What happens when you apply the Lean principles of Kanban with those of Complexity? Kanplexity!

In our recent Lithe Talks, John Colemanscrum.org/john-coleman – Professional Kanban Trainer and Professional Scrum Trainer – talked us through Kanplexity, which relies on the concept of sense-making developed as part of the Cynefin framework. John specialises in transformations for business agility using change models based on change leadership, change evolution and behavioural change.

Using tenets of Kanban and Cynefin® sense-making, Kanplexity offers an optional approach to those in non-software settings. Sense-making is a mental exercise that evaluates situations against our understanding of how the world works.

The origins of Kanplexity came about when John Coleman was working with an energy company where the team were told to use Scrum. When John was working with the team, he found that the personal values of the team members were completely different and strong in their own right, which resulted in work taking a longer to finish than needed. John used Cynefin as the leadership sense-making framework to guide decision making.

What we lack, is guidance for potential agilists in other sectors; we must strike a balance between the competing forces of doing the right thing, the right way, quickly, (more) predictably, repeatedly, and sustainably (inspired by Magennis, 2022). We should think and act appropriately for the complexity in front of us and only have formal interactions needed for the situation at hand.

Let’s have a look at how the two concepts meet together to make an agile and adaptive system that you can implement within your organisation.

The Cynefin Framework

A sense-making compass, Cynefin offers five decision-making contexts or "domains" – clear, complicated, complex, chaotic, and confusion.

The Cynefin Framework was developed to help leaders understand their challenges and to make decisions in context.

By distinguishing different domains (the subsystems in which we operate), it recognises that our actions need to match the reality we find ourselves in, through a process of sense-making. This helps leaders cultivate an awareness of what is complex and what is not, and respond accordingly so that no energy is wasted in overthinking the routine. This does not involve trying to make complex decisions fit into standard solutions.

Kanban

Using the Kanban framework is a strategy for optimising the flow of value through a process that uses a visual, pull-based system.

There may be various ways to define value, including consideration of the needs of the customer, the end-user, the organisation, and the environment, for example.

Kanban tries to strike a balance between effectiveness, efficiency and predictability:

Efficiency: what are we doing really well?
Effectiveness: are we doing the right thing? And is this the right problem to be focusing on right now?
Predictability: in terms of workflow, how long are things going to take and how much work can get done?

It is also important to ensure that the predictability over time and the complex space predictability cannot become a second citizen because learning is the first citizen in the complex space in Kanban.

Mind the flow: Operating an efficient workflow

Kanban comprises the following three practices working in tandem:

Making work transparent: Defining and visualising a workflow (making work transparent)
Making policies explicit and actively managing items in a workflow
Limiting work in progress, which leads to improving the flow of work

In their implementation, these Kanban practices are collectively called a Kanban system. Those who participate in the value delivery of a Kanban system are called Kanban system members.

The definition of workflow in the Kanban Guide is clear: what’s the minimum you need on your workflow?

The key is to define what the individual units of work are.

There can be more than one flow on a Kanban board, but the key is to define what the individual units of work are.

At a minimum, system members must create their DOW (Definition of Workflow) using all of the following elements:

A definition of the individual units of value that are moving through the workflow. These units of value are referred to as work items (or items).
A definition for when work items are started and finished within the workflow. Your workflow may have more than one started or finished points depending on the work item.
One or more defined states that the work items flow through from started to finished. Any work items between a started point and finished point are considered work in progress (WIP).
A definition of how WIP will be controlled from started to finished.
Explicit policies about how work items can flow through each state from started to finished.
A service level expectation (SLE), which is a forecast of how long it should take a work item to flow from started to finished.
So, how can you operate an efficient workflow in Kanban?

Optimise the flow of value

Through the Kanban workflow, there will be a flow of work items that have a value, which is different to Scrum, where there are product backlog items going through cycles (often called Sprints). An effective kanban workflow contributes to the generation of value but in and of itself does not generate value. With a strong kanban workflow, we optimise the flow of value, we don’t optimise the flow of activities.

Focus on finishing

John Coleman advises that we want to get people to focus on finishing. We want to get people focusing on finishing work items, rather than progressing subtasks and many items at the same time, with them likely finishing at the same time in a big batch within the designated cycle time (the start and finish point, which can border different columns and hence, also result in multiple start and finish times based on backlogs and feedback).

Avoid perfectionism

Each workflow needs to have the states defined: What does each column represent? Furthermore, you also want to get as many signals as possible. This isn’t about perfectionism, however.

“Do note that you can’t have a perfect board as there is no such thing, and if the Kanban system members don’t love the official workflow, they are not going to use it. So, be careful about going for perfection…” John Coleman

Maximise the signals

You want to try and maximise the signals going from the starting point to the finishing point. These signals and touch points will help you understand where the bottlenecks are and stopping the workflow from moving between the columns and finishing.

