The selection of the Scrum Master is not defined anywhere in the Scrum Guide. Who chooses them? With which criteria?

Different methods can be observed across Agile Organizations: in some cases, the Scrum Master set up voluntarily and is then appointed by the management; in others, he was born as a member of the Team and is elected within it, and often partly continues to carry out development activities, as a team member.

At the beginning of 2018, one of my clients needed to reorganize the development department to harmonize his teams. His Scrum Masters, elected among the team members, carried out their part-time role and continued to work on development as well. Various problems were emerging: in some teams, there was an overabundance of aspiring Scrum Masters while in others, the role had been played by the person with more seniority of service, without however manifesting a great enthusiasm. Playing a role partially also reduced the focus. It provided an alibi for not devoting time to the study of coaching practices. 

I suggested taking advantage of the reorganization to experiment with a different method of choosing the Scrum Master. 

Considering the existing constraint, which required maintaining the number of people employed (headcount), Scrum Master would become a full-time assignment. Each Scrum Master would be assigned to two teams. Those wishing to become Scrum Masters should apply to an internal Job Posting and pass an assessment interview held by the Coaches who led the transformation. After that, the teams would indicate their preferences among the selected people. The management would finalize the pairings, trying to satisfy as many people as possible.

The idea of having full-time and professionalized Scrum Masters convinced everyone in the company, and the managers decided to experiment with this different method. The top management assured me that Scrum Master would be a role and not a mere job title. With the mantra "Job and Salary safety, Role unsafety," the new organization was announced to the whole company development department.

Selection Criteria

I began to think about what could be the way to select candidates. Scrolling back to the book by Lyssa Atkins, "Coaching Agile Teams," I pondered on the diversity of roles that the Agile Coach must play, and in particular on its three main functions:

• Trainer
• Facilitator
• Coach

I also had a look at "Succeeding with Agile" by Mike Cohn with his acronym ADAPT, which is about Awareness, Desire, Ability, Practice, and Transfer. I had a crazy intuition: if ADAPT works for organizations, why shouldn't it be suitable for an individual?

Therefore, I tried to place the meaning of each letter within the context of an aspiring Scrum Master:

• Awareness of the role and of oneself.
• The desire of the role and how they see themselves in it.
• Ability to read coaching situations.
• Practice. The practical skills as a facilitator.
• Transfer. The ability to transfer concepts, hence to teach.

This skills' checklist seemed to cover the roles of trainer, facilitator, and coach of the book Coaching Agile Teams.

The Interview

The interview was held in a dedicated room. With the candidates' permission, it was audio-recorded to avoid taking notes during the interview and focus on the candidate. 

Evaluation

The evaluation of the candidates' responses was carried out later, listening to the audio recordings and analyzing the photos of the artifacts produced. At first, the critical aspects of the answers were noted. A grade was then assigned on a scale from 1 to 5. A second evaluation tried to harmonize the judgments once all the interviews were listened to have a balanced vote between the candidates. A summary table was then created in which there was a single vote for each component of ADAPT. However, the individual grades were not added to get the final ranking; instead, to try to absorb the complexity of selecting a suitable Scrum Master, those with serious fails were first discarded. Then, case by case, the remaining ones were ordered, trying to evaluate the candidate in its entirety and not based on the mere algebraic sum of votes. The goal was to build a list of ranked people to enter the role to select those who had the smallest gap to fill.

Outcomes

The development department involved had about 100 employees distributed between systems engineers, support groups, and 10 Scrum teams. The goal was to identify 5 Scrum Master candidates.

As stated to the management before starting the interviews, I would consider this process a success if it surprised me. I other words, if the suitable people emerged from it who had not already seemed such since the beginning of the Agile transition. Of the five new Scrum Masters selected, two were former developers. They had managed to overcome in the rankings people who acted as Scrum Masters for about two years, despite not having held this role previously. This showed how much potential and hidden talent there were in them. However, many people excluded from the final choice had demonstrated that they had characteristics of suitability. This was reassuring in the eventuality of having to repeat the process for an enlargement of the workforce or in case of refusals. The new Scrum Masters all showed great enthusiasm, and the first feedback after some Sprints are very encouraging. At the time of the writing of the book book "Scrum in AI," beginning of 2021 - three years after the first use of this practice, I used this approach multiple times in different companies, with a total of nearly thirty candidates. I consider this a valuable tool that helped me identify suitable candidates for several companies. In particular, three times, my assessment showed a not suitability for candidates. Still, they got employed anyway. In all the cases, they showed significant difficulties in entering the role and progressing with the learning.

(partial extract from the appendix of the book "Scrum in AI")