What does the Kanban Guide for Scrum Teams (at https://www.scrum.org/resources/kanban-guide-scrum-teams) say about work in progress limits?
"Medicine had been 'successfully’ practised without the knowledge of germs. In a pre-germ theory paradigm, some patients got better, some got worse and some stayed the same; in each case, some rationale could be used to explain the outcome …” (1) (2)
Doctors practised what was taught
Doctors explained what they observed through practice using models that they learnt: Bad air, Contact, Generated by the body.
Follow these five steps to improve your system of work
1 Understand the end to end system of the value flow to the customer
- Study how the flow of work and visualise your work and map the flow
- Run the map, watch work items flow through the map
- Limit the number of tokens
2 Understand the system capabilities
- Types of work
- System lead time for fulfilling each type of work
- Arrival rate for each type of work
- Departure rate for each type of work
- Failure demand for each type of work, demand caused by failure to do something or to do something right for the customer – John Seddon
3 Understand your system variability
- For each type of work, measure
- Common cause variation, act on the system to change the common cause variation
- Assignable cause variation, act on the cause to remove assignable cause variation
4 Measure the flow efficiency for each type of work
- Measure the flow of each major knowledge discovery step
5 Purposeful change
- understand the root causes of the failure demand and tackle them
- Use the arrival, departure rate and the flow efficiency to target your improvement efforts and achieve better results for your investment
- Evolve your system of work and improve your system of work
We talk a lot about self-organisation, autonomy and empowerment. While this is a good start it is not sufficient to achieve the organisational goals and deliver the agility that the organisation is after.
Agile teams will learn and improve how they deliver their work through increasing the number of feedback loops, shortening the time to feedback, transparency of how they work and what they work on built on frequent inspection and adaptation.