“The more signals you get from your system, the more likely you're to respond to that and hopefully get better feedback loops as a result.”

Have robust policies

Having some clear policies and criteria within the workflow can also be beneficial in understanding how the flow of activities should move between columns.

John Coleman in his talk gave the example of how he was once asked if an item could be moved from one column to the next, to which he responded “I don’t know, what does your policy say?”

Having these policies in place can help system members understand the criteria of each column and what steps need to be completed for the task to move within the columns. Having this defined can serve as clear exit criteria to finish the task.

Work in Progress

John Coleman also encourages us to have a definition of how Work In Progress (WIP) will be controlled from start to finish. Work in Progress is the number of work items started but not finished.

This could be that you have Work in Progress limits or individual podiums or for different teams there are different swimlanes which in turn have WIP limits themselves. And with a service level expectation on how you’re going to forecast completion of these items. The trends for this can be determined by historical flow metrics:

How many items started but didn’t reach the finishing point?
How long was the Cycle Time from the starting point to the finishing point?
What was the amount of time between when a work item started and the current time? This applies only to Throughput.

Practices in Kanban

There are three practices in Kanban.

The first practice is defining visualising the workflow. Visualisation of one or more workflows is essentially the Kanban Board. Kanban helps us optimise signalling to see what we need to do to help our workflow.

Using Service Level Expectations, which is initially based on a guess and then on data, can help us understand how many projects can be delivered in a given timeframe and also which ones are likely to require longer cycle times if they are not in the usual boundaries of data.

Focusing on the following allows us to navigate complexity more easily:

What active work should we focus on or bring in today?
How much work we’re comfortable doing at the same time
Our aspiration for more predictability
Visualising:

relatively aged work
blocked work
dependencies, including those neither aligned or acknowledged

The second practice in Kanban Guide is active management of items in a workflow is about Kanban System Members addressing items to focus on, unblock, finish and feed. For all intents and purposes, this is based on focus and finishing the tasks at hand.

By reviewing what the members need to work on together today, they continually refresh their thinking to address the complexity they’re facing.

The one thing that John Coleman loves and encourages about Kanban is that it allows slack time for thinking and unplanned activities. Allowing for a bit of slack in the system means system members can work together to finish tasks and then move on to the next.

When the Kanban System Members release items sooner, feedback loops also get tighter.

The third practice (often forgotten!) is improving the workflow(s). Kanban systems members can change the definition of workflow(s) at any time, including which columns on the Kanban board to add or remove. However, the members might need to strike a balance between allowing the system to settle to observe trend changes following a policy change and adapting to the current reality by making multiple changes in the definition of workflow.

In order to improve the workflow it is important to ask the following questions:

Is this the right board for you?
Do you have the right WIP limits?
Do you have the right exit criteria?
Do you have the right pull or move policy for more items and workers?
How are you prioritising items?
Have you reviewed your block policy?

Everything is and should be up for debate as it is important to step back and check if things are on track, particularly assessing the different signals and feedback from the board.

Kanban and Lean UX

Combining Kanban with Lean UX can strike a balance between effectiveness, efficiency and predictability. When you combine the two of them together, the signalling is already capitalising the frequency of feedback.

A lot of teams make the mistake of focusing on delivery. According to John’s estimation, 60 - 90% of items should not be built. He gives the example of using discovery and Lean UX Canvas as a tool to look at the different business perspectives of Problems, Solutions, Outcomes and the Hypotheses around the tasks to prioritise what is absolutely necessary to put on the board.

Using discovery, you can weed out some of the ideas that maybe you shouldn't deliver and you might discover much better ideas.

John Coleman conducted this with 240 students in UCL earlier in the year. What he found remarkable across the 36 teams (circa 160 of them were in the room and 80 of them were online) was that people were inclined to persevere, even though the evidence was compelling that they should change direction. This further highlights that you need some kind of opportunity to step back and reflect and ask if this is the right thing or the right personas and prototypes to keep going in the same direction or change course.

Red vs Blue work

John Coleman references David Marquis’ metaphor in his book “Leadership is Language”, oscillating between blue work (review, retrospective) and red work (doing the work).

Red work is about being the executors and doing the work, and the blue work is about more about stepping back and reflecting and thinking what we want in the right direction.

This can be demonstrated in Scrum with Kanban, where authentic Scrum meets authentic Kanban, and offers the opportunity to focus on blue work alongside the increased frequency of feedback loops allowing for the review and taking the step back.

