What does Scrum (not Sprint) Performance mean to you?
I'm curious. I believe some confusion exists around what happens within the framework (and it's artifacts thereof) as compared to the output of the framework. I think this stems from the fact that what happens within the framework is very important. What is more important though? What happens within the framework, or the output of the framework?
Using the Scrum Guide as a source of truth (and for now purposely avoiding the identification of who/role that would be responsible):
- What is Scrum framework performance, in your opinion/experience?
- What is your interpretation of the use case for it?
- Should it be something different?
Setting Scrum aside for just a sec, how would any framework consider its own performance? I like taking a black-box perspective. Provide Inputs, let the framework do it's thing, and inspect the Output.
Jumping back into Scrum now, Sprints live within the framework. Naturally, we can use Sprint data to measure success and performance of the Scrum Team, but what about the output of Scrum framework and its performance?
I took a look at the Scrum Definition...
Scrum makes clear the relative efficacy of your product management and work techniques so that you can continuously improve the product, the team, and the working environment.
If you break down "efficacy" it would read: "Scrum makes clear the relative ability to produce a desired/intended result of your..."
- The Inputs would become "product management and work techniques"
- Within the black box is where the framework does its thing.
- The probable reason why "Customer Value" is not represented at all when defining Scrum, is because it's a bi-product of the Inputs being considered in the Scrum Events instead.
- Wouldn't the Output of Scrum be "so that you can continuously improve the product, the team, and the working environment."?
So at a high level, a good/bad determination of Scrum performance would be the capability to "improve the product, the team, and the working environment" existing or not. What then does performance of the capability mean?
- Number of trainings/workshops held?
- Number of processes changed/updated/optimized/removed?
- Number of improvements made to the product?
- Number of improvements to the team?
- Number of improvements to the working environment?
- Etc, etc...
Finally, to introduce the roles into the equation ... how then should a Scrum Master (compared against the Scrum Team or all of Leadership) be responsible for the performance of the Scrum framework?
Lots of questions here but I am so much looking forward to the communities shared experiences and thoughts.
a good/bad determination of Scrum performance would be the capability to "improve the product, the team, and the working environment" existing or not. What then does performance of the capability mean?
I’d suggest it means how well empirical process control is established. Scrum is held to be suitable for complex problems, and is arguably less applicable for others.
Scrum makes clear the relative efficacy of your product management and work techniques so that you can continuously improve the product, the team, and the working environment.
I think a good example of this is a group of developers, business analysts, etc. who previously receive their work from many different channels. Perhaps they receive work from a team lead, a manager, a project manager, or peers.
Suppose these people are then put onto a Scrum Team where they are asked to self organize and prioritize their own work based on a Product Backlog that a single Product Owner is accountable for. The Scrum Framework would likely make transparent (very quickly) whether or not the Scrum Team is receiving other work from leads, managers, etc. that are not related to the Sprint Goal.
To your black box analogy....
1) Inputs - Work techniques and work allocation
2) Framework - Makes evident the Scrum Team's inability to meet a Sprint Goal due to prioritize lots of outside work
3) Output - Missed Sprint Goal and Learning
I think this is where organizations / teams or what have you can succeed or continue to fall into the same traps. All parties involved in the inputs need to understand the output the team is trying to achieve - a releasable product increment - and their behaviors, processes, etc. that are hindering the ability to do so.
So, did the framework do it's job? I'd say yes...but without applying the learning to future inputs and across the board agreement to change behavior, processes, or assumptions the the team will likely see the same undesired outcomes.
I like what you said Tony.
Question. I'm looking at it like there are two levels, Scrum level and then Sprint level separately. In your example, do you see it that way too?
I try to separate the goal of Scrum and the Sprint Goals. I believe they serve different purposes and it's easy to forget there is a responsibility to improve Scrum as a whole, rather than just the Sprints/Teams within it.
I think it's important to start off with the why in these cases. Why bother using Scrum in the first place? If it's to establish empirical process control in order to solve complex problems while delivering value and improving our product, team, and working environment then yes it's different than a Sprint Goal.
A Sprint Goal would be a short term success the team can continue to build on. The team could potentially hit their Sprint goal every time and still not be delivering value or improving any of the above. The Scrum Framework could make that apparent though.
To your point about using it only to improve teams...I've seen things like Agile or Scrum used as a way to 'improve team performance' without the willingness to use it to improve all the patterns and factors outside of the teams control that are inhibiting team performance in the first place. In my opinion, there's only so far a team can improve without help from leadership and larger working environment changes.