What is Large Scale Scrum (LeSS)?
What is LeSS?
LeSS is a de-scaling framework. Also referred to sometimes as a scaling framework.
One of the principles in LeSS is queuing theory. Kanban comes from a similar place in that you try to stop starting and start finishing.
The more work that we have running at the same time, the higher the cycle time goes up, the longer work takes and the less throughput that we have. Large Scale Scrum is still Scrum, although there is one case study that is non-Scrum.
My thoughts: I think there are possibilities for a Large Scale Scrum implementation to be without Scrum in careful hands. Like Scrum, Large Scale Scrum is based on transparency.
Being open about what’s really going on, looking at what’s really happening, and it’s also based on empiricism.
Empiricism is where you decide what to do next based on the evidence and the learnings from what you did last.
In Large Scale Scrum, we try to achieve more with less. Instead of providing something that’s really built out, we try to provide a barely sufficient framework.
So as well as being based on a set of principles, LeSS has some rules. A minimal set of rules so that chaos does not prevail.
Large Scale Scrum has a whole product focus.
What is a product?
My thoughts: For me, a product is something that an external customer or end user would recognize and pay for with:
- their money;
- their time; or
- their data.
Someone consumes it and someone produces it, and it’s a vehicle for delivering value.
The role of the Customer
In Large Scale Scrum, the customer is at the center.
For example, in product backlog refinement, often the customer and the end user clarify problems to be solved and opportunities to be availed of directly with developers in the teams.
How does it differ from Scrum?
Unlike Scrum, Large Scale Scrum does not have a product goal. It doesn’t have a vision, but there is a perfection vision for how we do the work.
Just because we can’t achieve perfection doesn’t mean that we shouldn’t strive for it. In the same way that, for example, a car manufacturer strives to build a vehicle in one cell, it might never get there. That doesn’t mean it’s not worth pursuing.
Large Scale Scrum and Lean Thinking
Large Scale Scrum like Scrum is based on lean thinking. We’re trying to eliminate waste and at the same time, we are not so focused on efficiency, we’re more focused on effectiveness and adaptiveness.
In fact, in LeSS optimizing a goal is essentially adaptability and adaptiveness.
Central to Large Scale Scrum is systems thinking.
A common practice within Large Scale Scrum is system modeling where you talk about a problem, you can have a conversation about a problem and a skilled system modeler can draw a system model of what the conversation is.
By modeling, you can find virtuous loops and you can also find vicious loops. Some loops make things better, and some loops make things worse.
My thoughts: We have to be careful about system models because in Kanplexity™ we have unknown unknowns, and so we don’t want to be going along with a delusion that we understand everything about the system.
Do we ever understand everything about the system that we’re operating in?
We can use it as a social exercise to align our assumptions about how the system is working so we can try to find those key variables so that if they change, they will actually change the behavior of the whole system.
We call them driving variables.
You might find some negative driving variables and some positive driving variables.
You can also discover some potential interventions that change the health of the whole system, and it’s a fantastic way of getting alignment about how to improve the way we work towards that perfection vision.
Large Scale Scrum is a framework. It’s a barely sufficient framework and it has officially parted ways with the 2020 Scrum guide.
My thoughts: I personally like the 2020 Scrum guide. See the blog post I wrote about it here. I isolated the differences between the 2017 guide and the 2020 guide. I agree that the definition of product is suboptimal in the scrum guide.
LeSS supports the 2017 version of Scrum
The people in the Large Scale Scrum community have mostly aligned around the 2017 guide and officially, LeSS supports the 2017 version of Scrum.
If you look more closely at LeSS, it really feels like a more old-fashioned Scrum.
My thoughts: There is a Scrum primer and LeSS is more aligned to the Scrum primer than it would be to the Scrum guide, in my opinion.
Differences between the Scrum Primer and LeSS
So what’s different about LeSS?
In sprint planning, there is:
- sprint planning part one; and
- sprint planning part two.
In modern scrum, we have sprint planning. Topic one, topic two, topic three, where you’ve got the why, the what, and the how.
Sprint Planning Part One
In Large Scale Scrum, because we are talking about many teams working together, in sprint planning part one, essentially the product owner would be meeting team representatives and a scrum master to talk about what might we be planning to do in the coming sprint and what planning can each individual scrum team do?
The entire team can come if they want to, but it’s usually team representatives with the product owner and one of the scrum masters.
