I recently wrote an article for GOTO Magazine which has just been published its latest issue.
Given that it is embedded within the magazine, I thought I’d reproduce it here.
Steering the Agile Enterprise with Kanban Thinking
Large scale, enterprise organisations consist of many initiatives, programs, projects and development teams. While it is well understood how each individual team can achieve the benefits of agility, it is more difficult for the enterprise as a whole to achieve those same benefits. This article will look at how Kanban Thinking can inform the design of a Portfolio Kanban System, with a view to steering the Agile Enterprise to attain those benefits in order to respond to today’s competitive environments and unpredictable change.
Kanban Thinking is the framework I use when approaching the design of a Kanban System within a given context. The central concept is that the design approach is based on principles of Systems Thinking.
On the right side are the leverage activities for evolving the system design; study, share, limit, sense and learn. Studying the current system helps understanding of the existing context. Sharing that understanding gives everyone knowledge about what is happening. Putting limits in place bounds the system to help stabilise it. Sensing how the system is performing gives an understanding of current capability. Finally, learning from the systems performance allows its capability to be continually improved.
On the right side are the anticipated impacts the system design should have; flow, value and capability. Achieving flow involves a smooth, regular and frequent progress of work. Kanban Thinking looks to eliminate delays rather than eliminate waste. Delivering value involves focussing on doing the right thing in order to delight the customer (and other stakeholders). Kanban Thinking looks to maximise value rather than minimise cost. Building capability involves developing people and knowledge as a foundation for business success. Kanban Thinking looks to develop people as problem solvers rather than their tools to solve problems.
The Portfolio as a System
As a system, a portfolio is more than the sum of its parts – that is the initiatives, projects, features, teams and people – but is a product of the interactions of those parts with a particular tendency. Managing a portfolio, and as a result steering the enterprise, is the job of designing and guiding the portfolio such that its tendency is to have a positive impact.
First I will describe how the levers can be used to discover a Portfolio Kanban System design, and then show how to understand the impact of that design.
When studying an existing portfolio, it is useful to begin by identifying all the work that is currently known about, from early ideation, to already in production and being maintained. That work can then be clustered and arranged into themes based on similarity or relatedness. Common patterns which emerge are often based around investment and work item types, hierarchies and their governance workflow.
Examples of investment types might be areas such as architecture, urgent customer requests, current market segments, expanding market segments or future opportunities. Those investment types could be made up of initiatives, which are progressed through the development of features, which are iterated on by breaking down into stories.
Sharing a portfolio involves creating a model of the work which everyone can easily access and understand. Visual kanban boards provide a powerful mechanism for achieving this.
The most common approach is to create columns for the various stages of workflow that work items go through. For example, initiatives might begin as options, then have some discovery work done, then be assessed for suitability, then built and released before the results are reviewed for learning.
An important question to ask when designing the mechanism to share the portfolio is what you want to understand in order to learn. The TIP (Token, Inscription, Placement) heuristic is one I find useful to think about how to amplify the important signals, and dampen any noise.
Limiting a portfolio is the means by which it can be effectively steered. By limiting the number of initiatives or features being worked on, they can be completed sooner, and with greater predictability, allowing organisations to get earlier feedback and respond better to new information and changing market conditions.
Work can be limited at a number of levels. Investment allocation can help keep the portfolio focussed on the right mix of work. The number of initiatives and features can be limited to focus on finishing work before new work is started. Flowing work through stable teams can be used to balance demand against the capacity and capability of those teams.
As well as just limiting work in process through a Portfolio Kanban System, another form of limit is explicit policies. Creating transparency of the boundaries of the system design mean that the system can be stable within those constraints and the policies can be evolved to allow the system to evolve. A simple approach to policies is to add a checklist of exit criteria to each stage of the workflow.
Sensing the performance of the portfolio is what tells us how well it is being steered, so that direction can be adjusted effectively. There are generally two elements to sensing; establishing a cadence and measuring outcomes.
Establishing a cadence creates a sense of rhythm, and helps lessen the co-ordination cost of getting people together. Setting up a regular meeting to plan and review the portfolio enables a forum for gathering new information and feedback. A Portfolio Council is then able to respond by scheduling and re-prioritizing work in a timely manner and ensure focus is on the most important work.
Establishing appropriate measures generates insights which can aid decision making about what can be done to enable better outcomes. For a Portfolio Kanban System, financial measure seems to be appropriate. An economic model, which uses high level estimates of the costs and benefits of initiative or features, along with an understanding of the run-rate of teams, allows effective trade-off to be made between portfolio items. A timeline of planned and actual progress, for example, provides the basis for a rolling forecast as an alternative to annual plans.
Whatever choices you make, the design of your Portfolio Kanban System, and the work within it, it will be wrong! Learning, the detection and correction of those errors, is key to evolving the portfolio to have a greater impact.
Steering a portfolio is about updating the portfolio to match the reality of the current situation, rather than managing the portfolio towards a future situation. By sensing current performance with a regular cadence, work can be advanced, delayed or even killed to keep focus on the most important work.
Evolving a Portfolio Kanban System is about running deliberate experiments, with prediction and validation of how changes to the system design will affect its impact. Again, by sensing current performance with a regular cadence, visualisations, WIP limits and other explicit policies can be adjusted as knowledge is gained about what a better design should be.
Achieving flow across a Portfolio Kanban System means that the initiatives, programs, projects or features in the system progress with as few delays as possible. As a result the enterprise will be more responsive to the competitive landscape, so that when new information surfaces it can be reacted to effectively. Further, when flow is smooth across a portfolio, the work delivered by the system becomes more reliable as variability of lead times is reduced to within understood ranges.
Delivering value from a Portfolio Kanban System means the initiatives, programs, projects or features in the system are the most important things that could be worked on at that time. As a result stakeholder satisfaction will increase as customers get their needs met. Maintaining stakeholder satisfaction sustainably means that quality will also increase.
Building capability with a Portfolio Kanban System means that the right initiatives, programs, projects or features can be effectively delivered over the long term, and not just the short term. As a result employee satisfaction will increase, leading to greater retention of people and their knowledge. This long term building of people, teams and their skills will also lead to an increase in overall productivity.