BT

Facilitating the Spread of Knowledge and Innovation in Professional Software Development

Write for InfoQ

Topics

Choose your language

InfoQ Homepage Articles The Sustainability Agenda in Kanban

The Sustainability Agenda in Kanban

Lire ce contenu en français

Bookmarks

This is the first article in the series “Nine Values, Three Agendas”, which covers the changes Kanban is capable of bringing about when a particular adoption approach is taken.

The Kanban Method (often just “Kanban”) has from its early beginnings been described as an evolutionary approach to change.

This was reflected in the title of David J. Anderson’s book Kanban: Successful Evolutionary Change for Your Technology Business. Another apt description would be this:

             Kanban is the humane, start with what you do now approach to change.

That humane side is captured in a system of nine values. These explain the motivations of the method and they provide a helpful starting point for exploring how it actually works. Each value identifies a goal of one or more of Kanban’s principles or practices and suggests some of the benefits to be derived from following them. Conversely, they suggest that the pursuit of those goals and benefits may be well served by the adoption of their respective method elements.

Neither the presence of this value system nor Kanban’s special emphasis on what you do now should be taken to imply any lack of commitment to change. On the contrary: change is what Kanban is all about! Unsurprisingly though, we have found that different organizations respond to Kanban in different ways and use it for different purposes.

We have identified three main “agendas for change” that represent three main patterns of engagement. These agendas work in a similar way to the values, characterising the kind of changes Kanban is capable of bringing about when a particular adoption approach is taken. Their purpose is to help organizations and change agents make contextually appropriate choices about where and how they will start and to guide what follows.

The values align to the agendas in groups of three:

  • The sustainability agenda draws on the values transparency, balance, and collaboration. It describes a common approach to Kanban adoption at the level of individuals and teams, often motivated by the need for relief from unsustainable practices and workloads.
  • The service orientation agenda brings together the values customer focus, flow, and leadership. Building on the sustainability agenda, the service orientation agenda describes a more outward-looking approach to change.
  • The survivability agenda is the most overtly cultural agenda of the three, and the most ambitious. Its values understanding, agreement, and respect represent commitments and disciplines that support organizational survival strategies based on adaptability.

Because the values correspond to principles and practices, they can be used to help explain not just what the agendas mean, but how they actually work. Welcome to the “nine values, three agendas” model. We’ll look in detail at the sustainability agenda in this article and at the other two in future instalments.

The sustainability agenda

Initially, I called this “the continuous improvement agenda”, a name that does a poor job of conveying the relief that Kanban can bring to unsustainable working environments. Of course it can be hard to adopt anything new in such environments, but even here, the sustainability agenda’s values of transparency, balance, and collaboration can be embraced where other approaches have been deemed unworkable.

Value: Transparency

The Kanban Method owes its name to a Japanese word meaning (roughly) “signboard”; its strong emphasis on visualisation should come therefore as no surprise. Its most important visualisations are kanban systems, a collective term that covers a broad range of visual management systems and techniques. They have in common the use of visual tokens (“the kanban”) that represent work items and mechanisms to coordinate and smooth the flow of work.

In knowledge work - technology development and support, HR, law, media production and so on - much of what we do is invisible; it takes place inside our heads and on our computers. Good visualization can make knowledge work very much easier to manage, but given its diversity, what constitutes a good visualization depends very much on context.

I have based this example on a composite of real-world examples of which I have direct experience. Details vary from board to board (and they evolve over time) but it is quite representative of kanban board designs used in software development:

 

A typical electronic kanban board

A guiding principle of this now-popular form of kanban system is that it organizes work, not people. There is nothing about this particular board’s design that indicates whether it supports one team or several, or whether people are organised by (say) function or by product line. Here, the colors organize work items by type, highlighting some dominant risk dimension or class of service. The columns organize work items according to their current need or state; as their needs are satisfied and their states change they move rightwards towards completion. In this way, the kanban system captures the essence of a high level workflow.

With this deliberately work-centric approach, kanban systems such as this encourage self organization. People have readily available the information on which good decisions can be made; effort gravitates quite naturally to where the need is greatest. This process of self organization doesn’t have to stop at team level: it can extend up through the organization as transparency makes it apparent that existing structures are not up to the challenge.

As people self-organize around the work, their perspectives change. This prompts improvements to the visual design of the kanban system, or ­– as the limits of visual language are reached – the crystallization of written policies, such as per-column “definitions of done”, quality standards, and global prioritization policies.

Such lightweight and easily-maintained expressions of process help to make process change a cheap and high leverage activity. With well-designed feedback loops built into and around the delivery process, the improvement process is locked in.

