The very first Program Increment (PI) planning is an exciting time for any team–it is filled with a hope and anticipation of moving towards positive outcomes by embracing agility. Each participating team member has something that they wish to get better at and collectively achieve as a team. This collective desire to work as a team is what moves the organization one step closer towards business agility.
I ran my first PI planning in a similar setup. Getting ready for the event required logistical, content and organizational fervor – our team of coaches and Agile Champion(department lead) had all the help they could get along with a significant amount of visibility across the organizational chart.
We played by the book, ran the leading safe training, helped prepare strategy, vision, roadmap presentations and held regular meetings, leading up to a week before PI planning when we felt that we were in great shape.
For an event with such large in-person participation (five ARTs planning simultaneously), the event went well. In fact, it exceeded our expectations …. the momentum going into the event helped create a positive impact. It also provided an opportunity for geo-distributed teams to meet their counterparts in-person. For this we had planned to extend the lunch hour by 30 mins to include a team-building activity, the marshmallow challenge, which turned out to be quite a blast. It’s an exciting game that encourages lateral thinking with some uncanny results, and who doesn’t like creating tall structures will spaghetti sticks and a marshmallow at the top.
PI planning, which typically spans two days, is meant to be an interactive and a structured event to encourage opportunities for cross-pollination. All teams on the ART are invited, and everyone on the agile team participates. They plan together, discuss dependencies, risks, issues and present a workable plan to the business. Managers and Business owners review the drafts at the end of the first day and their feedback is incorporated in finalizing the plan the next day.
The first PI planning event is always special–it has the new car feel to it. Over the years, different types of PI planning formats have been offered. Here is a peak into some of the formats adopted by different organizations.
In-Person PI planning
In-Person PI planning has been the de facto method as conceived by creators of ScaledAgileFramework. The two-day event has a sample agenda providing guidance on how best to structure the event. This setup assumes that participation is in-person for an entirety of two days and that the entire ART is invited. When all stakeholders are present in-person, systemic and chronic issues plaguing the teams can be discussed and resolved. Important decisions can be accelerated and thus, this setup provides maximum benefit.
Distributed PI planning
When teams on an ART are geographically distributed across various time-zones, the distributed PI planning setup is employed. The event is spread across three to four half-days such that draft plans across various time-zones can be reviewed during an overlap time-slot and assimilated at the central hub. Participation is a mix of in-person and video-calls. Important decisions are still made, however greater coordination (across TZs) is needed to ensure offline conversations are factored into decision making.
Virtual PI planning
The pandemic pushed many organizations to adopt remote work. And, this work from anywhere culture has had a significant impact on PI planning, moving it from an in-person event to a purely virtual event over Zoom or Microsoft Teams. This setup provides the most flexibility and it does so at the lowest cost. Team members can participate without major disruptions to their schedules and could do so even when they are located in a different time zone. However, this set-up also risks losing some of the benefits of PI planning.
One could run into technology failure causing a break in communication, it is difficult to maintain momentum, and hold the attention of all the team members in an online virtual PI planning. Constant distractions, being pulled away to other high priority activities and multi-tasking prevents team members from being fully engaged and feel connected with the team.
Over the years, there has been a gradual decline in the effectiveness of PI planning events. Shift to a virtual PI planning leading to a loss of engagement from team members is definitely one of the reasons, but it is not the only cause resulting in this decline. Some other factors that have contributed to this decline are:
Throw away PI plans and start over
When teams throw away their PI plans and start over with a fresh plan (during iteration planning), then it begs the question–should PI planning be conducted at all? When this pattern is repeated across the ART, it could indicate structural and/or behavioral dysfunctions plaguing the teams. Coming into a PI planning event, teams start from a point of familiarity on the backlog of items. The refinement sessions leading up to the two-day PI planning event prepare teams (and stakeholders) to progressively elaborate the backlog items, create an estimate and a plan that they can commit to.
Iteration planning sessions, conducted subsequent to PI planning, are the final opportunity for teams to review and commit to the plan. However, in many instances, the earlier plan and its deliverables committed to in the PI planning event are discarded. Teams create new plans at the start of an iteration that have no connection with PI objectives or the larger plan they committed to, during the two-day event!
Big room planning
An alternative to PI planning that has gained popularity is the big-room planning event. It brings teams together in a big-room but that’s where the similarity (with PI planning) ends. Each team would create a plan and present it to stakeholders for approval. Other teams present in the room benefit from the direct interaction and may provide feedback. The event can be cadence based and may be preceded by a demo from the just completed iteration. Big room planning adopts the prevailing organizational planning process in a large room setting where teams are brought together. It does not subscribe to PI planning agenda or ask the teams to commit to the plan. It could be run as a strategy and planning session that includes team-building activities, workshops and presentations.
Difficulty demonstrating on-going value / spend
Initial cadence of in-person events conducted every 10-12 weeks may quickly become expensive to run. Unless these events result in marked improvements, there is a tendency to scale back in-person events to once or twice a year and resort to virtual setup for remaining events. As discussed earlier, the virtual setup may further slowdown adoption of agile practices and impede discussions that are necessary to accelerate important decisions.
Spread out over multiple weeks
Parts of PI planning are staggered over multiple weeks due to limitations with securing time slots on the calendar. A few days separate the “Draft plan review” and “Management Review and Problem Solving” sessions. “ART Risks” are reviewed a week later. Stretching out PI planning over multiple weeks may help with busy calendars for some, but such staggered schedules mostly discourage others and they are likely to skip the sessions which further causes communication delays.
So, has PI planning truly lost its mojo? If you ask someone new on the agile release train, who has primarily been exposed to virtual PI planning events, they might tell you that PI planning never had mojo to begin with. However, those with decades of Agile experience will give you a different answer. PI planning can provide ongoing value to an organization when conducted appropriately. This article lays out the important issues, which typically impede the value derived from PI planning events.
You may identify with some or all of these Issues laid out in your organization. The key is to remember that there are multiple ways to work around these issues. For example, setting up an effective LACE (Lean-Agile Center of Excellence) may help preemptively identify the issues and work with senior leaders to resolve them. Encouraging communities of practice and leadership training can help persuade leaders to invest in effective PI planning sessions.
If you have encountered any of the aforementioned or another issue and have found ways around them, do share your views in the comments section below.
Leave a reply