One of the book’s most counterintuitive ideas: idle people are often better than idle information. Keeping people fully busy creates queues that starve downstream activities of information. Instead, deliberately keep some spare capacity to absorb variability and enable rapid response to new information. A 5–10% slack can cut cycle time by 30% or more.
Every feature or task delayed has an economic consequence. Cost of Delay quantifies how much value is lost per unit of time. Once you know CoD, you can: One of the book’s most counterintuitive ideas: idle
Reinertsen shows that queue size is directly proportional to cycle time – reducing WIP by half roughly halves cycle time. Reinertsen shows that queue size is directly proportional
Reinertsen argues that product development is primarily a network of queues (e.g., backlog of features, designs awaiting review, tests pending). Long queues increase cycle time, hide waste, and amplify risk. The key insight: utilization is not free. Running people at 100% utilization creates queues that dramatically slow throughput. The solution? Keep queues small and visible, and limit work-in-progress (WIP). backlog of features
This is perhaps the most counter-intuitive principle for traditional teams. Most organizations try to bundle large amounts of work into a single release to be "efficient."
Traditional top-down control fails in complex development. Instead, Reinertsen advocates: