alacrity in action
| Tags:

When I first started participating in and then running retrospectives we would always start our meetings, typically on a Friday afternoon, with a certain ritual. We would write down in bold letters and then read out to each other a piece of text know as the Retrospective Prime Directive created by Norman Kerth. I know many agile teams use and cherish this little ritual. However, over the years, I have grown increasingly weary of it. I tend not to use it anymore as I think it can be harmful to the goal of a retrospective.

I have had many discussions with people about the prime directive’s application, usefulness and effects. I have promised to share my thoughts in a more coherent fashion after a few email and twitter discussions on the topic with Yves.

In order to better understand the Prime Directive I decided to return to the source and read Norman Kerth’s bookProject Retrospectives: A Handbook For Team Reviews”. It turns out to be an excellent resource and on page 7 we find Kerth’s Prime Directive:

Regardless of what we discover, we must understand and truly believe that everyone did the best job he or she could, give what was know at the time, his or her skills and abilities, the resources available, and the situation at hand

BTW. Notice how often people skip the word “must” for the directive.

As I read the book I realised that the Prime Directive is really just a tool and as with any other tool is has been created to work in a certain context. Let’s try to uncover and understand this context.
Norm rightly points out that a successful and effective retrospective requires safety. Part of feeling safe means knowing that “there will be no retribution for being honest”; in other words, we’re trying to take blame off the table in order to establish trust. Other actions that Norm suggests is hiding particularly uncomfortable information discovered in preparation for a retrospective or taking managers out of the room in his “Session Without Managers” exercise. These are all sensible steps to take when you’re trying to create a learning experience in an environment which exhibits, what we could label after Bob Marshall, an analytic mindset.

Another crucial piece of context for the prime directive is that Norm used it in retrospectives that analyse 6-, 12- or even 18-months projects, last for three days and are held off-site. A three-day retrospective starts with the Prime Directive but at the end of day three it is usually nowhere to be seen because other exercises like “Emotions Seismograph”, “Repair Damage Through Play” or “Cross-Affinity Teams” have improved and maintained the feeling of safety.

My main objection to the Prime Directive however is that it explicitly pushes some issues into the undiscussable zone. We are told to believe and make ourselves believe that people acted with best intentions. This might be right most of the time but with any social interactions there will be times when a colleague annoys and hurts us in a way that affects how much we care about our job. After the prime directive is out we can no longer discuss these things. It would violate the directive. Norm is perfectly aware of this limitation. That’s why, at the end of day three, he runs an exercise called “Let The Magic Happen”. This is where the team is given an opportunity to discover any elephants in the room and bring to light issues that were so far undiscussable.

If your retrospectives last for two hours and happen every two weeks starting every single one with the Prime Directive might be the best sign that your retrospectives are not working, the team is not getting on together and that potentially the most pertinent and important problems will never be tackled.


| Tags: , ,

Last week I was at one of my favourite Agile events of the year – the ACE! Conference in Kraków. It always attracts interesting speakers and this year was no different. But talks are only a small part of such event for the real value is often is the conversations that follow.

During a few talks this time I happened to have been sitting near Bob, in two of the talks speakers shared their ideas about retrospectives.

Monika talked about “Gamified Retrospectives” and Mark talked about “Retrospectives, the most boring meetings ever”. Both are very passionate and excellent speakers. Their talks were enjoyable, made important observations and shared interesting ideas about novel ways of improving retrospectives. However both Bob and I felt that they were focused on doing the wrong thing righter. Bob, it turned out, has already had some more thoughts on the topic in his “Retrospectives – Wronger and Righterblog post. We followed it up with a few conversations with others who were slightly concerned about our attitude. After all retrospectives are a cornerstone of many Agile teams. The main argument of the discussions was that retrospectives should reflect on a hypothesis stated before the work has started and too often they don’t.

Another aspect that I would like to draw your attention to is that retrospectives are prone to suffer from the Hawthorne effect. Read the full Wikipedia article but the essence is that

subjects improve or modify an aspect of their behaviour being experimentally measured simply in response to the fact that they know they are being studied

In other words retrospectives may be improving the way we work simply because we run them and not because they result in meaningful improvements to the way our system of work is organised.

You might ask, whether it’s that relevant. After all the team is achieving visible improvements anyway so do we need to care why? Well, we should, we should indeed and the two talks starkly brought this to my attention. They both gave many interesting and innovative ideas about how we could make our retrospectives more engaging, more motivating, different. (And, no, this summary doesn’t do justice to the two talks). So I asked myself: Why do we need to re-invent retrospectives, why do we constantly strive to turn the dial up to eleven? Perhaps that’s because the standard way of running them no longer works.

Our retrospectives are not sustainable.

The workers are being observed and they improve but as they get more used to this observation, improvements fade away so we need to keep changing the observation and experimentation to keep fuelling the Hawthorne effect and this loop eventually runs out.

Instead, let’s focus back on the loop that was always there at the heart of retrospective – the PDCA cycle. By all means, make your retrospectives fun, engaging, challenging, use the great ideas shared by Monika and Mark but make sure that you used them for validation of your hypotheses and to seek to create meaningful, objective changes to the way your work works.