Sprint Planning Part Two
Then, each team does its own sprint planning in sprint planning two.
In sprint planning two, the teams might do their own individual planning, but the teams could also decide to plan multiple teams together because it makes sense for them to do that.
That’s a big difference from Nexus because in Nexus we don’t have product backlog items that go across different Scrum teams, but LeSS does allow for that.
Product Backlog Refinement in LeSS
It has a similar structure. There is overall product backlog refinement where the product owner again meets team representatives and a scrum master and they talk about what work might be coming up in the upcoming sprints.
We don’t know for certain what’s going to be in the upcoming sprints, but we’re trying to line up our ducks in terms of what might be happening so that we have some level of professional preparation done before sprint planning for the upcoming sprint. We don’t want to be surprised.
There might be some new ideas that need refinement, and that can be done last minute in sprint planning. Although, when you have got multiple teams working together, it’s really important we look ahead. If the teams are very good at doing product backlog refinement to support overall backlog refinement, then the product backlog will be in good health.
Product backlog refinement can be done at a single team level but these days there’s a trend towards teams working together on product backlog refinement, and you can think about maybe having product backlog refinement where people might do a shift and share, where they go from one session to another to learn what other groups have learned from their product backlog refinement.
It’s not a playback to each team, so everybody learns what’s happening in each session. It’s more about what did people learn in their individual product backlog refinement sessions?
This is because in Large Scale Scrum we’re striving for ‘slice of cake teams’ as I refer to it. Teams that deliver value that can be consumed by customers and end users.
And to enable that, we have teams that are truly cross-functional.
Dependencies in LeSS
In Large Scale Scrum, we are trying to eliminate dependencies and we are less concerned about dependencies that we can just do simultaneously. The only dependencies that we really are concerned about would be the asynchronous ones where one thing has to happen before something else.
We try to minimize or eliminate that by making sure that we have all the required skills in the team.
So in LeSS we still have a daily scrum and one slight riff from Scrum is, people from some scrum teams can visit the daily scrum of other teams. That seems sensible.
The coordination approach isn’t just talk, it is a little bit more elaborate than that.
If the teams are delivering software, they’ll be using continuous integration and technical excellence.
When we say continuous integration, we mean that when the build is broken, it’s like there’s a fire in the building. I’m slightly exaggerating, but it doesn’t mean that all the code is on Jenkins or some other type of continuous integration build server.
It means that we care when the build is broken because we don’t want to have branching. We want to use trunk-based development so that we can be only hopefully two hours away from production. If there’s an issue in production, we can fix it within two hours without having to eliminate code.
Sprint Review in LeSS
In Large Scale Scrum at the end of a sprint, just like in Scrum, there’s a sprint review and representatives from all teams should be there, ideally as many people as possible, as well as the customer’s end users or the stakeholders.
Just like in Scrum, it’s a working session, so it’s not a boring PowerPoint session.
People are encouraged, customers, end users, stakeholders are encouraged to play with the latest increment.
We might even be looking at the results of feedback:
- from experiments;
- from interviews;
- from experiences of people using the product since it was released during the sprint because you could actually be releasing during the sprint as well.
And there’s discussion about what might be the next thing to work on and what are the new insights from the market and so on.
We have a look at how we are doing overall in terms of delivering value in a really customer-centric way because LeSS has a true customer focus and product focus.
After the sprint review, you have the retrospective. Each team has its own retrospective.
My thoughts: What I really like about LeSS is that there’s an additional event called overall retrospective, and the idea of the overall retrospective is that managers, leaders would attend along with representatives from teams as well as a scrum master and the product owner, and they would typically talk about issues, problems, risks, things that are slowing teams down that are beyond the control and influence of the teams.
The teams need help to resolve some of these things. The teams in their own retrospectives will have come up with some improvements and they might even have done some of those improvements immediately. They might have done some system modeling, for example, to understand the system more, or they might have used Open Space, LeSS isn’t really prescriptive about how you do that.
But what’s important is that there’s improvement, and I think this is really nice because it’s unlocking this message that agility is improved when executives help teams to solve problems that are beyond the control and influence of the teams.
After all, agility is not a team sport as Klaus Leopold said.
Large Scale Scrum, a barely sufficient framework, a de-scaling framework based on principles, a small set of rules. It’s where you can start. It’s not necessarily where you end up. You might end up with something even much better, but what a great start.