The standard for project management

C

ASE

There is and probably always will be a lot of discussion around the Kanban Method and whether 

it belongs to the lean or agile movement. It was conceived in and around lean manufacturing, 

but is widely used in agile settings.

---

��

14 

  

Section 2

Requirements Uncer

tainty

High Uncer

tainty

Low Uncer

tainty

Technical Degree of Uncertainty

Low Uncertainty

High Uncertainty

Fundamentally

risky

Adaptive

approaches

work well here

Linear

approaches

work well here

Complica

ted

Comple

x

Cha

os

Simple

Figure 2-5. Uncertainty and Complexity Model Inspired by the Stacey Complexity Model

Teams can plan and manage projects with clear and stable requirements and clear technical challenges with little 

difficulty. However, as the uncertainty in the project increases, the likelihood of changes, wasted work, and rework also 

increases, which is costly and time consuming.

---

��

15

Some teams have evolved project life cycles to use iterative and incremental approaches. Many teams discover that 

when they explore the requirements iteratively and deliver more often incrementally, the teams adapt to changes more 

easily. These iterative and incremental approaches reduce waste and rework because the teams gain feedback. These 

approaches use:

u

u

Very short feedback loops,

u

u

Frequent adaptation of process,

u

u

Reprioritization,

u

u

Regularly updated plans, and

u

u

Frequent delivery.

TIP

What do simple, complicated, and complex projects mean? Consider large projects, such as the 

Boston Big Dig construction project. On the surface, the project seemed fairly straightforward: move 

the elevated highway underground. There was high agreement on the requirements (see the Y axis in 

Figure 2-5). There was low uncertainty on how the project would proceed until the project started. And, 

as is the case for many large projects, the project encountered surprises along the way.

When  a team works on a project where there is little opportunity for interim deliverables or little 

opportunity for prototyping, the team most likely will use a predictive life cycle to manage it. The team can 

adapt to what it discovers, but will not be able to use agile approaches to manage the iterative discovery 

of requirements or incremental deliverables for feedback.

The Big Dig project was not simple by any means. However, many projects that start out in the lower 

left part of the Stacey Complexity Model have no real means of moving to other approaches. Assess the 

project, both in the requirements and the means of delivery, to determine the best approach for the life 

cycle of the project.

---

��

16 

  

Section 2

These iterative, incremental, and agile approaches work well for projects that involve new or novel tools, techniques, 

materials, or application domains. (Refer to Section 3 on Life Cycle Selection). They also work well for projects that:

u

u

Require research and development;

u

u

Have high rates of change;

u

u

Have unclear or unknown requirements, uncertainty, or risk; or

u

u

Have a final goal that is hard to describe.

By building a small increment and then testing and reviewing it, the team can explore uncertainty at a low cost in 

a short time, reduce risk, and maximize business value delivery. This uncertainty may be centered on suitability and 

requirements (is the right product being built?); technical feasibility and performance (can this product be built this 

way?); or process and people (is this an effective way for the team to work?). All three of these characteristics—product 

specification, production capability, and process suitability—typically have elements of high uncertainty.

However, iterative and incremental approaches have their limits of applicability. When both technology 

uncertainty and requirements uncertainty are very high (the top right of Figure 2-5), the project moves beyond 

complex to chaotic. In order for the project to become reliably possible, it needs one of the variables (uncertainty 

or disagreement) to be contained.

---

��

17