Step 2: Run a Trial

After you determine that you and your organization fulfill the prerequisites for self-selection, the next step is to run a trial self-selection event to become familiar with the process and mitigate risks by anticipating and addressing them in a smaller context. A trial in this sense is a scaled-down self-selection with fifteen to twenty people (or two to four teams), which will give you a lot of information at little cost. (If your self-selection involves fewer than thirty people, your life will be easier, and there’s no need to run a trial first. You can go directly to Step 4 .)

We opted to carry out a trial self-selection event at one of our satellite offices, which gave us a more controlled environment and fewer people with whom to test and refine our process. We ran the event with just twenty employees and started with a crudely defined process, which involved iterations, discussions, and lots of sticky notes with names on them. By the end of the day, twenty people had self-selected into three squads, and we knew that we now had a basic process that we could refine to work with any number of people.

All of the employees in the newly selected squads were either in the team they wanted to be in or fully understood the reason why they had chosen differently. The level of buy-in to the outcome was far higher than anything we’d seen before. Had it not worked, we wouldn’t have lost too much either—maybe some of our own time—but we still would have learned a lot about the people and the teams involved.

In order to run a trial, you need only a few things: a group of people you believe are willing to try something new, a list of the teams to form, and photographs of those involved. For a small-scale trial all you need to do is place empty pieces of paper with the squad names on the wall, one for each squad, and hand everyone their photographs, asking them to place their photo in the squad they want to work in. If they don’t come up with a good solution during the first iteration of ten minutes, discuss what’s missing and try another iteration. If you keep it informal and low key, you’ll have the opportunity to learn a lot.

When we ran our initial trial, we formed our first squads in this way, which gave us the opportunity to fine-tune our process for the bigger self-selection event. The main issues highlighted by our trial were these:

• We needed to speed up the process. Going into our trial we believed that a full day would be plenty of time, but we almost had to continue the next day. We didn’t want this process to take so long, so rather than extend the time frame, we focused on identifying ways to speed up the selection process. We realized that better visual management was key.

• We needed to make everything bigger, bolder, and brighter so that it would be easy to interpret the status of a squad at a glance. As you can see in the figure, our early attempts at visualization were not that great. Big color photographs and green check marks and red Xes would be very helpful to indicate whether a squad is complete.

• We needed to create momentum and cadence. We decided to build in specific time slots and focus on moving forward quickly, thus keeping the momentum until the final squads were formed.

• We needed to provide pre-event communication. We found this to be incredibly important, perhaps even more important than what we do on the day itself. We had made assumptions for the trial, and in hindsight it would have made sense to explain things in far more detail and make sure that people knew what to expect. For example, one of the things we missed was offering a detailed explanation of the roles of scrum master and product owner. While focusing on the process of self-selection, we hadn’t put enough effort into making sure that people’s key questions about these roles were answered.

Of all the things we learned during our trial, one of the most powerful was that many of our original fears were unfounded: there were no fights, no people crying in the corner, and no empty squads at the end of the day.