“The key is to look at data informed decision making…”

This is quite a popular approach which can inject ‘rocket fuel’ into your product management systems, but one which needs to work with Lean UX so that the rocket goes in the right direction. The challenge is using these three systems collectively, so it is imperative that they are applied in their full form rather than picking and choosing elements as that can affect results.

Kanplexity

Kanplexity offers more flexibility in this regard. Instead of ‘inspect and adapt’, you get ‘sense-make and respond’. This does involve a team because for complex work, that element of trust is needed.

Kanplexity uses Cynefin to guide the team and the guide, which is the agile leader, who acts with the team. This team does the work and in the face of complexity, is small, diverse, self-managing and cross-skilled where members care about each other’s work and learn how to do their work. These team members give, reward and earn trust.

In Kanplexity, executives need to cultivate an environment where the team is ready, willing and able to work in the problem space, discovery, delivery and value-validation.

“Organisational agility is distinguished by how well the environment is cultivated for agility to grow”

John Coleman feels really strongly about fostering an environment for organisational agility by:

Listening to what frustrates people
Discovering value, weeding out what’s not and experimenting to better value
Delivering value
Decluttering workflows, processes and systems
Organising for value
Changing the ecosystem of decision-making
Decoupling leadership from the position of leader, because leadership is everywhere

In Kanplexity, a person acting as a guide, monitors the Cynefin Compass for the team or crew and facilitates the discovery of “the next right thing”. For this guide, leadership is not necessarily designed as a position. Provided they behave as per the tenets of Kanplexity, examples of guides might include change agents, coaches, mentors, and so on.

The guide has to:

Be careful to avoid too much focus on the process at the expense of value
Promote value validation over output
Resolve issues that teams and crews cannot solve.

What makes a good guide? “Someone who is curious, someone who is humble, open and someone who promotes psychological safety…”

Kanplexity also has a direction of travel. A direction of travel is a loose purpose that can flex depending on what we learn; the direction of travel evolves, pivots or stops. While goals are helpful for complicated work, it’s crucial to inform decision-making with evidence for and against the goal’s validity.

The direction of travel should be clear to everyone and the team, as well as to customers and focuses on the short and medium-term possibilities; it should be revisited during reviews as new evidence is discovered.

Kanplexity also encourages rhythm, which of course helps reduce complexity. However, it is key to note that in Kanplexity, the old adage of “everything in moderation” has to apply: finding the balance between rhythm and avoiding a one-size-fits-all approach. Too little rhythm and too much rhythm can be a problem.

Kanplexity has five interactions, each without a time box:

Replenishment (just in time) - as needed: in Replenishment, teams can use Throughput or probabilistic forecasting to guide selecting the number of items they can reasonably get done in a time period, e.g, a cycle.
Cycle - recommended for complex and liminal complicated-complex work
Standup - recommended for complicated, complex and liminal complicated-complex work: a review of blockers and relative work item ageing can help manage items in progress in the direction of travel
Review - recommended for complex, liminal and positive chaos: probabilistic forecasting can help manage expectations, with its caveat that we’ll have more accurate forecast for the next week, next cycle or next month
Retrospective - recommended for liminals, complicated and complex work: encourages the team or crew to assess where work is getting stuck, the definition of workflow and monitor their aspiration for the service level expectations against real world data

“Agility is not a team sport, it’s a company sport” - Dr Klaus Leopold

Agility needs to cut across value streams, which is what Kanplexity hopes to achieve.

In Kanplexity, in the absence of a Scrum Master, the team or crew should be clear about facilitation and preparation for each of the following:

Pruning, refining, full-kitting or rightsizing the work as needed for all domains
Replenishment, as needed for all domains
Standup for complicated, complex and liminal complicated-complex work
Review for complex, liminals and positive chaos
Retrospective for liminals, complicated and complex work

Having explicit policies that cover the above is crucial to ensure the direction of travel.

In summary, Kanplexity:

Offers cycles and interactions that catalyse more frequent feedback loops and the opportunity to course-correct
The Guide cultivates an environment where agility can grow
Cynefin is the compass for decision making
Iteratively revisits the problem space
Discovers to deliver better ideas
Builds empathy upstream and downstream to allow for work items to operate at different levels of granularity

It is important to remember that context is king.

Kanplexity is deliberately flexible, but not so flexible that it enables simple ‘rebranding’ or non-agile work.

John Coleman hopes that Kanplexity helps people who are authentic about agility, whether in product management, project management, Lean, Agile, Theory of Constraints, DevOps, The Vanguard Method or otherwise.

Kanplexity recommendations are based on the results of many experiments and benefits. Being open to the recommendations helps us avoid the road to negative chaos.

For more information and to read the Kanplexity guide, please see the website of Orderly Disruption.

“Kanplexity” – Kanban for Complexity