Value: Balance

The coordination mechanism in kanban systems of this kind is the work-in-progress (WIP) limit. WIP limits are constraints applied the number of work items present in each state.

WIP limits can be applied at the highest level of organization (limiting for example the number of initiatives pursued concurrently) but they are most often implemented at levels that bring obvious and direct benefit to the people actively engaged in the work. Their effect can be dramatic when previously overburdened parts of the system are protected, bringing the number of unfinished work items into balance with the supply of effort available.

Controlling the amount of unfinished work in the system does more than just reduce multi-tasking:

  • It provides system support to the old adage “stop starting and start finishing”. We find that WIP limits encourage a strong psychological bias towards successful delivery and an aversion to premature commitment, permitting further reductions in WIP.
  • With limits are applied not just to activities but to the queues between them, highly effective pull systems are created. As work moves downstream towards completion, the availability of capacity is immediately signalled upstream (see figure), keeping systems properly coordinated end-to-end.
  • With less work in the system, the time it takes for individual items to move through the system is both shortened and made more predictable.
  • Stuck or otherwise problematic work items are made that much more visible; issues have fewer places to hide. As well as prompting direct intervention, opportunities for further improvement are brought to the surface more often.

There is capacity in TEST, allowing an item to be pulled across from BUILD Done

The pursuit of balance manifests itself in mature Kanban implementations in other ways. These include:

  • Actively managing the mix of short-term risks, maintaining for example an appropriate balance between urgency-driven work and deadline-driven work.
  • Deliberately allocating effort across short, medium and long-term time horizons.
  • Explicitly considering the interests of customers, those inside the system, and the wider organization when evaluating potential improvements. As a rule, “improvements” that benefit one of these stakeholder groups at the expense of another should usually be rethought.

Each of these contributes towards sustainability at individual and organizational level.

Value: Collaboration

The collaboration value represents the core practice “Improve collaboratively, evolve experimentally”, to which is often added the rider “using models and the scientific method”.

As the means to deliver improvements, collaboration takes the form of creative, problem-solving relationships that involve interactions of high quality. We see collaboration as an end as well as a means however, a tool for achieving improvement and something to nurture for its own sake.

Collaboration as an improvement focus

Low quality interactions ­– inspections, reviews, handovers etc – often add significant delay and unpredictability for poor return. More collaborative approaches – pair programming being a classic example – smooth and simplify the process, make feedback much more timely, and spread both knowledge and skill. Add to those benefits the basic human need to relate to others in meaningful ways and the advantages of collaboration seem clear.

And the small print? “Using models” can refer to several things:

  • The application of external models such as Agile, Lean, and Theory of Constraints
  • The development of internal models of how things currently work and perform
  • The development and replication of exemplars of successful change

Each of these has its own priorities, heuristics and tools. The ease with which multiple models can be integrated will depend to a large degree on the alignment of values and core assumptions. Practices can be made to fit once the underlying principles are understood; incompatibilities between values cannot be glossed over so easily and choices may need to be made.

“Scientific method” refers to the disciplined application of experiment-based improvement, framed stereotypically as plan-do-check-act (PDCA), plan-do-study-adjust (PDSA) or similar cycles.

Implementing the sustainability agenda

With its echoes of the Agile manifesto’s “sustainable pace”, the sustainability agenda resonates strongly with Agile teams. Moreover, many of their existing practices and artefacts can be interpreted in terms of transparency, balance or collaboration, the last of these being of course a key theme of the Agile manifesto.

Overlaying Kanban on existing processes at team or departmental level helps to bring supply and demand into balance, simultaneously delivering gains ­– often dramatic – in performance and predictability. Collaborative, model-driven improvement then deepens and sustains these benefits into the future.

It should be acknowledged that it can be hard to maintain the momentum of change. Kanban implementations aren’t completely immune from this, but they do have much in their favor:

  • Their visualisations and feedback loops bring the need for change to the surface
  • Their systems are cheap to change and the effects of doing so are visible quickly
  • Awareness for the need for change is further heightened as WIP is lowered
  • Collaborative, model-driven change speeds the change cycle

The Kanban Method’s trick is that it does not treat change as a standalone activity ­– it is integrated into each of its most basic of practices, those of the sustainability agenda.

Of course there’s no inward-looking approach that can (or should) entirely protect teams from external realities. The values of customer focus, flow and leadership need be brought to the fore when internally focused change is failing to meet the expectations of people outside. In the second instalment in this series we’ll explore these values and the service orientation agenda to which they belong.

About the Author

Mike Burrows (Twitter:@asplake) is UK Director of David J. Anderson & Associates (djaa.com) and Director of Training Programs at LeanKanban Inc. He blogs here. His book Kanban from the Inside is due later in 2014.

Rate this Article

Adoption
Style

BT