A Good Year - Ending

Our little distributed heterogenous team managed to meet at one place - Bremerhaven - to conduct our first yearly retrospective and celebrate together. Although we did not have an external facilitator, we handled the self-inspection quite well by using a couple of well designed and often practices exercises. We reserved a full day - and it was worth it. I would like to share this experience - which exercises we used, how we managed time and facilitation, how we came to positive results and actions.

Setup

A retrospective is not an Open Space session, although it can leverage Open Space techniques. "Whoever shows up are the right people" is fine for general experience exchange and community meetings. We needed to ensure that "the right people" show up in the first place. So we decided to reserve a full day at a location we all could meet - Pit, myself, our office manager Julia, our accountant Tristan and our student Sebastian. If you want to follow the Lean principles, for instance Optimize the Whole, you should not limit the discovery process of a retrospective to the field players. The scheduling and travel arrangements were not easy, but surely worth it. 

Another thing to regard is whether to include an external facilitator or not. We decided against that, since we have been perfoming retrospectives for years. We were pretty sure that the well-known rituals would guide us through the day. However, this can be a dangerous assumption. You should only self-facilitate a retrospective if you can ensure that you can separate facilitation from contribution. It worked for us, but I think we could have achieved even better results with an external.

Create Safety

I think we did not do this part very well. Instead of an opening exercise, I just quoted Norm Kerth's Prime Directive and we went on to the next step. The opening of the retrospective does not only ensure that the participants feel safe enough to contribute, it's main purpose is to encourage them to contribute. This is particularly important for teams with a variety of characters.

Create a Shared Picture

We picked our favorite exerzise for this part of our retrospective, the Retrospective Timeline. We changed the setup a little to separate events from insights. We cleared our large and long conference table, attached a couple of sticky notes to indicate the months from September to December, and started placing events and artifacts on the table. We wanted to gather events without 'any' evaluation, to avoid premature conclusions and create an unbiased shared picture of what had happened. This exercise took about 30 minutes. The table started to fill slowly, after a while several weird or significant or both artifacts appeared. So I placed my reading glasses on the table to indicate that I had taken over a lot of bookkeeping work during the past months. Some of the sticky notes contained more data, like the number of course participants. That gave us a good base to mine the timeline and dig out insights to talk about.

Gather Insights

Depending on the time you can spend for your retrospective, this part can take up 20 minutes or several hours. We limited the categories to two - "what went well, what could be improved" and started with the practices and competencies that brought us forward in the past three months. After the process of going through the events, it is quite helpful to have this kind of mental rest with a mainly positive focus. The time came soon enough to look at our shortcomings, errors, issues and challenges. We collected each of the sections separately, allowing the group to brainstorm about their points, write them down on sticky notes, before posting them on a flip chart sheet. 

Plan Actions

Now that we had talked so much about what had happened and how we interpreted the outcome of several events, we started to device concrete measures and actions. Most of the actions were based on our identified issues. Nevertheless, some measures emerged out of some good things we wanted to explore further. This section took another 20 minutes for collecting and 10 minutes for discussing and prioritizing the items. Our improvement backlog came to life!

With the improvement backlog in place, we were able to start working on some of the issues, talk about possible solutions or - at least - identify caretakers for the different points. One method that helped us pretty much was Gojko Adzic's effect mapping. We used it to phrase our primary goal for the next year, and identify several actions. I point out some details about the analysis techniques we've been using in this section.

Effect Mapping

We used effect mapping as one of the first "solution mode" exercises after the retrospective timeline, and issue determination. We ran the effect mapping after lunch - not recommended, since creative thinking is a little impeded by digestion. The mapping took not much longer than 40 minutes. The most difficult thing was to formulate the business goal. Stakeholder / contributor identification went fast - so the main part was the "How". 

The first challenge was to formulate a good goal - SMART enough, without being to specific. Not actually a plan, but rather a very concise formulation of the main goal, that was not obvious before. So the starting point is crucial: "What is the goal? Why do we want to change something? How can we measure success? "

It was quite easy to identify possible contributors after that- from training clients to existing coaching clients, partners and ourselves. The next level was a little tricky, as we sometimes confused "How" and "What". The overall process worked good enough to point out some weaknesses in our business performance 2011, and provide learning points for 2012.

Effect maps can be brought to a - seemingly - different level by focusing on organizational development goals. 

Root Cause Analysis

Another issue was approached using a simple Root Cause Analysis of a small but painful accounting problem. We collected the main effects and connected them on a flip chart to identify influence factors and vicious circles. We found out that some VAT difficulties were actually related to our high percentage of travel expenses not directly covered by the company. Picture that. 

Domain Modelling

Well, this is not exactly a retrospective technique. But a simple drawing containing a couple of boxes - our classes - and lines - their relationships, made painfully clear that we need to improve our training management system(s). I won't go into too much detail, but 8 classes spread over 3 separate systems with manual transfer between them surely gives lots of improvement opportunities. Thanks to Zsolt Zsuffa for pointing out the importance of light weight domain modelling for agile teams.

Closing the Retrospective

As with many events, the most remarkable and reminded parts are often the beginning and the end. After we'd identified the next steps after our analysis, we added dates and point persons to ensure the improvement backlog is worked on. But that was only the formal closing. We ended the day with a nice little celebration at a great fish restaurant in Bremerhaven. The retrospective is no leisure time - it's the time-out, not the break. So adding a social event to an important retrospective rounded up the day. 2011 is almost over now. We are looking forward to a new year with a strengthened team - and a full improvement backlog.

Andreas