At the recent Agile Australia conference Linda Rising gave a keynote talk on Organizational Myyths and Patterns which can be viewed here. Afterwards she spoke to InfoQ about using continuous retrospectives to motivate teams to take an experimentation approach to making constant small changes to the way they work. She advocates the importance of an experimenatal, learning mindset.
She describes the process as follows:
The idea is that in the normal retrospective, there are lots of exercises that we do and one of my favorites is called the Timeline. I am sure you have seen it. You use cards or stickies of different colors and you begin. If it is a reiteration retrospective, you begin with the beginning of the iteration and you put the date and then the end date and then across the time line you put the stickies or the cards that reflect events and they are of different colors. Then you reflect back and you use that to drive actions that you are going to take at the end of the retrospective because, unfortunately, most of us, not just old people like me, cannot remember what happened. So it is an exercise to help you remember what happened.
What I have done is I changed that a little bit and now, instead of waiting until the end of the iteration, we begin in real time, at the beginning of the iteration. So here we are at the beginning and every day, when we do a stand up, we do it in front of the timeline and as we are chatting and actually doing the stand up and reflecting afterwards, it will happen that people will remember and think of something and will put events on the timeline in real time. Now, it is not just the normal kinds of events which would be things that make you happy or things that make you sad, or things that make you angry.
It would also be ideas driven by those events that suggest possibilities for experiments. So you use a special card, either of a different color or a different kind of sticky or card to put that on the timeline as well. What about an experiment that would and then here is the suggestion on the timeline for the experiment. Sometimes we will not wait until the end of the iteration, but we will actually start the experiment and in fact, sometimes, just putting it on the wall will engender some interest. People will say “Well, that is a great idea. Let’s do it”.
So instead of waiting until the end and then debating on action items, I have experiments running all the time. What that drives is continuous learning instead of “Now we learn, now we reflect” it is “Oh, we could do that in the middle of the iteration” and the quality or the quantity of the items that come up for the timeline are different. They are more immediate, they are more about what is happening now. We do not have to remember because it is in the moment. So, it does become continuous reflection and it does become continuous action in the sense that we are running experiments all the time.
Now, we still have our end of iteration retrospective just as when we started iteration retrospectives we did not get rid of “end of release” or “end of project” retrospectives. We still have those because the focus is different. The focus at the end of an iteration is more tactical, the focus at the end of release or project is more strategic, whereas these continuous retrospectives that we hold daily are more about doing the little experiments. And it does not just happen during the stand ups, it is that someone else will come by, maybe someone on another project will have a conversation with somebody in our team “Oh, I see you are doing this experiment. Maybe we could try that” And so we are spreading information about what is going on and encouraging them to get involved in the kinds of experiments that we do and that it is sort of shared learning and it happens all the time.
We can also have a conversation between people on our team that is not a stand up, but they gravitate toward the timeline. So now, there, in that meeting they are also reflecting, also adding to the timeline, also coming up with ideas for experiments. I have only done it a couple of times, but it seems to work pretty well and in the talk at XP 2015 I said “Well, you could try it and you could let me know” And I did hear from a couple of people who said “We sort of have been doing something like that” but they had not done the experiments, so now they were going to try the experiments.
Since I talk about experiments – that is what I am going to be doing at the Agile Conference in Washington DC. I will be trying to make those experiments more scientific and to get us to learn a little bit more about what an experiment is and how it could be. It cannot be a completely randomized, double- blind placebo controlled experiment we just are never be able to do, but we could be more scientific, we could define an hypothesis, we could do some kinds of randomization, we could do a little bit of analysis. I do not want to give that talk away, but that is going to build on talk on continuous retrospectives.
About the Interviewee
Linda Rising is an expert in helping people make changes in their organisations, with proven change management strategies to enable you to be a more successful change agent in your organisation. She has authored four books and numerous articles and is an internationally known presenter on topics related to patterns, retrospectives, influence strategies, Agile development, and the change process.