The adoption of SAFe as a framework to scale up agile in organizations has been gaining traction, and it’s not difficult to see the appeal. One of the hardest things to do is to let go of control, and this framework injects the good old reliability of waterfall into agile teams. Now, since not all my articles can be rages against the machine, I won’t go into the Agile vs. SAFe discussion, but rather take the opportunity to look at a practical way to take baby steps towards Agile through the Program Increment planning.
These are my takes on PI planning and getting the most agility out of the teams.
The key goal of a good PI planning is the alignment of the business goals across all the teams. Of course, this means the business can provide features that will be crucial for the next period. A pitfall here is when they try to come up with a complete planning to fill up the teams; then the focus shift from required value to filling backlogs.
It should be added here that a team can have a lot of ‘local’ value. One can think of technical debt, proof of concepts for team innovation, quality and security enhancements, initiatives that fall outside of a global scope.
Try to not align all aspects of what a team does, give them room to develop and innovate on their own. Independence and responsibility are core values of agile.
I’ve now seen several times that SAFe gets adopted from a situation where teams aren’t communicating as much as they should on topics where they are dependent. The PI planning is then adopted as the means to get them to communicate better. Then when it comes to the PI planning, it’s organized to such a degree that team members prefer to lean back and let it happen, apparently it was all thought out.
A central part of PI planning is fine: it gives stage to present some ideas. But when the topics for alignment are presented, lean back and only facilitate. It can happen that three teams, an architect and a business analyst need a couple of refinement sessions and time to investigate, give them the room to do that in their own time.
A PI planning should not have to be two consecutive days to get it done, if you take two half days a week apart and give freedom to use the time in between to refine then that should be fine.
Know where the value lies
Those who facilitate the PI planning rarely provides the value you need for the next increment, though they often like to take the stage. There’s an interesting effect that out of a ‘panic’ mode these sessions are filled with fillers to keep everyone entertained, and the entertainment leads to less input from the knowledge holders. It can get to the point where the trust in the team’s value is broken down, in favor of architects designing everything and going in a complete waterfall all over again.
If you want these sessions to work, ensure that they are a collaboration. Get every one at the table – from business owner to architect, team members and any other necessary roles – and facilitate an open conversation. The more each member is encouraged to add value, the better prepared you will start the next increment.