread has a ton of different retrospective formats in the answers, so I won't rehash them: Different ways of running Retrospectives

As for when to change, you want some consistency so the team gets used to it, but if it starts becoming mundane, you need to switch it up. For a lot of teams that ends up being around 4 or 5 sprints, but see what works for you. There are also some formats that are easy to switch between. The plus/minus/delta, the boat retro, and the starfish retro are all similar in concept, so jumping between them is easy. Going from a starfish retro to a 4L's approach is a bigger mindset shift and should be done with more care.

Also, as a Scrum Master, you may find situations that call for a certain technique. If the past few sprints have been really low energy in the team, doing a retro that asks what was exciting and what was frustrating get the team talking about those items in a way that "What went well" wouldn't.

Answer 4

No, you should always keep doing retrospectives. It is one of the 12 core Agile principles:

At regular intervals, the team reflects on how to become more effective, then tunes and adjusts its behavior accordingly.

There is always something to improve or to try different to become more effective. I haven't seen the perfect team yet.

One of my teams had great success with Retromat to spice up their retrospectives. Retromat generates a new fresh retrospective format for each time, give it a try :)

Answer 5

Process for the sake of process is a bad idea, and one of the big downfalls of scrum is that it is so process heavy. For this reason I think it is okay to occasionally not hold the retro, especially if the sprint went well. You do not need to hold a special meeting to celebrate wins, as wins can be celebrated at any point in time. I had a team that was on its 67th sprint and the team had terrific chemistry and was a very successful team. I almost always held the retro, but occasionally during the stand up on the day of the retro I would do a Fist of Five on the success of the sprint and if the fists were unanimously Five I would offer that the retro be skipped. I like scrum. Scrum is a fun way to develop software, but I loathe process.

My advice to you is to try the Fist of Five method, look for unanimous Fives, and then make the offer to skip the retro. Maybe not every single time, but occasionally is okay.