The Journey Toward Self – Organizing, Self-Directing Agile Teams: The Role of the Agile Coach

The agile ideal is for a team to be composed of equals, peers who neither direct each other nor oversee each other, but who work collectively toward a common goal. Team members are self-directing in the sense that they, seeing the goals of the team, choose to work on things for which they are most capable. Since many teams start from a traditionally managed, externally directed world, the journey toward this idea takes time and adjustment for everyone involved. A team typically cannot become instantaneously self-directing.

Typically the journey is assisted and facilitated by a coach who helps the team learn new techniques and overcome barriers. The very existence of the coach suggests that the team is not wholly self-directing, at least not yet; the coach will naturally provide some direction to team members during the learning process. The degree to which the coach will "direct" depends on the progress the team has made toward becoming truly agile.

A typical problem is reflected in a recent question raised by a client. We recommended use of a particular newer framework (SOAJ) in place of the Hibernate/Spring framework familiar to the team, based on our understanding of the goals of the project and the skills of the team; in short, the new framework would reduce coding and increase code quality while improving performance. Some team members felt more comfortable with their familiar framework, however, and the team was divided on the use of the new framework. If the team is self-directing and self-organizing, who makes the decision in the event of a split decision?

It often falls to the agile coach to make these decisions. The coach generally has greater experience and is better equipped to make the choice when the team lacks experience. Coaches must be careful not to become dictators, however benevolent, and teams must take care not to become over-reliant on the coach. Over time, the incidence of coach intervention should decrease as the team gains experience and confidence.

Coaching is more than mere facilitating new practices: there is always an element of leadership in successful coaching engagements. Teams learn best by doing, and the coach must lead by doing as well. Over time, team members step forward and the coach steps back until, one day, the team is leading and directing itself. At first, however, the coach takes more of a directing role.

  1. Mike MacDonagh | April 30, 2012 at 3:02 pm Reply


    As an agile coach with IJI clients in the UK I introduce decision making techniques with teams that help them establish how to make decisions. In simple terms it works like this:
    - identify the different categories of decisions the team needs to make
    - identify (using a variation of the Vroom-Jago model) the mechanisms the team will use to make each decision type
    - facilitate the decison making process reminding the team of their own chosen mechanisms.
    I find that using this approach avoids me making the decisions for them and encourages a mature approach to conflict within the team. Lots of detail here: What does a self-organising team really mean? Organisation!

  2. Kurt Bittner | April 30, 2012 at 8:30 am Reply


    Yes, that’s a potential problem. It can be a challenge to balance the risk that the team, because they are sometimes not yet experienced enough to even perceive the potential problems, cannot make effective decisions, and the risk that they will become overly dependent on the coach for making decisions. The best solution is probably to let them make their own mistakes so they can learn. This takes time, and often looks bad to almost everyone at the time – the team and their management wonder why the coach let them fail (at least in small measure) when the coach clearly knew better. To a very large degree the coach is there to save time from having the team have to make all the mistakes on their own.

    The best solution is one where the team makes, and owns, its own decisions. My philosophy is that teams should be allowed to make their own mistakes so long as they don’t endanger people or the project. The coach can advise and teach, but they can’t decide. If the team chooses the wrong path, so long as it is reversible, so be it.

    A postscript to the original story: after working through the architectural issues, the team itself came to the conclusion that the original framework chosen was not the right solution and they found a number of better solutions. It turned out to be a good team-forming exercise, and they worked through the issues as they proved out the scalability and concurrency issues for the application. This is a good example of how a risk-driven agile approach can work, and validation that the team had come together to work through a set of technical issues. So even though the initial decision was made by the coach, the team stepped up to own the architecture and came to their own conclusions after working through the technical issues.

  3. David Grabel | April 9, 2012 at 3:07 pm Reply


    Team’s will have a hard time becoming self-organizing as long as the coach has any decision making power. The coach should observe the team as it goes through the decision making process, ask powerful questions, and teach the team to resolve their differences quickly. If they can count on the coach to make the hard decisions for them, they will use it as a crutch and never self-organize.

  4. Kurt Bittner | February 10, 2012 at 8:21 am Reply


    Some of you have asked about information on SOAJ. Here’s a link:

  5. Prabhakar Karve | February 6, 2012 at 4:09 am Reply


    I think the “ShuHaRi” approach to learn a new technique could be useful in this context for the coach to progressively take the team towards self-organization.

    You have raised an important question “If the team is self-directing and self-organizing, who makes the decision in the event of a split decision?” As you suggest, coach should be required only till the team becomes truely self-organizing. In that case, after the coach leaves who should have the last word in case of a split decision?

    Any suggestions?

    • Kurt Bittner | March 12, 2012 at 8:59 am Reply


      @Prabhakar: the ShuHaRi approach is interesting, and yes, I think it would be very useful. For others, there is an interesting introduction at

      Regarding the question of who makes the decision in the case of a split decision, this also relates to Stuart’s comment. Yes, it is dangerous for the coach to make the decision. If the split does not affect the working of the team, then a split decision usually means no decision. This sometimes breaks up the team if the division is deep enough. Not an easy situation, and there isn’t a clear-cut answer.

      @Stuart: yes, this is a potential problem. The problem is often time. If the team takes 6 months to investigate the alternatives and decide for themselves, they will learn a lot and will be better for it in the long run. Unfortunately, the project will have failed and the team will have been broken up. There is always the pressure of budget and schedule breathing down the neck of the team, and one cannot always wait for consensus to emerge. Ideally, if the coach is good, they will figure out a way to step back from having to exert stronger influence early in the project so that by the time they leave it is the team directing things. If the coach cannot leave at some point then they will have failed, regardless of the outcome of the project.

  6. Stuart Turner | February 4, 2012 at 3:49 am Reply


    Have you not taken away the risk from the team and placed it onto the coach?

    Now, if the project succeeds, the team didn’t direct themselves.

    If the project fails, the team can, in part, blame the coach.

    There are other ways the team could have made the decision themselves and learned a lot more in the process.

  7. Chris R. Chapman | December 7, 2011 at 11:30 am Reply


    Kurt – I respectfully disagree: The coach’s role isn’t to make these decisions because it requires knowledge of the problem domain that they are likely ignorant about. It’s irresponsible to pull the trigger on the team and force a technology decision on the team that they then have to assume responsibility for.

    What happens if the technology or framework isn’t well-suited to the problem domain? Who takes responsibility for the decision? The team will wear it, but shouldn’t have to.

    A better strategy is for the coach to facilitate the team maturing into self-organizing behaviours by showing how to arrive at a decision within a time constraint. No decision will be 100% perfect, and that’s the point. It is more important for the coach to help the team become more capable in their roles early in the process, when it is most critical. Making decisions for them robs them of this independence and does set a precedent, however well-intentioned, that the coach will save them (irrespective of any desires to never do it again, etc.). You do the team no favours, otherwise.