Over the years having been involved with, and having observed, a number of development teams, one of the most vital activities I have found a team should do is retrospect at a regular cadence.
A concerning pattern that I have observed in organizations and teams that are moving from a hierarchical or ‘traditional’ structure to a more collaborative structure is they forget to intentionally retrospect. In many instances it becomes the forgotten meeting.
Retrospectives often become the forgotten meeting…
Why this is? I’m not sure - I suspect it is because stand-up’s and so forth can be mapped closely to more traditional meetings whereas retrospectives are a totally new type of meeting.
Another pattern that I have observed from teams is that when they get to what they think is an acceptably good place they become comfortable in how they work. At this point they move the retro from a regular cadence meeting to something they do on demand. This worries me allot.
Don’t make the retro an on-demand meeting
To implement incremental iterative improvement you need to meet at a regular cadence. Meeting sporadically defeats the purpose of incremental improvement.
What are we trying to achieve from a retrospective?
I see two types of things we generally get from a retrospective:
- Recognize we are human
- Implement iterative and incremental improvement
Let me explain…
Recognizing that we are all humans
When a team is NOT in a good place socially, I see the primary purpose of retrospective to:
- Vent, discuss, bring up issues in a safe environment
- Blow off some steam
- Draw closer together as people
Team’s that are still struggling to emotionally connect really struggle to incrementally improve.
Implementing iterative and incremental improvement
When a team is in a good place socially, I see the primary purpose of the retrospective to:
- Review the success or value of the improvements they set out from the previous retro
- Identify the one or two small improvements (or experiments) that they can do in the next period
When a team embraces the fact that they can try something and if they see value then enhance it, the team starts to move towards being a self organizing team. Team’s that are self organizing are generally happier and more effective.
My Four Step Incremental Improvement Retro
There is no always ‘right’ or always ‘wrong’ format for a team retrospective - context and time matters. That said, I generally use the following structure:
- Ice Breaker / Change Mindset
- Review of Previous Retrospective Goals
- Main Activity
- Confirm goals to be achieved by next retro
In general, for a team of about 8 people that retrospects once every two weeks you need about an hour and a half to cover this format properly. If your cadence for retrospectives is on a more frequent interval retrospectives can be shorter. That said, you need to give some time between retrospectives for people to do things otherwise it becomes a futile meeting.
Step 1 - The Ice Breaker
I use the ice breaker to help the team make a mental shift from immediate work to taking a step back and looking at how we are working together.
Be strategic about your ice breaker
Don’t just do an ice breaker for the sake of an ice breaker. Be strategic about what your ice breaker is. An ice breaker is a great way to get people to get to know each other better.
- Are there any long term objectives that the team is trying to develop?
- Are there activities we can do that help team members know other team members better?
Examples Ice Breaker Activities
- Fizz Buzz Group Number Game (also known as Ping Pong)
- Organize seating based on distance that one was born from the current location
- Tasty Cupcake Ideas
Step 2 - Review of Previous Retrospective Goals
Review of previous retrospective goals is critical if you want to incrementally improve. There is little point setting new goals if you do not gauge how successful you were with your previous goals.
Typically, I like to go through each goal we had in the previous retro and ask the following:
Did we achieve this?
- If Yes, did we see value in it?
- If Yes, do we want to do something to enhance it
- If No, why did we not achieve it? Was it to vague, unmeasurable, to large, we were to busy, not really valuable?
- If to vague, what should we be careful of going forward so that we don’t make goals like this again?
- If it was unmeasurable, how could we have measured this (generally the goal was not specific enough)
- If to large, how could we make it smaller?
- Do we want to do the smaller goal and carry it over for the next period?
- If to busy, do we still see value in it, do we want to carry the goal over for the next period?
- If not really valuable, how do we avoid making unvaluable goals like this going forward?
- If Yes, did we see value in it?
I’ve found going through each goal individually is useful. Do not rush through this process, if the conversation starts getting to deep on a specific goal and you feel you are running out of time to review the other goals, ask if you can park the topic till after the review of the other goals. Don’t forget to come back to it afterwards. If you don’t, you loose trust as a facilitator.
Step 3 - Main Activity
Plan a main activity, but don’t be tied to it. Forcing a team to do your activity because you planned it when there may be something else the team want’s to discuss is selfish. Go where the heat is.
Usually giving the team an opportunity to propose items to discuss as part of the main activity is a good thing. Most of the time it makes sense to go through their proposed topics. Sometimes it is more useful to take a bigger step backwards to look at the bigger picture.
Be careful not to let the main activity take up all your time, you need enough time for confirming goals for next period
Lean Coffee Style Retro
In the teams I’ve been with we typically have a retro topics wall in our team area. During the iteration before the retro we put topics up we would like to discuss. These become useful if you run your main activity in a lean coffee style retro.
Quick Wins vs Deep Discussions
Use techniques that allow you to identify what the hottest topics are. Generally I find two types of topics that come out:
- Quick wins
- Deep discussions
Try identify quick wins first. Don’t assume something is a quick win because you think it is. Get consensus from the team that a topic is a quick win. If everyone in the team agrees it is a quick win and they want to adopt it, adopt it - make it a retro goal outcome. If only some of the members of the team think it is a quick win, but others do not - then it is not a quick win - move it to the deep discussions section.
With deep discussions you are only going to be able to cover 1 or 2 topics in an hour retro. Find a mechanism for the team to identify what they feel is the most important topic. Facilitate discussion, but if it is going nowhere don’t feel shy to suggest the topic is parked. Talking about something without generating goals / outcomes can be frustrating. A good indication if a deep discussion is still worth pursuing is:
You don’t have to cover everything, focus on where the heat is…
- Gauging how involved the team is in the discussion (if only one or two people are engaged, it should be parked)
- Are proposed goals emerging (if they are not, push the team to find a small improvement, if they can’t suggest it get’s parked and move on)
Ideas for Main Activities
Step 4 - Confirm Goals for Next Period
This is probably the second most important step in a retro, with Review of Previous Goals being the most important.
Make sure you have enough time at the closing of a retro to do this - do not rush this step
By this stage you should have a number of ‘proposed’ goals for the next period (until the next retro). Remember, retrospectives are aimed at incremental improvement. The more improvements or experiments you introduce, the more chaos. Typically I have found small teams can only really achieve 2-4 ‘improvements’ in a 2 week period. If a team has a lot of proposed improvements you want to get some sort of priority on where they feel the most value will be added.
Having a retrospective at a regular cadence is vital for incremental improvements. For a team to create trust they need to have regular touch points to evaluate if they are doing what they said they were going to do. Retrospectives are one of those touch points.