You may be familiar with this meeting by other names such as Product Increment Planning (PIP in SAFe) or Big Room Planning (BRP) or QP (Quarterly Planning).
While there are some variations in mechanics, this meeting happens roughly every three months. Attendees include all teams, stakeholders and management related to a product or value stream.
In some circumstances the teams are not related by a common product or value stream but share a management hierarchy. Or the teams are doing operations work or what is often called BAU (Business as Usual) or KTLO (Keep the Lights On) work. Those environments magnify the dysfunction.
The Desired Purpose
The theory is that by bringing all the people together they will be able to create a plan for the next 3 months. Management usually opens the event describing their aspirational goals for the quarter, or even longer. Teams are often encouraged to document any dependencies on other teams that they discover. Then teams go off in their working groups to develop their plan. Or to simply document their plan as it’s already done in what many people call pre-work or pre-planning.
After the first day management reviews the current plan and makes any desired adjustments. The next day the team adapts as necessary and creates their final plan. It’s common that teams will do a confidence vote on the overall plan to show they support the final result. Which, far more often than not, becomes a “commitment”. After all, the teams voted that they could do it.
Once that is all done the delivery dates within the plan are finalized and then published. Management is now happy because they have a schedule which will be used to hold teams accountable. There will be little tolerance for any deviations from the published plan.
What they have actually created is a Gantt Chart in disguise.
A False Sense of Certainty
I admit that after QP, or whatever you call this event, a physical board that is full of stickies, many swimlanes and columns, red yarn connecting work to visualize dependencies, looks very impressive. You can’t help but think “wow, that’s a lot of stuff” and “look at all we accomplished”!
And therein lies the trap.
Almost immediately after publishing the plan reality has a nasty way of stepping in and blowing things up. The more uncertain your environment, the more unknowns, the faster and more severe the impact.
To create a plan of that length, spanning several to many teams, requires a lot of assumptions of what will happen in the future and decisions made. We are making a huge bet that all of those assumptions and decisions will prove to be correct in order for the plan to be accurate.
The problem is we do not know ahead of time which of these assumptions will prove to be incorrect. We cannot know ahead of time which decisions we made before or during planning is wrong.
No matter how much buffer time teams add for unknowns and unplanned work (priority interrupts) there will be more of it than anticipated. Always. Because it’s not only the unplanned work, it’s the unknowns that will add time to an already overloaded plan. It’s exposing the incorrect assumptions and decisions as the work happens.
We planned assuming everything will go right, or at least mostly right. Once the work begins, teams discover the real requirements, the real unknowns, the wrong decisions. Because the map never matches the terrain.
Within a few weeks reality will drift from the plan, maybe only a little at first but likely a lot in environments with many dependencies and high uncertainty.
Expensive Multi-Day Theater
The more people and teams that attend this event the more costs involved. If you are hosting an offsite event which involves travel, the costs are even higher. Don’t forget to factor in catoring, travel, hotels and all those extended costs.
All in, over two days the teams will spend about 6 hours, or even a little more, doing their planning in small working groups. The rest of the time is presentations to hear a little about strategy and watching other teams present their plans. If the teams are not all working to create the same product attention will wander during presentations that are not related to a team’s immediate goals.
As mentioned earlier, after teams have attended one of these events, they will likely conclude that they need to do some work before the event. So they will do pre-planning. Planning to plan.
As you are calculating costs to determine if the juice is worth the squeeze, do not forget to add in time spent pre-planning. This would not be happening if there was no such event. So you need to factor in the time the team spends preparing for the event.
I have witnessed teams taking 1 to 2 weeks, full time for the entire team, to prepare for the event. That is A LOT of lost productivity to develop a plan whose accuracy degrades quickly.
In one occasion, several of those teams, after 2 weeks of preparation, put their prebuilt plans on their boards right away and then sat back to celebrate. During the draft plan review, near the end of Day 1, the teams had their entire plans blown up. A Vice President said that goals had changed and none of the work they spent weeks planning was needed.
Talk about theater. Talk about waste.
Anti-Agility Weapon
Plans that become commitments are dysfunction.
Worse than the waste is when the plans become weapons. As reality begins to drift from the plan management will pressure teams to get back on track. Status reports and get-healthy plans, if not already in use, will become required. Watermelon reports attempt to hide plan deviation.
The plan has become a schedule. Tell me how this does not resemble traditional project management.
When plans become commitments and become a schedule they disincentivize agility. Teams are far less likely to pivot away from the plan when they should.
If performance reviews and bonuses are dependent on delivery of “commitments” made during planning you can almost guaratee agility will suffer.
Sticking to the plan becomes the goal rather than adapting to change to deliver higher value items discovered after the planning event. Or even pivoting to work items deemed lower value at planning but changing circumstances now indicate an elevated priority.
“Tell me how you measure me and I will tell you how I will behave. If you measure me in an illogical way…do not complain about illogical behavior.”
-Eli Goldratt
Trust Your Teams, Leverage Your Data
Now I will admit that identifying and documenting dependencies has value. It has a lot of value. But there are far less expensive ways to identify and visualize them. Creating a dependency graph depicting the structural dependencies is a great way to start.
It is also my experience that while dependencies are identified and perhaps visualized in this event, little happens as a result. It’s very rare that organizations will do anything beyond develop efforts to better orchestrate the dependencies. They rarely make efforts to remove the dependencies. That’s where real wins and improvements in flow happen, by removing dependencies.
Planning has immense value. As long as you also understand that the resulting plan has little value. It’s a point in time perspective, not an enduring, rigid schedule.
Shorten your planning horizon, one month at most, and adopt rolling planning. The majority of teams are using Scrum and always have a Sprint or two of ready work items. So no pre-planning is necessary, it’s already happening as part of backlog refinement.
Similarly, Kanban teams using an upstream workflow and commitment point will have ready-to-work items for this purpose.
At this point planning becomes more negotiating dependencies and integrations. Actually plan to coordinate and integrate, schedule conversations and put them on people’s calendars. This will take hours and not days.
Use lead time for teams to understand when you need to “get on their schedule” for dependent work. Make the conversations happen, not agreeing to talk someday as happens during QP.
Generate longer term probabilistic forecasts, meaning more than a month, using real team data. Update them often, not once per quarter. Teach others that forecasts are ranges, not specific dates.
Estimates and derivatives, such as velocity, are not needed to forecast. Use lead times and throughput. Leverage Monte Carlo simulations.
By embracing Lean metrics and probabilistic forecasting events such as Quarterly Planning will become overhead in your organization. Expose it as the dysfunction it is.
By doing so you will both increase your predictability and agility while reducing waste. Wins by any measure.
Until next time!
